T-(L MAR 2 9 1976 CORNELL UNIVERSITY LIBRARY ENGINEERING DATE DUE APR !: \m i GAYLORD PRINTED IN U S.A Cornell University Library The original of tiiis book is in tine Cornell University Library. There are no known copyright restrictions in the United States on the use of the text. http://www.archive.org/details/cu31924004454249 Gerard H Matthes STATE OF OHIO THE MIAMI CONSERVANCY DISTRICT Rainfall and Runoff in the Miami Valley IVAN E. HOUK District Hydrogiapher; Assoc. M. Am. Soc. C. E. TECHNICAL REPORTS Part VIII .%l % ^/ V'i \ . '■, DAYTON. OHIO V>!v;T;Tf3 ^,,. 1921 .:,.',,:.'''''" THE MIAMI CONSERVANCY DISTRICT DAYTON, OHIO EDWARD A. DEEDS, Dayton Chairman HENRY M. ALLEN, Troy GORDON S. RENTSCHLER, Hamilton Board of Directors EZRA M. KUHNS, Secretary OREN BRITT BROWN, Attorney JOHN A. McMAHON, Counsel ^^^^*^^*" !«' ^^ ', .' ARTHUR E. MORGAN, Chief Engineer iM CHAS. H. PAUL, Ass't. Chief Engineer PREFATORY NOTE This volume is the eighth of a series of Technical Reports issued in connection with the planning and execution of the no- table system of flood protection works now being built by the Miami Conservancy District. The Miami Valley, which forms a part of the large interior plain of the central United States and comprises about 4000 square miles of gently rolling topography in southwestern Ohio, is one of the leading industrial centers of the country. Out of the great flood of March, 1913, which destroyed in this valley alone over 360 lives and probably more than $100,000,000 worth of property, there resulted an energetic movement to prevent a recurrence of such a disaster by protecting the entire valley by one comprehensive project. The Miami Conservancy District, established in June, 1915, under the newly enacted Conservancy Act of Ohio, became the agency for securing this protection. On account of the size and character of the undertaking, the plans of the District have been developed with more than usual care. A report of the Chief Engineer, submitting a plan for the protection of the District from flood damage, was printed in March, 1916, in three volumes of about 200 pages each. After various slight modifications, this report was adopted by the board of directors as the Official Plan of the District, and was republished in May, 1916, under the latter title. This plan for flood protection includes the building of five earth dams across the valleys of the Miami River and its tributaries to form re- tarding basins, and the improvement of several miles of river channel within the towns and cities of the valley. It is esti- mated that the dams will contain nearly 8,500,000 cubic yards of earth ; that their outlet structures will contain nearly 200,000 cubic yards of concrete; that the river channel improvements will involve the excavation of nearly 5,000,000 cubic yards ; and that the whole project will cost about $35,000,000. At the time of the publication of this volume the flood con- trol works are about three-fourths completed. The Germantown dam and a considerable portion of the channel improvement 6 MIAMI CONSERVANCY DISTRICT work are entirely finished, and the remaining dams and channel work are rapidly approaching completion. In order to plan the project' intelligently many thorough investigations and researches had to be carried out, the results of which have proved of great value to the District and will also, it is believed, be of widespread use to the whole engineer- ing profession. To make the results of these studies available to the residents of the State and to the technical world at large, the District is publishing a series of Technical Reports contain- ing all data of permanent value relating to the history, investi- gations, design, and construction of the flood prevention works. The following list shows the titles of the reports published to date and the price at which they may be purchased. Part I.— The Miami Valley and the 1913 flood, by A. E. Morgan, 1917, 125 pages, 44 illustrations ; 50 cents. Part II. — History of the Miami flood control project, by C. A. Bock, 1918 ; 196 pages, 41 illustrations ; 50 cents. Part III. — Theory of the hydraulic jump and backwater curves, by S. M. Woodward. Experimental investigation of the hydraulic jump as a means of dissipating energy, by R. M. Riegel and J. C. Beebe, 1917 ; 111 pages, 88 illustrations ; 50 cents. Part IV. — Calculation of flow in open channels, by I. E. Houk, 1918 ; 283 pages, 79 illustrations ; 75 cents. Part V. — Storm rainfall of eastern United States, by the engineering staff of the District, 1917 ; 310 pages, 114 illustra- tions; 75 cents. Part VI. — Contract forms and specifications, by the engin- eering staff of the District, 1918, 192 pages, 3 folding plates, and index; 50 cents. Atlas of selected contract and information drawings to ac- company Part VI; 139 plates, 11 by 15 inches; $1.50. Part VII. — Hydraulics of the Miami flood control project, by S. M. Woodward, 1920; 344 pages, 126 illustrations; $1.00. Part VIII. — Rainfall and runoff in the Miami Valley, by I. E. Houk, 1921; 236 pages, 51 illustrations; 75 cents. Technical reports on the following subjects are contemplated. Laws relating to flood control. Structural design, construction plant and methods. Methods of appraising property benefits and damages. Orders for Technical Reports should be sent to : The Miami Conservancy District, Dayton, Ohio. Dayton, Ohio, ^^t^^^ E. Morgan, Jan. 1, 1921. Chief Engineer. CONTENTS Officers of the Miami Conservancy District 4 Prefatory Note 5 Contents 7 List of Illustrations 9 List of Tables H CHAPTER I. INTRODUCTION General 13 Scope of this report 14 Acknowledgements 16 The Miami Valley 16 Rainfall and runoff relations 18 CHAPTER IL RAINFALL AND RUNOFF RECORDS Records prior to 1913 22 Stations established since 1913 23 Records being secured 28 Gages in use 29 Discharge measurements 30 Stream flow records 33 Publication of data 33 CHAPTER III. MORAINE PARK EXPERIMENTS Description of plats , 36 Methods of measurement 37 Results of observations 39 Soil moisture 54 Surface runoff 71 Annual surface runoff 91 Summary 92 CHAPTER IV. SPRINKLING EXPERIMENTS Description of plats 95 Methods of experimentation 97 Results of experiments 98 Rainfall, retention, and runoff 106 Rainfall and runoff rates on saturated soils 119 Rainfall and runoff rates, soil not saturated 126 Conditions before runoff begins 130 CHAPTER V. MONTHLY, SEASONAL, AND ANNUAL RAINFALL AND RUNOFF Introductory 134 Compilation of the data 136 7 8 MIAMI CONSERVANCY DISTRICT Annual rainfall and runoff 138 Seasonal rainfall and runoff 150 Monthly rainfall and runoff 155 Surface and ground water flow 164 Mass curves 167 CHAPTER VI. RAINFALL AND RUNOFF DURING 1913 FLOOD Rainfall ' 177 Runoff 182 Relation of runoff to rainfall 186 CHAPTER VII. — RAINFALL AND RUNOFF DURING FLOODS SINCE MARCH, 1913 Rainfall, runoff, and retention during floods 190 Descriptive notes* 191 Total retention 203 Maximum values of retention 286 Rates of rainfall and runoff 209 CHAPTER VIII. FLOOD FORECASTING The present service 216 Reports being secured 217 Forecasting methods 218 APPENDIX Bibliog:raphy 227 LIST OF ILLUSTRATIONS Figure P^ce 1 Map of Miami Valley showing gaging stations 17 2 View of gaging station at Lockington Pacing 24 3 View of cable station at Tadmor Facing 28 4 Diagram showing Moraine Park records for 1915 Facing 40 5 Diagram showing Moraine Park records for 1916 Facing 40 6 Diagram showing Moraine Park records for 1917 Facing 40 7 Diagram showing Moraine Park records for 1918 Facing 40 8 Diagram showing Moraine Park records for 1919 Facing 40 9 Rainfall intensity at Moraine Park, July 7, 1915 80 10 Rainfall Intensity at Moraine Park, March 15-17, 1919 81 11 View of plats 1 and 2 at Taylorsville Facing 96 12 View of plats 3 and 4 at Taylorsville Facing 98 13 Mass curves showing experiment 1, level bare soil at Moraine Park 107 14 Mass curves showing experiment 3, level bare soil at Moraine Park 108 15 Mass curves showing experiment 4, sloping bare soil at Moraine Park 109 16 Mass curves showing experiment 5, plat 1 at Taylorsville 110 17 Mass curves showing experiment 6, plat 2 at Taylorsville Ill 18 Mass curves showing experiment 7, plat 3 at Taylorsville 112 19 Mass curves showing experiment 8, plat 4 at Taylorsville 113 20 Mass curves showing experiment 9, plat 1 at Taylorsville 114 21 Mass curves showing experiments 10 and 11, plats 4 and 3 at Taylorsville 115 22 Rates of rainfall and runoff at Moraine Park 120 23 Rates of rainfall and runoff at Taylorsville 123 24 Relations between rates of rainfall and runoff 1 24 25 Rates of rainfall and runoff, soil not saturated J 27 26 Intensity and duration of rainfall before runoff begins 132 27 Annual rainfall, runoff, evaporation, and temperature above Dayton • • • • 143 28 Departures of annual rainfall, runoff, evaporation, and temper- ature above- Dayton 144 29 Relations between annual rainfall, runoff, and evaporation above Dayton . - 145 30 Diagram showing monthly rainfall above Dayton 147 31 Effect of temperature on annual rainfall, runoff, and evaporation above Dayton 149 32 Seasonal rainfall, runoff, retention, and temperature above Dayton ^-52 33 Maximum, mean, and minimum monthly rainfall, runoff, reten- tion, and temperature above Dayton 162 9 10 MIAMI CONSERVANCY DISTRICT 34 Maximum, mean, and minimum monthly rainfall, runoff, reten- tion, and temperature above Dayton 163 35 Mass curves showing hydrology of Mad River Valley in 1915. . . . 168 36 Mass curves showing hydrology of Mad River Valley in 1916. ... 169 37 Mass curves showing hydrology of Mad River Valley in 1917. . . . 170 38 Mass curves showing hydrology of Mad River Valley in 1918. . . . 171 39 Mass curves showing hydrology of Mad River Valley in 1919. . . . i72 40 Curves showing monthly evaporation in Mad River Vajley 175 41 Maps showing daily rainfall during storm of March, 1913 178 42 Maps showing cumulated rainfall during storm of March, 1913. . 179 43 Diagram showing hourly rainfall during storm of March, 1913 . . . 181 44 Map showing maximum rates of runoff during 1913 flood 184 45 Hydrographs of 1913 flood at Piqua, Dayton, and Hamilton. . . . 185 46 Mass curves of rainfall, runoff, and retention during 1913 flood 188 47 Crest relation diagram for Pleasant Hill and West Milton 219 48 Crest relation diagram for Springfield and Wright 220 49 Crest relation diagrams for Sidney, Piqua, and Tadmor 221 50 Crest relation diagrams for Dayton, Miamisburg, and Franklin. . 223 51 Crest relation diagrams for Dayton, Middletown, and Hamilton. . 224 LIST OF TABLES Table Page 1 Stream gaging stations in the Miami Valley 26 2 Weight per cubic foot of Moraine Park loam 39 3 Rainfall, runoff, and soil absorption at Moraine Park 42 4 Maximum amount of moisture in Moraine Park soil in June, July, and August 58 5 Minimum amount of moisture in Moraine Park soil in January, February, and March , 59 6 Maximum rates of evaporation and transpiration at Moraine Park 62 7 Soil absorption at Moraine Park during summer storms 67 8 Soil absorption at Moraine Park during winter storms 70 9 Effect of slope and surface cover on runoff at Moraine Park .... 72 10 Surface runoff and soil moisture at Moraine Park during summer storms 75 11 Winter storms at Moraine Park which did not cause appreciable runoff 78 12 Rainfall intensities and percolation rates at Moraine Park 88 13 Rainfall, runoff, and retention at Moraine Park during storms 86 14 Annual surface runoff and rainfall at Moraine Park 91 15 Analyses of Taylorsville soil 96 16 Results of sprinkling experiments at Moraine Park. . ^ 100 17 Results of sprinkling experiments at Taylorsville Dam 102 18 Summary of results of sprinkling experiments 104 19 Intensity and duration of precipitation before runoff begins. ... 105 20 Stations used in studies of rainfall and runoff 135 21 Annual rainfall, runoff, and evaporation in the Miami Valley. ... 137 22 Annual rainfall, runoff, and evaporation above Hamilton 140 23 Annual rainfall, runoff, evaporation, and temperature above Dayton 142 24 Seasonal rainfall, runoff, retention, and temperature above Dayton 151 25 Monthly rainfall above Dayton 156 26 Monthly runoff above Dayton 5 57 27 Monthly retention above Dayton 158 28 Ratio of monthly runoff to monthly rainfall above Dayton 159 29 Monthly temperature at Dayton 160 30 Surface and ground water runoff in the Miami Valley 166 31 Monthly evaporation above Wright 176 32 Seconal evaporation above Wright ". . 176 33 Daily rainfall and runoff during the 1913 flood 189 34 Rainfall, runoff, and retention during flood periods 192 35 Rainfall intercepted by trees 205 36 Maximum retention during various floods 206 11 12 MIAMI CONSERVANCY DISTRICT 37 Maximum surface storage, during various floods 208 38 Storage in river channels during various floods 209 39 Rates of rainfall and runoff during various storms 210 40 Ratio of maximum 24-hour discharge to maximum discharge during various floods 213 41 Forecasted and actual conditions during flood of January 5, 1917 226 CHAPTER I.— INTRODUCTION GENERAL The purpose of this report is to present to the engineering profession the results of rainfall and runoff investigations car- ried on in connection with the Miami flood control project. When an engineering examination of the Miami Valley was begun, immediately after the great flood of March, 1913, in order to determine the best plan for preventing damage by fu- ture floods, an investigation of rainfall and runoff conditions was naturally one of the first lines of attack. It was recognized at the start that a knowledge of rainfall and runoff would be essential in determining the size of the flood to be provided for, in the design of the flood protection works, and in the assessment of the benefits and damages which would result from the con- struction of the works, as well as in the many other problems which probably would be encountered as the development of the plans proceeded. However, as the work progressed and as the magnitude of the problem became apparent, the importance of collecting such data became even more pronounced than had been originally anticipated. Consequently the collection of rainfall and runoff records and the studies of rainfall and runoff rela- tions were more or less gradually expanded during the first few years of the work. While there were several rainfall stations in the Miami Val- ley at the time of the 1913 flood, there were but three river gages, one at Piqua, one at Dayton, and one at Hamilton. The work of establishing additional stations was begun, in coopera- tion with the U. S. Weather Bureau, almost immediately; and within a few months daily records of rainfall and river- stages and periodic measurements of discharge were being obtained at several stations on the Mad and Stillwater Rivers, at German- town on Twin Creek, and at several additional places on the Miami River. Arrangements were also made with the various observers for special readings of river gages during flood periods. The number of stations and the amount of flood data 13 14 MIAMI CONSERVANCY DISTRICT being secured, was increased from time to time as the work progressed, as will be described in detail later. Extensive hydrographic surveys of the 1913 flood in the Miami Valley, and investigations of the rainfall over the valley during that storm, were carried on during the summer and fall of 1913. Studies of the relation of the flood runoff to the storm rainfall were made as soon as the data was available. Similar studies for subsequent floods were made from time to time as the floods occurred. As a practical aid in the study of the relation of runoff to rainfall, a number of small experimental plats were established at Moraine Park, about five miles south of Dayton, where rain- fall and surface runoff could be measured on varying slopes and with varying soil conditions, as well as the rapidity and depth of soil saturation caused by different rains. After about four and a half years of records had been secured experiments were undertaken, using garden sprinkling cans to reproduce rainfall effects, in an effort to develop a method by which rainfall and runoff relations could be determined for a given watershed without waiting the comparatively long time required for the collection of sufficient data from natural rainfall. The results obtained were so suggestive that similar plats were established at the Taylorsville Dam where data could be obtained on different types of soil. SCOPE OF THIS REPORT Chapter II describes the rainfall and runoff records obtained in the Miami Valley. The records available at the time of the 1913 flood, the stations established since that time, the records secured at the various stations, the gages in use, and the methods of measurement are all discussed in detail. The actual records are not reproduced since the more valuable data is being pub- lished elsewhere. However, the places of publication, the par- ticular records being published, and the manner in which the unpublished data may be secured are fully described. Chapter III takes up the rainfall, runoff, and soil moisture data secured on the small experimental plats at Moraine Park. The records are given in full, in tables and diagrams, and are discussed in detail. The effects of variations in rainfall intensity and in soil moisture content on the surface runoff are taken up, as are also the total rainfall, runoff, and retention during storm RAINFALL AND RUNOFF 15 periods. A summary of the principal conditions shown by the data is given at the end of the chapter. Chapter IV is devoted to the sprinkling experiments at Moraine Park and Taylorsville. The results are shown graph- ically, by means of mass curves. Summaries of the more im- portant data are given in tabular form. An interesting relation was found to exist between rates of rainfall, runoff, and re- tention when the surface soil is saturated. The total rainfall, runoff, and retention during the various experiments, as well as the rates, are discussed in detail; and some data is given re- garding the intensity and duration of precipitation before sur- face runoff begins. In chapter V the monthly, seasonal, and annual rainfall, runoff, and retention throughout the Miami Valley are taken up. Annual conditions in the different drainage areas are shown by means of tables and diagrams. Monthly and seasonal condi- tions are discussed only for the drainage area above Dayton since the records available for the other stations are of com- paratively short duration. A method of studying the hydrology of a valley by means of mass curves is shown, using the data for the drainage area of Mad River above Wright as an example. Discussions of the proportions of ground water runoff and flood runoff are included for the Stillwater, Mad, and Miami Rivers, and Buck Creek. Chapter VI discusses the rainfall and runoff during the great flood of March, 1913. The data is shown by means of maps and diagrams, but the complete station records are not included. The distribution of the rainfall as regards time as well as drain- age area, the characteristics of the flood hydrographs, and the relation of the flood runoff to the storm rainfall are discussed. Chapter VII takes up the studies of rainfall and runoff which have been made for floods occurring since March, 1913. The total rainfall, runoff, and retention during flood periods; the maximum rates of rainfall and runoff; and the maximum values of retention are given in tabular form and are described in the text. Data is also included relating to storage in stream chan- nels and on the ground, and to precipitation intercepted by trees. Brief descriptions of the various floods are given but the de- tailed rainfall and runoff records are not included. Chapter VIII contains a brief description of the flood fore- casting work of the District and of the methods used in making the forecasts. 16 MIAMI CONSERVANCY DISTRICT ACKNOWLEDGMENTS Acknowledgments are due the officials of the United States Weather Bureau, both in Washington and in Dayton, for their cooperation in establishing river and rainfall gaging stations and in maintaining records. All of the rainfall data used in this report except that obtained in the experimental work at Moraine Park, was secured from the Weather Bureau records. Acknowledgments are also due the U. S. Geological Survey, the U. S. Bureau of Soils, and other governmental bureaus. The engineering organization engaged on the Miami flood control project up to the time construction began was described in detail in an earlier report.* The investigations described in this volume were conducted by the writer assisted at different times by G. N. Burrell, H. W. Wesle, B. H. Petty, H. R. Dau- benspeck, F. E. Davis, and others. Professor S. M. Woodward in his capacity as consulting engineer for the District has been frequently consulted. The work has been outlined and super- vised at all times by Arthur E. Morgan, chief engineer. THE MIAMI VALLEY Rainfall and runoff conditions vary so widely with variations in geology, topography, and climate that it seems pertinent to give a brief description of the Miami Valley. As may be seen by referring to figure 1 the Miami River flows in a southwesterly direction through southwestern Ohio, entering the Ohio River at the Indiana and Ohio state line. It drains a rather fan shaped area of about 4000 square miles lying almost wholly in Ohio. The Whitewater River which joins the Miami near its mouth and which drains an area of about 1400 square miles lying almost entirely in Indiana, has not been shown since it is not affected by the works of the Miami Conservancy District. The Miami River is about 163 miles in length. Its drainage basin, which includes parts of 15 counties, measures about 120 miles on the longer axis and about 70 on the shorter. The more important tributaries below Dayton, following northward up the west side of the Miami, are: Indian Creek, emptying just above Venice; Four Mile Creek, a flashy stream entering just above Revo^s^%7rt°\l%h^^mI!^r ^°''*''°^ ^"t"^' ^^ ^- ^- ^ock, Technical pale 115. ' Conservancy District, Dayton, Ohio, 1918, RAINFALL AND RUNOFF n SAL *a40///a < 1 1 ^^ ^ i It ^ 1 IV ^ 9 ^^HR| " \ 1 ^^ 'Jh ' ■T ■ V ^ /IT ^ JLJi _^^^ . IS liBH o O iz; o o 1-1 H <; K O « o hJ iz; o Is o H <; H 02 o g 5 ) J'S :5' * - : : : - - ... - J o ra :: :: pj g «» • a -a JLnd m n ^- QgO = .- = Pd^. d •il^ ^ - n : ^ d^d ; = SS 03 S CO ^- CO ^ 1 COCOCOCO CO '^ CO CQ lis -§ i-H .A ■A 1-H .A II 111 lO lO uo CD CO CO CD CD CD CD -^ =^ L^ Q CO S 7 CO CN CM (M (M (N II III COCO coco CO CM CN (M (N C Oi CO CN ■^ woo O O ^ O O CD o s I>- m '* lO lO T-H 03 1— < CO OiOOOO 00 00 t^ CD s (N 1— W (N 00 CN CO (N C-J CO CO CO CO , !>. <-) OS CiOI>- CJO »^OS r^ 00O5 1^ 00 Oi CO 1— 1 1—1 3^ ;iS4 00 I-H i>-- 00 i-f 1 CO 1 Jt 11^ V lO lO CM t^ CD CO t^Oi-H C^ CO CN t- 1 T-H 1 O o O'-H g<^0 O-HOi-H o 1 o O o o lO OO OOiO iCOOiO lO O o CM CO ii: cc CDOOiC OO O r- (-•J « (^4 (M (N -^ CO CD COCO UOCO ■* IV -* T— 1 o o lO OO OOO 000»0 Q O CD IJO lO CD ^ O o OC l# c~. O CD Tt^ lO *0 o ^~* CO I-H <-l i-H CO COCO iNCNrJiCSI CO CO CO ^m O iM '<*' (M Oi 00 tM OOO • l>- t^ l>- W CD OO iO Oi lO CD I-H r- '"' •— t i-H |>,-t .,- ■1 r-( ,-( I-H 1-H i-H CM 1-H CD lO O '* -^ 00 CO OS O 1-H lO CD CM (N CO CO lO ?g-'S' TtlCD rHt^CM t^cD t^ O O ■'^ CD lO (N ■«* O CO 00 o l-E = Si COTt OiOOi COCO O CO '-H O CX) '^ '<*^ O 00 Tt< CO (N CO 3 N CN (> ■<*< t-i O 00 CD O (M 00 00 »0 CS CD CM T-H O "0 000000 t^ 1>- I>- t-- CO CD CD CD lOiO IC »0 "^ i-H OS O" t^ OS 13 TjHCniOOOCD 4 ^J> cn oc^ 03 i-H 4 ic lA ci ci -Ji 00 -4^ c»lO■4^ b* COIN !>. O CO Oi CO CD ,-1 CD CO ^ CO J, 4 # « * * * : : :"s : 'e • ^ •; 1M CC :pq fcl . ---flN rt -K a «J^ o & CO g«£fS|fS(S| '^ o me pq $ c 03 a 1 ^1 .a-S £••3 tag U) ..MM . cgHSS"=SS"= II o ^ a^lS-i-^ bOT3 »• ^ > s p^ - c5 1 = = = = : : : : : : = : : : 1 > : : ; : £ B 1 i C ^ 1 1 1^ g:i & T! >i >> OS o & S3 "a g 1 1 e09 RAINFALL AND RUNOFF 27 Drain- age Area Square Miles lO CO O r-(CO 00 lO »0 O lO CD GO CM -^ CD CD^ -^ C35 (M t-< i-H C0O(N 00 W t^ lO t^ t^ Oit^l-^ CJ t^ II "5 - "• ■" - - s: 11^ Q d III Q CO »— t 1 »o IM 1 CO CO CC CO CO CO CO ^ ^ i-H 1-H ^ .-H 1 1 1 II 1 m to lo lo lO lo (N OJ (N CN CM CM CO CO CO coco CO CO CO CO CO CO CQ CO CO CO rHl-Ht— (1— 11— ll-H.— 1 ,_(,_( 1 1 1 1 1 1 1 11 to »o ":> w3 lo »o lO lo lo CNCMCjtC^lCMC^C^ CMCM CO CO CO CO COOOCO CO CO 1 o CO to O 00 CO CM "t) t^ OO ^ CM OS O LO lO OOOCO o o Tt^ C^J ■<* GOCD1>-00 t^ OC ^ i-H .-H CM CM ^ --H ^ --1 1^^ 1 T 1 00 GO I>- 00 Cflt^ 00 1 1 n^ T'T 1^ 05 .-H t^ CDO CM 00 w CO t-CM o6 t- CTi CC' T T n^ t^ »o ^ T 7 "1' CM Oi 00 C:iCQC3i -^ 00 r-l ,-H rH »--4 -i-H Co4 4 CD IlO T-t 1-H CM Oii-H O ^ 00 s o 00 C^ O CO O CO 00 ■* O CD 1-1 o; O rl O coo .-H O O -H OCOO 1-H o 1 o CM Oi O OO O ^ "^ O (M OJ O '-< CM CM iOO O O -^ o -^ t^ lOt^O o o i-H CM .-H CM i-( CM C^ .-H ■o CM O O OO O T-) O ^ Oi t^ lO l-H CM ^ .-H io o o o iooo lO uo OC t^ CM CO ^lOO C* .-H 1— 1 .— 1 1— 1 1— 1 II 00 r^ CO t^ O t^ O GO b- CO CM C^Ol>G0 O t> Eleva- tion of Zerot g <35 "o r^ o 1— 1 en UD ca CM 00 ' CD i-H CD CM CO O OS ^ CTi t~ Tt< Ot^ --H CD ?3 UO CD CM O 00 t^ t- Tt* rH O GO 00 00 00 Oi GO t^ CO O CO lOOrtiCM CO 00 CO OC 00 ■* .-HCMi-ti-H CsJ rH 00 r-- t^ i> oot^t^t^ CO CD 1 1 CO T CO 4 un CD Tt< OO '"^ 'rj^ CQ 1^-' CO CO lO i ! 1 1 T 1 O tH ^ W3I>- CM CO -^ 00 CO CDOiTt^Tt* rt^ Ti^ ,— 1 .— t r-l i-H ,— 1 ,— 1 ,— ( 1— 1 ,— 1 ,-( 1 1 1 1 1 1 1 1 1 1 t^ 05 CO C^ CMi-HOS'^ CD l> CO ^ CO CD 4 -*j s X s ^ .3? t— 1 "E. < > 5 1 o "e i -3 2 O d; c CS a> < o a: 1 ^ H-J 2 i^ a; 3 & 3 < OJ X a ^ -.-5 .— J -^ s. aa ^ <: C!> o '"' -fj 1 -C" be ^ c. RAINFALL AND RUNOFF 29 In general, these readings are taken every hour during the rising flood, every two hours during the day following the time of maximum stage, and then three times each day until the water has fallen to about the stage existing before the rise began. Each day's readings are recorded on a special postal card form and mailed to the headquarter's office as soon as possible. Special reports from the greater number of both river and rainfall stations, for use in forecasting flood heights, are made direct to the Conservancy District during critical periods, as well as to the Weather Bureau. These reports are made by tele- phone or telegraph as soon as the rainfall amounts to 0.70 of an inch, provided it has fallen in 24 hours or less ; or whenever there is a sudden rise in the river stage amounting to three feet or more. A confirmation of each report is made by mail as soon as the message has been telegraphed or telephoned. These re- ports make possible the accurate forecasting of flood conditions and also furnish valuable information regarding flood runoff and storm rainfall. Rainfall measurements are recorded to the nearest hundredth of an inch. Where the precipitation is less than a hundredth of an inch the amount is indicated by a capital "T" meaning "trace." River gage readings are observed and recorded to the nearest tenth of a foot at all stations except Venice. At Venice, where the gage is of the Mott type, the observations are taken to the nearest hundredth of a foot. Readings to hundredths of a foot may be practicable at times during ordinary and low water stages where the stations are equipped with chain and weight or Mott gages, or with vertical staff gages graduated to hundredths. However, where the gages are of the vertical staff type, graduated to tenths only, it is doubtful if such precision is ever warranted, especially where the observers have had no technical training, as is generally the case. During flood conditions, or"if there is a strong wind blow- ing, the water will rise and fall, intermittently, from a tenth to a half a foot or more ; so that readings to hundredths, while me- chanically possible with certain gages, are accurate only to tenths of a foot at the best. GAGES IN USE Standard U. S. Weather Bureau rain and snow gages are be- ing used at all rainfall stations. The regular Weather Bureau station at Dayton is also equipped with tipping bucket gage. 30 MIAMI CONSERVANCY DISTRICT A chain and weight river gage is in use at the New Balti- more station^ and Mott tape gages at Piqua and Venice. The other river stations are equipped with vertical staff gages. Au- tomatic recording river gages of the electric transmission type are in use at Hamilton and Dayton. The engineers of the Miami Conservancy District prefer the vertical staff gage to any other type, leaving out of consid- eration the sloping gages which are so expensive that they are feasible only in exceptional cases. The principal objection to the chain and weight gage is that the chain gradually stretches, thus requiring the continual checking of the chain length and the correcting of the observer's reports. Another objection, which applies also to the Mott gage, is that the boxes, having a somewhat mysterious appearance, are frequently broken into and the gages damaged. DISCHARGE MEASUREMENTS Measurements of discharge are made by the District at all river stations in the valley except New Baltimore and the Ger- mantown, Englewood, Taylorsville, and Huffman dams. Meas- urements are not made at these places since they are close to the other stations and since the conditions due to the construction work are unfavorable for the securing of accurate data. The station at Venice was well rated by the engineers of the War Department, First Cincinnati District, during the flood of July, 1915. Measurements are made during flood periods and more or less periodically during normal or low water conditions. They serve to determine the relations between gage heights and dis- charge, thus enabling the calculation of station rating tables and the compilation of daily stream flow records. Moreover, the inspections by the hydrographers furnish checks on the accuracy of the observer's readings and also supply information regarding channel conditions, effects of vegetation on stages, and the like. Periodic measurements of discharge are also made on var- ious artificial channels, carrying water for industrial use, as follows : Miami and Erie Canal north of Fort Loramie Miami and Erie Canal Feeder at Sidney Tail Race at Slusser-McLean Company's Plant at Sidney Miami and Erie Canal at Lockington Miami and Erie Canal Feeder north of Lockington RAINFALL AND RUNOFF 31 Miami and Erie Canal at Piqua Tail Race at Waterworks Pumping Plant at Piqua Miami and Erie Canal at Troy Mill Race at Troy Miami and Erie Canal at Tippecanoe City Head Race at Tranchant & Finnell Mills at Osborn Miami and Erie Canal Feeder at Findlay Street, Dayton Miami and Erie Canal Feeder Wasteway below Findlay Street, Dayton Miami and Erie Canal at Warren Street, Dayton Dayton Hydraulic Company's Canal at Findlay Street, Dayton Miami and Erie Canal at West CarroUton Hydraulic Canal at West CarroUton Miami and Erie Canal at Miamisburg Tail Race at Grove & Weber Co.'s Plant, Miamisburg Tail Race at Ohio Paper Co.'s Plant, Miamisburg Tail Race at Miamisburg Paper Co.'s Plant, Miamisburg Miami and Erie Canal at Franklin Hydraulic Canal at Franklin Miami and Erie Canal at Middletown Hydraulic Canal at Middletown Miami and Erie Canal at Hamilton Hydraulic Canal above Reservoir at Hamilton Hydraulic Canal at Niles Tool Works, Hamilton Old River at Hamilton Head Race at Bentel Margedant Plant, Hamilton Wasteway at Ohio Electric Power Plant, Hamilton These measurements furnish the information needed in ben- efit and damage assessments as well as in design of local channel improvements. They also furnish the additional data needed in calculating total runoff at river stations. Gagings at some of the above sections have recently been discontinued due to changes made in connection with the construction work. All gagings are made with the small Price current meter, combination type, using the penta commutator whenever the velocities are so high that single revolutions of the meter cannot be accurately counted. Observations are taken by the two-point method whenever feasible. During low water conditions meas- urements are made by wading, using the six-tenths depth method if the water is less than two feet deep. During floods it is fre- quently necessary to resort to the surface method. In such cases coefficients of from 0.8 to 0.9 are used to reduce the surface 32 MIAMI CONSERVANCY DISTRICT velocities to mean velocities, the particular coefficient used in a given case being determined from a study of vertical velocity curves taken at the given station. Stay lines have been used in some cases. The two-point method of measurement has been tested by about fifty vertical velocity curves, taken at various locations among the fifty odd gaging stations, in artificial as well as natural channels. The average of the ratios of the velocity by the two-point method to the velocity determined from the curve was found to be 0.994 ; the average error of the two-point method, obtained by averaging, arithmetically, the differences between the ratios and unity, was 1.24 per cent ; and the maximum error for a single curve was 7.5 per cent. Gage readings are taken before and after each measurement, to the nearest half tenth of a foot wherever practicable. Sound- ings are recorded to the nearest tenth. Observations are taken in at least ten but not more than twenty vertical sections during each gaging, regardless of the width of the stream. If the ve- locity varies greatly across the stream the sections are spaced closer together than usual. In computing discharges from field notes the velocities and depths measured in a given vertical are assumed to represent average conditions in a width of channel extending, on each side, half way to the adjacent verticals. This method has been found to give fully as satisfactory results as the method of averaging velocities and depths in adjacent vertical sections to get the average conditions in the width of channel between sections. The latter method requires two operations not necessary in the former. Current meters are rated at least once each year, more fre- quently if necessary. However, the experience of the Miami Conservancy District has been that the ratings of individual in- struments, where the instruments have received proper care, seldom differ more than one or two per cent from the composite table furnished by the manufacturers. The meters were formerly rated by the Bureau of Standards at Washington. Recently, however, they have been rated in the river at Dayton, at a location just above an old concrete dam, where still water exists. Ratings are made with the meters sus- pended by cables and held in place by lead torpedo weights, the conditions being made as nearly as possible like those under which the meters are used. RAINFALL AND RUNOFF 33 STREAM FLOW RECORDS Daily stream flow records are being compiled for the follow- ing stations: Sidney — On the Miami River Piqua — On the Miami River Tadmor — On the Miami River Dayton — On the Miami River Franklin — On the Miami River Hamilton — On the Miami River Venice — On the Miami River Lockington — On Loramie Creek Pleasant Hill — On Stillwater River West Milton— On Stillwater River Springfield — On Buck Creek Springfield — On Mad River Wright — On Mad River Germantown — On Twin Creek Seven Mile — On Seven Mile Creek Four Mile — On Four Mile Creek The records are tabulated on forms similar to those used by the U. S. Geological Survey, one sheet being used for each year at each station. These sheets give the daily stages and dis- charges, the mean monthly discharges in second feet and in sec- ond feet per square mile, the monthly runoff in inches depth over the drainage area and in acre feet, the maximum and min- imum discharges for each month, the total runoff for the year, and the mean, maximum, and minimum rates of runoff for the year. The records are believed to be as accurate as it is feasible to determine such data on streams similar to those in the Miami Valley. They are, of course, more accurate for the larger streams having the flatter slopes than for the smaller streams having the steeper slopes. The records for the Four Mile Creek station are more unsatisfactory than those for any other station in the valley, due to the shifting of the control during floods. This shifting occurs during small rises of two or three feet as well as during the larger floods, owing to the sand and gravel deposits at the station and to the steep slope of the stream, about fifteen feet per mile. PUBLICATION OF DATA The daily rainfall records at Weather Bureau stations are published by the U. S. Weather Bureau in their "Climatological 34 MIAMI CONSERVANCY DISTRICT Data." The records at the Miami Conservancy District's stations are not being published. Daily gage heights at the river stations maintained by the U. S. Weather Bureau are published annually in their "Daily River Stages at River Gage Stations on the Principal Rivers of the United States." Summaries of discharge measurements, daily stream flow records, and descriptions of stations, for sta- tions where stream flow records are being compiled, are pub- lished by the U. S. Geological Survey in their water supply pa- pers. Records secured at the other river stations are not pub- lished. Discharge measurements made on artificial channels are not published except where the results are needed to determine total runoff at river stations. In such cases the results are pub- lished in the U. S. Geological Survey water supply papers. The data on rainfall, runoff, and soil moisture collected at Moraine Park is given in full in chapter III of this volume. The data on rainfall intercepted by trees is given in chapter VII. River or rainfall records secured by the Miami Conservancy District and not published may be obtained from the District at the cost of blue printing. CHAPTER III.— MORAINE PARK EXPERIMENTS In February, 1915, it was decided to make a series of field investigations of precipitation, surface runoff, and soil moisture at isolated plats of various characteristics, the object being to obtain data on the conditions under which surface, or flood, run- off takes place. For this purpose four small experimental plats were located in an orchard at Moraine Park, the home of Colonel E. A. Deeds, about five miles south of Dayton. It was recognized, of course, that these plats were too small and too few in number to be representative of the average conditions throughout the Miami Valley. In fact, it was known that the conditions are not typical. The area just south of Dayton consists of deep glacial deposits of sand and gravel, covered with a thin layer of surface soil, in the form of comparatively steep eskers and moraines; while the areas north of Dayton are slightly rolling glaciated areas with deeper surface soil underlaid by materials of various nature and geological age. However, it was thought that if a detailed study could be given to the rainfall and runoff condi- tions at selected places, by experienced observers, valuable infor- mation regarding the laws of runoff could be secured. For a study of the laws of runoff and the relation of runoff to rainfall small experimental plats possess certain definite ad- vantages over the much larger drainage areas which exist above the stream gaging stations. For instance the slopes of the ground surface within the plats, as well as the character of the soil and surface covering, can be accurately determined without making elaborate and costly surveys. In fact, the plats may be located so that definite comparisons can be secured between the runoff from areas having different surface conditions. Cer- tain questionable features pertaining to the larger areas are practically eliminated in the smaller, such as the absorption of runoff by the soil before it reaches the drains and the amount of runoff contributed by the ground water storage. While the Moraine Park experiments do not furnish conclusive evidence on all phases, of the subject, it is believed that the re- sults are worthy of presentation. 35 36 MIAMI CONSERVANCY DISTRICT DESCRIPTION OF PLATS Four plats, each five feet square, are located in open places in the orchard, two on level ground and two on a hillside, the two sets being about a hundred feet apart, and the two plats of each set being about ten feet apart. A standard rain gage was installed near each set. The plats on the hillside were placed where the slope of the ground is about eighteen feet per hundred. One plat on the hillside and one pn the level ground were located where the surface covering is a heavy blue grass sod. The other two were located where the sod had been removed leaving the soil bare. The upper two feet of soil on the hillside is a yellow, sandy loam containing some clay and gravel ; the upper two feet where the level plats were established is a similar material except that it contains a larger proportion of gravel. At both places the upper foot contains considerable humus. Of course the soil under the blue grass covering is practically full of roots, some of which extend to depths of 2 feet or more. The material underlying the 2-foot layer of loam, in both cases, is a mixed sand and gravel of glacial origin. On the hillside there is a fairly definite division between the loam and the underlying deposits. On the level ground the proportion of sand and gravel increases more or less uniformly with the depth below the surface until a depth of about two feet is reached. Below this depth the amount of silt and clay present is negligible. Mechanical analyses of typical samples of the surface soil taken on the level and on the hillside showed that the propor- tion retained on a quarter inch sieve is about 30 per cent, by weight, for the former and about 7.5 per cent for the latter. The analyses of the portions passing the quarter inch sieve, made by the Bureau of Soils, U. S. Department of Agriculture, gave the following results: Percentage by Weight Level Hillside Fme gravel, 2 to 1 mm 3.0 2.6 Coarse sand, 1 to 0.5 mm 11.8 9.8 Medium sand, 0.5 to 0.25 mm 12.2 9.0 Fine sand, 0.25 to 0.10 mm 29.6 28.0 Very fine sand, 0.10 to 0.05 mm 7.4 8.7 Silt, 0.05 to 0.005 mm 19.6 24*0 Clay, less than 0.005 mm 16.4 18.0 The plats were isolated from the adjacent ground by corru- gated iron strips set into the ground about eight inches and ex- RAINFALL AND RUNOFF 37 tending above the ground about four inches. In setting these strips care was taken not to disturb the ground inside the plats. Concrete was placed around the outside of the corners so as to prevent leakage at the joints. Of course it is quite possible that some water may creep down the inside edges of the iron strips thus slightly increasing the soil percolation. A galvanized iron tank, eighteen inches in diameter and four feet deep, to catch the surface runoff, was set in the ground just outside the lower corner of each plat, and was connected with the inside of the plat by a joint of three-inch sewer pipe, laid in concrete. A wire screen, to keep out vermin, was fas- tened over the upper end of each sewer pipe. The tanks were tested and found to be water-tight before being installed; and were tested at intervals after installation, no leaks being found at any time. They were provided with suitable tight fitting covers so that no water except surface runoff from the plats could be caught, and so that the evaporation within the tanks would be reduced as much as possible. The capacity of each tank is equivalent to a runoff of about 3.0 inches depth over the plat with which it is connected. Some trouble was encountered at times due to leaves stop- ping up the screens and causing the runoff to spill over the tops of the iron strips. This occurred mostly at the plat on the hill- side having the bare soil surface. The screens were later re- placed by wire mesh having openings about three-eighths of an inch square, after which more satisfactory records were obtained. The work of establishing the plats and installing the gages was completed March 4, 1915, and the measurements were be- gun the following day. METHODS OF MEASUREMENT Measurements of rainfall, runoff, and soil moisture have been made more or less regularly since the plats were estab- lished. The endeavor has been to secure observations just before and just after each rain, and also to secure measurements of soil moisture once or twice a week between rains, to determine the rates of drying of the soil. Owing to the pressure of other work it has not always been possible to adhere strictly to the above plans. Observations were also discontinued for short intervals during the winter months, as, for instance, during the severe winter of 1917 and 1918. During the first year the soil moisture determinations were made rather irregularly. Some- n MIAMI CONSERVANCY DISTRICT times samples were taken to depths of 18 or 24 inches, but more frequently they were only taken to depths of 12 inches. Smee March, 1916, however, samples have been taken systematically to depths of 24 inches. Where precipitation occurred on two or more days between successive readings of the gages it is possible to estimate the daily amounts at Moraine Park from the daily records taken by the U. S. Weather Bureau at Dayton. While such estimates may be considerably in error during summer thunder-storms it is not believed that they are greatly in error at other times. Val- uable information regarding intensities of rainfall is also fur- nished by the graphical automatic records being secured at the Dayton Weather Bureau station. Notes regarding rainfall and runoff conditions were made by the writer at his home in Carr- monte, about two miles north of Moraine Park. Rainfall measurements were made in the usual rtianner, that is, using the regular rain gage measuring sticks. The amounts of runoff were determined by measuring distances from the tops of the cans down to the water surfaces, using yard sticks, and reading distances to eighths of an inch. A depth in the can of an eighth of an inch corresponds to a depth over the plat of about 0.009 of an inch. The amount of moisture in the soil, under the sod and under the bare surface, was determined by taking samples, weighing them, drying, and reweighing. Samples weighing about a kil- ogram, or abont two pounds, were taken at intervals of about six inches in depth down to a depth of about two feet. During the first few months samples were taken close to the plats on the hillside as well as close to those on the level ground. Slightly different results under similar surface coverings were obtained at the two places, the moisture content of the soil on the hill- side generally being a little greater than that of the soil on the level grpund. This was probably due to the much larger proportion of gravel in the soil at the latter place. Later on samples were taken from beneath the sod and bare soil surfaces at a place on level ground, from 25 to 100 feet southeast of the level plats, where the soil was very similar to that on the hillside. The samples were placed in paper sacks and dried in the fur- nace room at Moraine Park. Harvard scales, reading to tenths of a gram, were used in weighing. Weights were tested and adjusted using standard scales of known accuracy. The sam- ples were dried and reweighed until their dry weight became RAINFALL AND RUNOFF 39 constant, before they were discarded. When the investigations were begun it was attempted to dry the samples by leaving them in small incubators which could be kept at a constant tempera- ture of about 100° Fahrenheit. It was found, however, that owing to poor air circulation in the incubators it required sev- eral weeks to dry the samples thoroughly. They were then placed in the furnace room, directly over the furnace, where they dried out in a few days; or, when the furnace was not being used, they were placed on shelves above a small coal water heater, which was used every morning. Determinations of the weight per cubic foot of the upper two feet of soil were made December 8, 1919. Samples were taken by boring down with a post hole auger, and were weighed, dried and reweighed in the laboratory at the headquarters of- fice. The cubical contents of the samples were obtained by weighing the amounts of dry sand of known density required to fill the holes from which the samples had been taken. RESULTS OF OBSERVATIONS Weight of Soil per Cubic Foot The data on the weight per cubic foot of the upper two feet of soil is given in table 2. Samples 1 and 4 were taken from beneath the bare surface on the level ground, where the samples for determining the moisture content of the soil have been taken regularly. Samples 2 and 3 were taken from beneath the sod surface on the level ground. Sample 5 was taken from beneath the sod surface near the plats on the hillside. Table 2. — Determinations of Weight per Cubic Foot of Moraine Paris Loam Sample Number Volume of Sample Weight when, taken Weight of Moisture in Sample Dry Weight of Sample Weight per cubic foot when taken Weight per cubic foot when dry Moisture in Sample 1 2 3 4 5 Cubic feet 0.410 0.652 0.710 0.510 0.713 Pounds 47.1 81.4 85.9 58.6 86.3 Pounds 9.9 15.4 13.6 8.8 13.1 Pounds 37.2 66.0 72.3 49.8 73.2 Founds 115 125 121 115 121 Pounds 91 101 102 98 103 Percent* 26.6 23.3 18.8 17.8 17,9 Average . . 119.4 99.0 20.9 *Based on dry weight. 40 MIAMI CONSERVANCY DISTRICT It will be noticed that there was about 20.9 per cent of moisture in the soil at the time the samples were taken ; that the weight per cubic foot of the soil when taken varied from 115 to 125 pounds, averaging 119.4 pounds; and that the weight per cubic foot when dry varied from 91 to 103 pounds, averaging 99.0 pounds. In order to simplify the calculations an average value of 100 pounds has been used for the dry weight in the studies taken up later. When the samples were taken there was, on the average, about 20.4 pounds or .33 of a cubic foot of water in each cubic foot of soil. While the upper two feet of soil at that time was about as wet as it ever gets under field conditions, the actual volume of the voids was probably a little greater than this. If a value of 2.7 is assumed for the specific gravity of the soil par- ticles, an average value based on several laboratory determi- nations, the weight of a cubic foot of soil particles would be 169 pounds, the volume of the particles in one cubic foot of soil in place would be 0.59 of a cubic foot and the volume of the voids in one cubic foot would be 0.41 of a cubic foot. Conse- quently the maximum amount of moisture that could be present in the soil would be 41 per cent by volume or about 25.6 per cent of the dry weight. Rainfall, Runoff, and Soil Moisture Table 3 gives the results of all observations of rainfall, run- off, and soil moisture, taken from the time the plats were es- tablished up to the end of October, 1919, about four years and eight months in all. Column 1 gives the date of observations. Columns 2 to 5, inclusive, give the moisture content of the soil under the sod covering, expressed as percentages of the dry weight. Column 6 gives the average moisture content of the soil under the sod covering calculated from the data in columns 2 to 5. Columns 7 to 11, inclusive, give corresponding data for the soil under the bare surface. Column 12 gives the observed rain- fall. Columns 13 to 15, inclusive, give the data on runoff from the plats having the sod covering. Column 13 gives the runoff from the level plat, column 14 gives the runoff from the plat on the hillside, and column 15 gives the average runoff from the two. Columns 16 to 18, inclusive, give similar runoff data for the plats having the bare soil surface. The maximum and min- imum records of soil moisture are set in bold face type so they can be easily located. __J^j^/y^J^A<^S^.-^Sepf.-J^Ocf.-^/ -May — y^June - .^rarrTTPTT cilTCUEED AT MORAINE PARK DURING 1915. FIG. 4.— RECORDS OF RAINFALL, RUNOFF, AND SOIL MOISTURE SEOUKJ!.!^ June PIG. 5.— RECORDS OF RAINFALL^ RUNOFF, AND SO Moisture in Soil in '% ofOryWeiffhf Rainfall and Runoff in Inches ^ ^ ^ Q INJ Q N ^ . ^ J 3 * \ \ tf I 1 ^ ' 1 > r- .■^; r .. " — ~" § S; < < ' 1 I ^ i\ :^ ^ ^ f ^ Y i- 1, :3 > 1 ► ^ r ff // ^ ^ "S a O 1 \ y \i'-->^ 1 r J" 5 i i 1 1 o m H c! en * 1 1 t Si. 1 \ < t - — T 1 1 '/ > ^ i ^ •^ ^ J_ - \ ^ — s ' 2 rf ^ y ^ H 7 /V s V 4i - Hh ir~ _ W 1 4r« 1 ^ Q s^^^ r O ^ ^y,. 1 § > J0 — ■■ ■ ' ^ ^-» ^^>Lk l_ o 4- j( r 1 '-' - ^ V -q O^ Tt ^l J i. 4 r f I* 1 1 . 1 1 s* r ' » "t ( o 1 i k ■III till 1 II' "• ^ "31 .a 1 CO O OOOOOOOOOOOOOOOOEhOEhOOOtHO a o 00 OS OOOOOOOOOOOOOOOOf-iOEnOOOrHp 1 CD T-HrH 000 0000 0000 00000f-(00000t-10 1 CO 2 1 OOOOOOOOOOOOOOOOfnOOOQf-lOO c 'a o rjt 1-! (M 00 oooooooooooooooooooooooo "3 1 ec OOOOOOOOOOOOOOOOf-iOOOOfnOO 3 s lOCD mC- T-IU5C0 t- 00 as CD "* CIS ^ U3 -* CO rt asOfHO <1 - OSIO coco T-H o ■ ^ • tH ■ O CO S 00 Oi co t- OJO C4 00 00 CM T-H lO ■ CO ■ T-H ■ * O 1 CO t- ot- tH t-H o T-H CD ■ ^ ■ T-H ■ * i o -•J t o o ■§ [/I is T3 a EV < ■ ■ T-H CD 00 00 oo tH t-H (M OOt-H-JI OOlOJ CD • N • T-H ■ O 05 CO-* CMIN •3 00 -* 00 CO .-(00 (M .-( >* T-H * CO T-H CO C0(MCO T-H T-H T-H * * ♦ i N CO •00 OCD * oo 00 t- tH r-i T-H * 00 T-H t- 005 00 * * *. o ■ ^ • T-H • > ! 1 <£> Cs] CDO rfCD rH I— < CDCC com T-H T-H o T-H * (NIOCO 05 coco * * * ■* IN [ ■* ■ tH ■ RAINFALL AND RUNOFF 43 I 00 OOOi-Hi-HNOOOOi-lOOOOi-tOOt-Oi-lOi-iOCOOCOOOrHOOO 1-H aiO(M cci-H CD oca t- (m oo >-( co (n Oi i-n t- oo lo o OOO^WCOOOOOr-HOOOOOaOt-lt-OrHT-lrHOt-rHCOOOi-lOOT-^ * * # CD OOrHOO CO ■** O (MIOO T-H W*!* 1-1 lO 1-1 OOOi-HOi-lOOOOi-HOOOOOOO^iOOOrHOtDONOOOOOO 1-H * 1-1 CQ 1-1 1-1 '^S* t- 1-1 00 1-1 t- N i-lTfrH O000f-(0000000000000(MOOOOOOSOOOOOOOO "* oq Tjt M N 1-1 Oi-ioo CO w*J* i-< OOOOOOOOOOOOOOOOOO'^Oi-tOOOCOOOOOOOOO CO OOOOfnOOOOOOOOOOOOOOOOOOOCDOOOOOOOO (M CO'o'^D (Mt>-^,-H i-HOO-«*^Dt-'^i-tC^OOi-lOfHW3Tj<{M-05Ot-0000t-00'«*C0C]"^ iTS^CDOit-CDCD 00"<#UD?D-«^lOWt-CD'OCJ OOS CDIOCDCDCDO ■^ CD • Oi CO 00 CO CD T}< «5 CD CO Oa N C- »-t 00 ^H CO 00 ''^ t-OSCDMOOO'^OS COOO t-OOOOUDWO CM O O m t- lO t- CO 00 CD CO 00 t* t- OO OS t- t- t- OS CO CO lO OS Ot- -lOOCM CDOQIO t-OOO mU3 t- oot- - CO o 00 1-H OS 00 00 o ca t> CO T-( 1-tT-H • I-H CM 1-t CM tH rH T-H CM 1-H i-H I-H tH COCDOO ,-HCO 'eJ t- CO OS t- COU30 1-1 T-H 1— ( T-H • . CM T-H i-H T-H 1-H i-t rH 1— I rH rH rH »-t r-f T|* lO lO OS CM LO iH i-t 00 CO CD CM OS CM t- CO U3 O OS t- '^ CO 1-1 -^ WOOS-^CDlftCD U5 O CD lO CO Cr- CO OS OS lO CO t- OS 00 CD CO t- tH ,-) rH ^ ^-^ rH i-( CM i-" i-l i-( rH ^H i-( rH rH i-l tH i-l i-l rH rH O CD CO CD OS CD i-H O CO O "^ OS 00 t- t- CM UO 1-1 00 t- rH Tj* CD -^ COOSOOCOmuSt- 000SCDCDt-t-m0SOSI>OSt-"osOSiOC0CD .r^£!?'~'53^S'~!°9-*^'^"'^t-OOOSCDCOi-(Ot-CMOSOOCMO a;s2^'^'^^cMC<'^'-''-|'-CDCD t- T-H T-H 1-H T-H 1-H T-H * * * o oos -^ lO OO -<# 1-H OS OS CO t- "^ t- ^ CO -"* U3 lO t- lO 00 iO o OOS OS t- O CO C^'^COrHCCCO'^OO'^'^OO IOC^COOOOOOOOSC<10'"* T— ( T-H I— 1 1— ( T-H T— 1 rH T-H 1-H T-H T-H T-H T-H rH a> oc- COi-HCOOOOlOOrHOT-10010lOC^!*-<:t<0SCaC0'<^C 00 C<1 OJ C<1 OS CO '<# T-H U3 OS O tr- CO '^^ -"^ t-OOOOCOCD t-CD ■ ococo t-Tf ■^ca waosooioc^ t- t-c-T-H t--^o CO -tr- .,_),_,,_( t-Ht-H t-HtH t-H t-H t-H t-H t-H t-H -t-H OOCDCDOOOOS T-H T-H rH CO rH rH N t-^ •50t-t-OOOOOOOOOOOJOS0501000'-l'-l'-iNNC -^OOCOrHt-CO JS T-H T-H rH 1-H CO CO -H 46 MIAMI CONSERVANCY DISTRICT lOi-HOOOO'^THOOOSOSOt-CCN'^'^OO'^OrHOO OO 00 CaCMt-i-l .-( CO CO 00 CD Tj< 1-1 OS O 1-H WNO COOOOOO-^i-HOOOOt-OCOCONt-CDi— (OWOt— lOO ■^ CM CD CJO O -^ O CO IC CO 00 t- Tf* O CD ^ CO CD t- OS -^ CD UD ^ ■^OOOi'<*"«*00(Mi-(00(Mt-TjCDC0Tj<-«*OC0i-HC0t-O'^00i-l'Oi05COTjOTj O^W«COT-HOLOi-(t-rH,-HC^?DOTitOiOstDlrHOO-«*' ?22S^?3'^^?2°°r:!'^^*^'»°o^^'^^^>•c--<*aST^CDlOlOTpcOCOOaU3'^OSCCOOCOOiOOCOl>OU301 !£!£22^^^^^S"^'"°^^'-^'^'^«=c»«»w3o^cocq(N 1-H »-(.-) T-H rH 1-H rH i-H rH N T-H i-H tH iH t-H i-H ,H y~H rH ,-H irH ^V*^ I 1''^'^ 1 1 ^<^77^7 , ^^C^ -HNCp .hS 2COCO^^-rJ<>^lOkOU5lOW3CDCDCDI^tlt-^-OOOOOOOOasCT>05 RAINFALL AND RUNOFF 47 00 OOOUSOlOOOOO 0(N CO i-( 10 1-1 N looooo 0000 0000000 00000 (M -<* F-l oooooooooo ■^ CD 1— 1 T-1 1— ( ,— ( ,_| T-l T-H t-i-] (M 1-1 CDfiSOOOW OQ O ■^lOc^T^OfnOcc CD Tj* lO Cr- UD Oi tr- 00 Oa CO 00 lO -^ (Mt-i-it>C> {MU3CDlOlOO.-lt-CO(MCg-^O0SCOO(MCOi-)lO-l • CD CO (M (M C=> CO lO 10 CO 1-1 CD 1-t U3 i-H CO lO 'ti' CO 00 00 CO 00 CO cot-'<:fi-*t-t-a5 1—1 1— 1 • UD CD 05 CO t- lO CD • tH T-H .— 1 rH rH 1— 1 i-H OOi-IOOQOCDCO'^OOtMOOOOOiOOOOt-CDCDOCD i-( (M 1-1 tH rH iH 1-1 1-1 .-H 1-t tH ,-H ,-( ^ 00 lOOO ■i-HOOt-lO-^ i-l(M (NQOO"«*001i-HlOCO.-HCD'"*CTiaiCD'<*'^'>*i- t- CO 00 CD O -^ - i-< i-H tH T— 1 rH 1— ( 1— 1 1— (OOOOt— OOCD'^OitNi— li— IU3i— I(M'<^IOIOCOCOCD-^ i£) ■^ (M (>3 CO -<* ^ CO CD CO lOlOCDC:'lCCat-CCOC:-lO'<*cocDoO'<*C']osini ,-<00 T-l ■ CO rH t- CO CO -^ CO - N 1-1 1-1 --* (M rH rH Oi-ICD-^OOCOlOOilOTHCO(MCOt-OOOCOast-CDU3 (M (M 1-H 1-H T-H rH rH ,-H rH rH rH T-( .-H i-l tH ,-H i-l i-( (N coo . CO 00 CO OC CD 1-1 «t-COlOOOOCgcO(35CnoOOOi-HCOrHCDOOOi-li-i 00 CD ■ 00 t- 00 CD t- »0 t- • i—j i—l rH 1—1 rH rH i—t TfoOOOOOt-t-'^05COOCOlO'*i-Hl005000CDiOCD N i—( 1—1 1—1 1—1 1—1 (M rH I— 1 T-H tH 1—1 1—1 rH 1-H i-H i-H rH iH ooCCU5!35 SNC^CO-^'<* t- t- CO t-fH CDOOCOOiHOOOi-COOOOOOOOO ooooc ' 1 < ooooooooooooooooooo ooooo bO 1 OOOOOOOOOOOOOOOOOOO ooooo 1 CO OOOOOOOOOOOOOOOOOOO ooooo 3 1 .s.s-g to a cq -^ «3 -^ Oi (M -^ rH O Oi as 00 O OS CO (N OS -^ (M (MOlOCOCDlOOi-H-^aSi-IONOOt-T-Hlrt-^cr) 00->*OO!D rH T-H rH r-( CO i— 1 & i .s 1 1 2 1 r-l COOOOt^lC(MTHlOCOrHl>CO-^CD005COir3«D CO 00 CO t--^ T— < 1— ( i—( tH 1-H rH rH rH rH tH 1—1 1-H 1-H T-H T-H »H I— ( o 1-i CDCOCOOSasOSOOt-OOi'^l'CONOOCOOCOlOCD O^DC^WW lOlOCOOOCOUSCOCO-^^lOW'^'^lOt-OiOlO 00 OOCOCO U5U3 00O5-** lOiC'^CO'XHO-^'^lOiX>CDlOI><:DCDOOt-?D 00 t- CD CD t- i-t T-< r-t T~< tH 04 « coco CO OO OS 00 t- C^ CO Oi O -^^ '^ CO rH ■ O CO 00 5D 1-H OS rtCO^OJOO t-U3 CJ CO CO Tj< CO Tl< t- CO ^ Od CO t- lO OS 00 OS T-Hi-HOO t-O 1 i o O 2 so CO O OS O rH O CO -^Ot-COOilO OS N'«*0'^0 CDQOCONOO TjOOOOS kA iOCONCOCOOSCOCO'^USC^t-NlOOOC^IMrHCO t-coou5eo Tf<'^OOt-OSt>OOt-!>CDt-COCOCDt-00 00 00 OOOOO k 00 -* OCOOCOCOCDCOlOTt*NOOOCDCOCOi-H'«;JOOb-t>t-COt-OSOOOOSOS ooeoeqr-ic>i ■-HNINOJC^ 1 N CO U3lOCOW3 00rHTtCOt--<1COCOG 00 lO -^ lO CD t* 00 t> -^ OJ W U3 lo ui CO d N CO U5 '^t ^' tJ (>■ ui (xi t>^ OS tH tH tH rH rH rH rH 1-^ rH ^ ,-( rH tH ^ 1-J in ^ S OCOCOt-CD O t- t- U3 t- s 5 ^ RAINFALL AND RUNOFF 49 3 >£> -^i-H-^ 00 N i-H t- Tj< (M O T-< -^O CO i-HM W 00 CO ^ W 00 OOOOOOOOi-tOrHOO(Mi-tOO^CJsOOC0rHOOOt-t-(NOOO»-H^ 1-1 t- «rH-^ 00 CO CM N W3 "^ t- OSi-H Oi 00 Tf t- 00 iH rH r-t OOOOOOOOi-tOi-iOOWi-lOOr-IOiOO'^OOOOOOt-WOOOt-t''* CO -^ CO CC 00 1- O W* Oi Tt< 1-H OOO ■'^t' i-H CO O O '^ "Tji -^ OOOOOOOOi-lO'rHOO»-t-CCOl0C0O00OlMl0t-Tt<00CDC0t- ,H (M ^ i-H »-H 1-1 1-1 r-4i-l CMOiooajt-cot-t-cococooooo o '^"^00rHCU5COt-tr-COCOt-t--^OOaiOOCOt:-ast-COtr-OOtr-t-COtr-U3COCOCO(NU^ia, OS lOOOOt-COCt-Q0 00lO^i-Hi-(O0S00aj0it-00O00t-00 0S0000'*:i OOO i-H050a500SOOOtr-OOi-HOt-05 1-100500:1 rHOt-t-OOOOi'^'^'^OO Cq,-HCMrH<>lTHWrHCOU5U3COt-OCO(Mt-i-llOCMT-ICONOt-COOt-05W3COt- O 00 05 t- OOCD OS 0005 1-1 CO t- OiOi OO CD t-t-t-rj* CO 00 t- CO CO OCC^WOOaOi OS (^jTHTHrHi-lrHi-(i-(i-IOai-l'-lrHi-INCMi-lrHiHT-(rHi-li-liHC^i-lC i-lt-u^TJ^cocOl-lC<^"«*■^0(^^ooaot>c3l-^o^co (M(MCq OI>^oioOOOOr-ioOi-(OOt-0'-lOOS-^Ot-tCO-^OOOOiOOWi-]i-jU3'<1* >• 1 N ^oooc<^'^cOT-lTHOlOOal-^oooccwcO'-t005CM'-lC<^oso^ooooscoOiOl-^"«*o^u^ T3 S i-i gc!lciciciDL.cici4.44.44^uLuLJ5uLcicitllitit^iitl.o!>o!>otcii^^ S^ so MIAMI CONSERVANCY DISTRICT moisture under the bare surface are connected by dotted lines. These connecting lines have been drawn to aid the reader in studying the soil moisture under a given surface cover, from point to point across the diagrams. No attempt has been made to show the daily fluctuations. The lines have been drawn prac- tically straight from point to point, whereas, in the case of an increase in the amount of moisture, the line, in order to show the conditions accurately, should have been drawn on a slightly downward slope until the day of heavy rainfall and should then have risen more or less abruptly to the higher percentage. Although the diagrams were prepared by platting percentages as ordinates against dates as abscissas, scales showing the ac- tual amounts of water in the soil in inches depth, for the two foot depth of soil involved, have been added at the edges of the sheets. Notes to Accompany Table 3 On account of the condensed form of table 3 it has not been possible to include descriptive notes. Since such notes are im- portant in any study of the individual records, they are repro- duced, herewith, arranged chronologically so that any date can be easily located. Asterisks have been inserted in table 3 after the quantities for which descriptive notes are available. March 6, 1915. — Average percentage based on two samples taken at this depth. March 16, 1915. — Average percentage based on three sam- ples taken at this depth. Values given for depths of 18 and 24 inches are questionable and have not been used in computing the average. April 19 and 28, 1915. — Average percentage based on two samples taken at this depth. May 3, 5, 22, and 28, 1915. — Average percentage based on two samples taken at this depth. May 30, 1915. — Runoff was some greater than value given, due to clogging of sewer pipe. June 4, 1915. — Average percentage based on two samples, taken at this depth. July 29, 1915. — Runoff was some greater than value given, due to clogging of sewer pipe. Sept. 7, 1915. — Ground within these plats spaded thoroughly to a depth of about six inches. RAINFALL AND RUNOFF 51 Nov. 30, 1915. — Light snow on ground. Ground frozen to a depth of about one inch. Dec. 11, 1915. — Sleeting at time of observation. Jan. 6, 1916. — Soil under bare surface frozen to a depth of about three inches. Jan. 14, 1916. — Soil under bare surface frozen to a depth of about five inches. Jan. 25, 1916. — Runoff due to rain falling on frozen saturated soil surface. Feb. 2, 1916. — Soil under the bare surface frozen to a depth of about two inches. Feb. 17, 1916. — Soil under the bare surface frozen to a depth of from one to four inches, but soft on top. Soil under sod not frozen. Feb. 28, 1916. — Soil under the bare surface frozen to a depth of about five inches. Soil under sod frozen to a depth of about four inches. Mar. 9, 1916. — Ground thawing out. Mar. 16, 1916. — Soil under bare surface frozen to a depth of about a half an inch. Soil under the sod frozen to a depth of about an inch. About six inches of snow on the ground. May 12, 1916. — Ground within these plats spaded thoroughly to a depth of about four inches. June 3, 1916. — Record probably low due to runoff overtop- ping side of plat at lower corner. June 19, 1916. — Surfaces within these plats covered with a dense growth of white clover and bluegrass about twelve inches high. July 5, 1916. — Grass and weeds removed from these plats, and ground spaded. Aug. 7, 1916. — Runoff from sloping plat with bare soil sur- face probably low due to leakage. Grass in sod covered plats about six inches high. Sept. 6, 1916. — Runoff record for level bare soil plat prob- ably too low due to can overflowing. Sept. 9, 1916. — Gravel encountered at depth of twenty-four inches under bare surface. Blue grass in sod covered plats three to six inches high. Sept. 22, 1916. — Ground within bare soil plats spaded to a depth of about three inches. Blue grass in sod covered plats about five inches high. 52 MIAMI CONSERVANCY DISTRICT Nov. 20, 1916.— Grass in sod covered plats about tw^o inches long and dying. Dec. 4, 1916. — Samples not taken under sod on account of rain. Dec. 27, 1916. — Runoff due to melting snow, and to rain falling on frozen ground. Jan. 8, 1917.— Soil not frozen. Dead grass from one to five inches long in sod covered plats. Jan. 22, 1917. — Soil under bare surface frozen to a depth of about three inches. Bare soil plats about half covered with snow about a half an inch deep. No snow on sod plats. Jan. 31, 1917. — Soil not frozen. May 5, 1917. — Data questionable, results not used in com- puting averages. May 25, 1917. — Grass two to twelve inches long on sod cov- ered plats. Weeds removed from bare soil plats. June 30, 1917. — Record may be slightly low due to leakage. July 14, 1917. — Record uncertain due to clogging of tile en- trance. Dec. 3, 1917. — Owing to the unusually severe winter weather no records were taken during the remainder of this month or during the month of January, 1918. The total precipitation in January, two-thirds of which was snowfall, was 3.46 inches at the Dayton U. S. Weather Bureau Station and 3.87 inches at the Dayton cooperative station. Feb. 11, 1918. — Snow and ice practically gone, ground frozen to a depth of about eight inches. Records of runoff uncertain due to overtopping of cans and due to water entering plats from outside snow accumulations. Precipitation measurements uncertain; value of 5.26 inches given was observed at the Day- ton cooperative station. Feb. 26, 1918. — Record low due to can overflowing. May 14, 1918. — Record probably low due to clogging of tile entrance. June 26, 1918. — Placed wire mesh over entrance to tiles. July 17, 1918. — Sample disturbed in drying. July 25, 1918. — Record probably low due to leakage around tile. Aug. 23 and 28, 1918. — Sample disturbed in drying. Oct. 3, 1919. — Grass about eight inches long on level sod covered plat, and about two to six inches long on sloping sod covered plat. RAINFALL AND RUNOFF 53 It is believed that all records of soil moisture obtained dur- ing the summer and fall of 1919 are slightly high compared with the preceding records, due either to taking the samples a little farther away from the plats, where the soil and topography ■were slightly different, or to getting them more thoroughly dried. Accuracy of the Data The data is believed to be sufficiently accurate for the pur- poses for which it was collected. Although difficulties were en- countered, particularly during the first year, they were finally overcome in most instances. The soil samples were taken for the purpose of learning about how much water was in the ground before and after rains, especially when runoff occurred, rather than for making thor- ough studies of soil moisture. Consequently the records should not be used indiscriminately in any detailed study of the subject. Any individual value given in table 3 may be considerably in error. Due to the difficulties encountered in taking, drying, and weighing the samples, no single percentage is probably accurate to within less than one per cent ; that is, a value given as 5 might actually be 4 or 6, or a value given as 20 might actually be 19 or 21. Possibly a few of the minimum records are low due to not getting the samples entirely dry. It must also be remembered that on account of the slight differences in soil texture and the variations in surface configuration within the limited area in which the samples were taken, samples taken on different dates may not be strictly comparable. Probably this effect is even more important than the errors in observation. These condi- tions, however, are not so important in considering the average moisture content of the upi)er two feet of soil, as given in col- umns 6 and 11 of table 3. The precipitation records are believed to be accurate in all cases except where the greater part of the precipitation occurred as snow. The two rain gages were well located with respect to obstructions and the readings generally checked to within .05 of an inch. The runoff records for the level plats and for the sloping sod plat are believed to be as accurate as the precipitation rec- ords, except where the runoff was caused by the melting of large quantities of snow. There is no doubt but that all of the runoff from these plats entered the runoff tanks and that the amounts were accurately measured. The depths in the runoff tanks could 54 MIAMI CONSERVANCY DISTRICT easily be measured to eighths of an inch, corresponding approxi- mately to hundredths of an inch on the plats. The records for the sloping bare soil plat are somewhat uncertain in several in- stances, as indicated in the preceding notes. SOIL MOISTURE The records of soil moisture, given in table 3, furnish in- teresting information regarding the conditions at this particular location and pertaining to this particular soil. From the per- centages given in table 3 and the weight per cubic foot of the soil, given in table 2, it is possible to discuss the dryest condi- tion which the soil ever reaches, the maximum amount of water that the soil can contain, the maximum amount that it can hold against the force of gravity, the variations throughout the year, the amount of water absorbed during rains, and the rate al which the ground drys out after the rain ceases. Minimum Records A study of the records given in table 3 shows that during the length of time covered by the observations the soil was dryest on August 2, 1916. The determinations made on that date showed an average moisture content of only 4.7 per cent for the soil under the sod covering and only 3.1 per cent for the soil under the bare surface, amounts corresponding to 1.80 and 1.19 inches, respectively, for the depth of two feet in which the sam- ples were taken. Although these values may be slightly low due to not getting the samples thoroughly dried, it is known from other information that the ground at this time was baked hard and was very dry, probably as dry as it ever gets. The water in the soil was probably all hygroscopic water. It is unlikely that any further appreciable evaporation or transpiration could take place. Practically all vegetation, including the larger bushes and trees, had been wilting for several days. The wilt- ing coefficient, calculated from the mechanical analyses given previously, by the methods explained on page 69 of Bulletin 230. of the Bureau of Plant Industry,* would be about 12.3 per cent for the sample taken near the level plats and about 13.7 per cent for the sample taken near the sloping plats. Although values as high as these, and higher, are given for loam and clay loam *The Wilting Coefficient for Different Plants and its Indirect Deter- mination, by Lyman J. Briggs and W. L. Shantz, Bulletin 230 of the Bureau of Plant Industry, U. S. Department of Agriculture, 1912. RAINFALL AND RUNOFF 55 soils, in the above mentioned publication, these values seem slightly high for the Moraine Park soil. Observations showing values smaller than these were made at several times when no evidences of wilting could be detected and when there was no reason to believe that the records might be low. According to Bulletin 230 the values of the hygroscopic coefficient would be 0.68 times the values of the wilting coefficient or about 8.4 and 9.3 per cent respectively for the two samples. The records at the different depths on August 2, 1916, were as follows : Depth in Inches 6 12 18 24 Ave. Moisture under sod, percent 3.5 4.7 4.5 6.1 4.7 Moisture under bare surface, percent- 2.1 2.8 3.5 3.9 3.1 It will be noticed that the amount of moisture in the soil increased as the depth increased, for both types of surface cov- ering; also that the amount under the sod was greater, at each depth, than the amount under the bare surface. The percent- ages at the different depths were, themselves, minimum values for the entire period of record, in all cases except at the depth of 24 inches under the bare surface where a value of only 3.1 per cent was obtained on July 24, 1916, the preceding date on which samples were taken. It is probable, however, that the soil at this depth was actually drier on August 2 than on July 24, and that the opposite condition shown by the data is due to errors in observation or in securing comparable samples. The amount of moisture in the soil was also very low in August, 1918, the measurements of August 19 showing the fol- lowing percentages: Depth in Inches 6 12 18 24 Ave. Moisture under sod, percent 6.1 5.5 5.5 6.1 5.8 Moisture under bare surface, percent. 4.5 6.9 6.9 7.9 6.5 It will be noticed that on this date there seemed to be a little more soil moisture under the bare surface than there was under the sod. Under the bare surface the percentage of moisture seemed to increase with the depth, while under the sod it seemed to be about the same at ail depths. It is interesting to note that Widstoe and McLaughlin in their experiments in Utah,* found that in one instance the amount of moisture in the first foot of soil on which, crops were *The Movement of Water in Irrigated Soils, by J. A. Widstoe and W. W. McLaughlin, Bulletin 115 of the Utah Agricultural College Experi- ment Station, Logan, Utah, May, 1912. 56 MIAMI CONSERVANCY DISTRICT growing was reduced to 5.64 per cent, 40 days after irrigation; and that the amount in the first foot under the bare surface was reduced only to 18.6 per cent, 36 days after irrigation. The value of 5.64 per cent is only about one and a half per cent greater than the minimum Moraine Park record obtained in the first foot of soil under a blue grass sod. However, the value of 18.6 per cent is rather large compared with the value of about 2.5 per cent obtained under the bare surface at Moraine Park. Although there are some differences in soil texture, the real rea- sons for this wide difference in eva-poration are probably the greater percentage of voids in the Utah soil and the differences in the climatic conditions at the two locations. At Moraine Park the percentage of voids in the soil, by volume, is only about 41 while in Utah, where the above experiments were made, it is about 55. In Utah the climate is arid, while in Ohio it is humid. The differences in soil evaporation due to differences in cli- mate were discussed by Buckingham in 1907.* He showed that a moist bare soil in an arid climate dries out rapidly at the sur- face at first, forming a sort of a dry soil mulch, after which it dries out very slowly; that a moist bare soil in a humid climate dries out less rapidly than in the arid climate at first, so that the dry mulch effect is not produced, and more rapidly later on ; the net result being that after several days more water had evaporated from the soil under humid conditions than had eva- porated from the soil under arid conditions. Maximum Records The maximum percentages of moisture at the different depths, as shown by the data in table 3, occurred on different dates, although some uncertainty exists in this connection due to the difficulties encountered in securing comparable samples. The actual maximum values, not considering a few erratic ob- servations which have been mentioned in the notes as being ques- tionable, are as follows : Depth in Inches 6 12 18 24 Moisture under sod, percent 24.8 23.3 23.2 23.6 Moisture under bare surface, percent- 23.2 21.8 21.9 21.5 These values seem to indicate that the soil under the sod at a given depth never contains more than about 24 per cent of moisture, and that the soil under the bare surface never contains *Studies on the Movement of Soil Moisture, by Edear Buckingham. Bulletin 38 of the Bureau of Soils, U. S. Department of AgricSe, 1907! RAINFALL AND RUNOFF 57 more than about 22 per cent. In the preceding discussions it was shown, by calculations based on the specific gravity of the soil particles and the determinations of the unit weight of the soil in place, that the soil would be saturated when it contained an amount of moisture equal to about 25.6 per cent of its dry weight, an amount slightly greater than those given above. The records seem to indicate that the total amount of mois- ture in the upper two feet never is more than about 21 per cent of the dry weight of the soil, an amount equivalent to a depth over the surface of 8.06 inches. Samples were taken at several times during the months of January, February, and March, when the soil was probably as nearly saturated as it ever be- comes under field conditions. The slight differences in moisture content at the same depth shown by the data at such times are probably due to the difficulties encountered in securing compar- able samples or in weighing and drying those taken. The ob- servations which gave the maximum average values for the up- per two feet are as follows : Depth in Inches 6 12 18 24 Ave. Moisture under sod, percent 18.2 19.7 22.8 23.4 21.0 Moisture under bare soil, percent_-_ 19.7 21.6 21.6 21.5 21.1 The values for the soil under the sod were obtained on March 10, 1919. Those for the soil under the bare surface were ob- tained on March 24, 1917. On these dates the percentage of moisture seemed to be slightly greater at the greater depths un- der the sod, but did not differ materially at the different depths under the bare surface. The average value of 21 per cent shown by the above data probably represents the maximum amount of water that can be held by the Moraine Park soil ; that is, the maximum amount of moisture that can be present without any appreciable down- ward percolation due to gravity taking place, — the quantity fre- quently referred to as the "moisture-holding capacity." That this is true is indicated, in a way, by the observations of January 24 and February 4 and 11, 1919. The average percentages of moisture found on these dates were as follows : Sod Bare January 24, 1919 19.8 18.8 February 4, 1919 19.3 17.3 February 11, 1919 19.2 16.7 The total loss in moisture in the 2-foot depth during the 18 days from January 24 to February 11, indicated by these per- 58 MIAMI CONSERVANCY DISTRICT centages, would be equivalent to a depth in inches of 0.23 for the sod and 0.80 for the bare soil. The total precipitation dur- ing this period was 0.08 of an inch, thus increasing the amounts of moisture to be accounted for to 0.31 and, 0.88 inches, respec- tively, or to 0.017 and 0.049 inches per day. As the weather during the greater part of this period was clear with tempera- tures above freezing and some wind blowing, it is quite likely that these amounts represent soil evaporation alone and that consequently no material percolation occurred. While the amount of moisture that can be held by the soil undoubtedly varies widely with its composition it is interesting to note that Widstoe and McLaughlin, in their investigations in Utah, previously referred to, found that the maximum amount of water that could be held by the Greenville soil under field con- ditions was a little less than 24 per cent. Variations in Soil Moisture The variations ^n the amount of moisture in the soil at Mo- raine Park throughout the year are shown graphically by the curves in the lower parts of figures 4 to 8, inclusive. The amount of moisture under the sod is shown by the continuous lines and the amount under the bare surface is shown by the dotted hnes. A study of these diagrams shows that the soil is generally dryest in the late summer or early fall, during the months of July, August, or September ; and wettest in the late winter or early spring, during the months of January, February, or • March. It has already been pointed out that the minimum val- Table 4. — Maximum Percentages of Moisture in the Upper Two Feet of Soil at Moraine Park During the Months of June, July, and August Year Moisture under Sod .in% Moisture under Bare Soil, in % June July August June July August 1915 13.3 14.8 17.2 13.6 18 11.9 8.3 17.0 16.4 15.6 "i^'g" 13.1 15.0 18.7 14.4 15.2 13.3 12.5 19.5 14.0 9.0 13.1 13.2 19.6 ■■14:2" 11.4 15.8 18.9 1916 1917 1918 1919 ues for the entire period of record were obtained in the month of August, and that the maximum values were obtained in the month of March. The curves also show that the amount of RAINFALL AND RUNOFF 59 Table 5. — Minimum Percentages of Moisture in the Upper Two Feet of Soil at Moraine Parit During the Months of January, February, and March Year Moisture under Sod ,in% Moisture under Bare Soil, in % January February March January February Marcli 1916 16,4 17.0 16.6 17,6 18.5 19.2 17.8 18.8 16.6 18.6 14.4 15.1 "is.k" 14.9 16.8 18.9 16.7 15.9 17.8 18,1 17,2 1917 1918 1919 18.6 moisture gradually increases in the fall, during the months of October, November, and December ; that it does not change much during the winter months, even in the absence of rainfall; and that it begins to diminish appreciably in the spring, during the months of April or May, due to the requirements of plants and the higher rates of soil evaporation, both of which are brought about by the higher temperatures. In the summer months the moisture absorbed during rains is rapidly consumed by transpiration and soil evaporation, as soon as the rain ceases, until the ground becomes so dry that capillary movement of the moisture practically ceases or until the amount of available moisture is replenished by additional rainfall. The rates of soil evaiwration and transpiration are so high that the upper two feet of soil at Moraine Park seldom, if ever, becomes filled with capillary water during the months of June, July, and August, even though the rainfall may be con- siderably greater than normal. The maximum percentages found during these months, shown by the data in table 3, are given in table 4. It will be noticed that the maximum amount of capillary water that the soil can contain, shown by the preceding discus- sions to be about 21 per cent, was not reached during any one of the months given in table 4 ; although the percentages were rather high in the case of the bare soil in the summer of 1919. However, it is believed that the records obtained during the summer and fall of 1919 are slightly high compared with those taken previously. The rainfall was considerably greater than normal during the month of July, 1915, when it amounted to 5.80 inches ; during the month of August, 1916, when it amounted to 5.98 inches; and during the month of June, 1917, when it amounted to 6.11 inches, the normal amounts for these months at the Dayton Weather Bureau station being 3.28, 3.01, and 3.96 inches respectively. 60 MIAMI CONSERVANCY DISTRICT During the months of January, February, and March the amount of moisture in the soil, even under the most favorable conditions, seldom gets much below the maximum capillary amount, since plant requirements are nil and soil evaporation is very low. The minimum percentages obtained during these months are given in table 5. Records obtained in March, 1915, are not included since the work had hardly become organized at that time. It will be noticed that while these values are all somewhat lower than the maximum capillary value of 21 per cent, they are all considerably higher than the minimum values of from 3 to 10 per cent which generally occur during the summer months. No records were obtained during the month of Jan- uary, 1918, due to the unusually severe winter weather at that time. It is known from other observations, however, that the upper foot of soil became practically saturated during the pe- riod from December 21 to 29 due to the melting of about 9 inches of snow; also that the ground froze before this water could percolate to a greater depth, and remained frozen until the thawing period which began February 6. It will be noticed from the curves in figure 8 that during the months of January and February, 1919, there was little change in the amount of moisture in the soil. Very little drying out seemed to take place between rains although the conditions were probably as favorable for the drying out of the soil as they ever are in the winter. The soil was not frozen ; the mean tem- peratures were comparatively high, being about four degrees above normal; and there was some wind blowing the greater part of the time. The curves in figures 4 to 8, inclusive, show that the changes in the percentages of moisture in the soil between successive ob- servations were considerably greater during the summer months than they were during the winter months, as, of course, would naturally be expected. An inspection of the data in table 3 shows that while the individual observations vary greatly, partly due to differences in soil texture and to errors of observations, the moisture content of the first 6-inch layer of soil seems to vary more than that of the deeper layers. Interesting data on variations in soil moisture at different depths was obtained near Akron, Colorado, by H. L. Shantz, in the summer of 1909.* ♦Natural Vegetation as an Indicator of the Capabilities of Land for Crop Production in the Great Plains Area, by H. L. Shantz, Bulletin 201 of the Bureau of Plant Industry, U. S. Bepartment of Agriculture, 1911, page 31. RAINFALL AND RUNOFF 61 It was there found that during the period from June 10 to Sep- tember 10 the rainfall did not affect the moisture content of the soil below a depth of 18 inches, although on July 7 the rainfall amounted to 2.40 inches. However, a rainy period during the last of May and the first part of June had some effect on the moisture content of the soil down to depths of about 3 feet. At Moraine Park, the moisture content of the soil from 18 to 24 inches below the surface seemed to be affected by the rainfall at times during the summer. This difference in depth of pene- tration at the two locations is probably due to a difference in soil texture. Records showing variations in soil moisture have been pub- lished by numerous investigators. To mention all such data is beyond the scope of this publication. However, the observa- tions by King, published by the Bureau of Soils,* should be re- ferred to since they gave valuable data on the moisture content, at different depths, of eight different soils, under various condi- tions of cultivation and fertilization, in four different states. Evaporation and Transpiration It is interesting to compute the daily rates of evaporation and transpiration for short periods of time from some of the records given in table 3. This has been done for a few selected periods where the data is most reliable. Periods have been chosen in which the rainfall was not excessive, in which there was no appreciable surface runoff, and in which it is believed that there was no percolation of moisture into the underlying beds of sand and gravel. It has been assumed that the decrease in the amount of moisture in the soil in each case, was caused by evaporation and transpiration, and that no moisture was drawn upward by capillary action from the sand and gravel, assumptions which are probably not greatly in error. The data computed in this manner is given in table 6. The moisture in the soil, in per cent, at the beginning and ending of each period, the amount of water in inches depth corresponding to the decrease in moisture percentage, the total evaporation, and the evaporation in inches per day, are given for the soil un- der the sod and for the soil under the bare surface. The term evaporation has been used in the table headings to include trans- piration as well as soil evaporation. The total rainfall, the num- *Investigations in Soil Management, by F. H. King, Bulletin 26 of the Bureau of Soils, U. S. Department of Agriculture, 1905, pages 167-191. 62 MIAMI CONSERVANCY DISTRICT o > |6? is? o C ^ >.S _ ^i CO "d o C i-Ht-05lOOOCDOCvJO?DOJt-00-<*«00«COOOS OCDOi-lU3CD;DCOasC003CO"^r-HOOOOO'^00'^10COi— (>Ji-l CO i-<C0 0a'^»-tt-CM "^OOTt?DOOt- C0W30JiOOC000C000t-OC0t-C00ii00St-i-H«DOS00l0 ffq i-H i-H N W (M i-H 1-H W CO 1-1 i-ti-H .-I iOOtr-CO»-Ht-NCOt-CO jt • 1 tAn Intensive Study of the Water Resources of a Part of Owens Val- ■S^' ^ onf'?^^'.'''' Charles H. Lee, U. S. Geological Survey Water Supply Paper 294, 1912, pages 49 and 118. **Water Resources of the Penobscot River Basin, Maine by H K. Barrows and C. C. Babb, U S. Geological Survey Water Supply Paper 279, 1912, page 120. ±-t- j r RAINFALL AND RUNOFF 67 Table 7. — Soil Absorption at Moraine Park During Summer Storms Rainfall Inches Under Sod Under Bare Soil Storm Period Moisture in Soil Absorp- tion* Inches Moisture in Soil .Absorp- Before Rain % After Rain % Before Rain % After Rain % tion* Inches Aug. 4- 8, 1916. Sept. 5- 6, 1916. Sept. 27-29, 1916. Nov. 23-24, 1916. June 26-30, 1917. Oct. 11-19, 1917. July 22-23, 1918. Aug. 12-23, 1918. Aug. 28-Sept. 2, 1918. 4.56 4.12 1.68 0.84 2.90 2.02 3.11 2.30 1.39 4.7 9.3 8.7 11.3 9.9 9.2 10.3 5.8 10 9 14.9 15.4 16.4 16.9 17.2 16.2 16.4 15.0 18.0 3.92 2.34 2.96 2.15 2.80 2.69 2.34 3.t3 2.73 3.1 3.7 7.4 11.7 10.4 11.1 5.8 6.5 6.0 14.2 9.0 11.3 15.1 13.3 16.0 9.0 15.8 12.7 4.26 2.03 1.50 1.30 1.11 1.88 1.23 3.57 2.57 *In upper two feet of soil. The preceding discussion has shown that throughout the period of about four and a half years covered by the observations the amount of moisture in the upper two feet of soil at Moraine Park varied from a minimum of 4.7 per cent, or 1.80 inches, to a maximum of 21.0 per cent, or 8.06 inches, in the case of the sod covering; and from a minimum of 3.1 per cent, or 1.19 inches, to a maximum of 21.1 per cent, or 8.10 inches, in the case of the bare soil covering; the difference in the actual amounts of water in the 2-foot layer in the two cases correspond- ing to 6.26 and 6.91 inches, respectively, averaging 6.58 inches. This average value would be the maximum possible absorption at Moraine Park as shown by the records. While the maximum percentages used above were practically reached each winter, the minimum values were reached only once. Consequently this average value of 6.58 inches is one which would very seldom, if ever, be attained during a single storm. For the amount of soil moisture to be increased from the minimum value to the maxi- mum during a single storm would require an exceptional combi- nation of conditions such as the occurrence of a very heavy pro- longed rainfall at a time when the ground was dryest. Taking a value of 8 per cent for the minimum amount of soil moisture, a value which is reached practically every summer, the differ- ence between the amount of water in the upper 2 feet during the ordinary dry periods of the summer and the amount present during the winter, would be 5.00 inches. Probably this value is also greater than the maximum amount of water ever ab- sorbed during a single storm. 68 MIAMI CONSERVANCY DISTRICT If the ground is not frozen the proportion of the maximum possible absorption that can be absorbed during a single storm varies with the amount of moisture present when the rain be- gins. The dryer the soil the greater is the space in which the water can be absorbed. If the rain continues long enough the moisture holding capacity of the upper two feet at Moraine Park will become' filled, after which the water will percolate through the underlying sand and gravel as fast as it can move through the surface soil. If the ground is frozen very little moisture will be absorbed unless the duration of the rainfall and the temper- atures are great enough to thaw out the ground. The actual amount of water that is absorbed during a given storm, of course, varies also with the nature of the rainfall. If the rainfall inten- sity is greater than the rate at which the water can soak into the ground, and the surface storage has been filled, the excess water will run off; whereas, if the same total precipitation is distributed through a greater time, it may all be absorbed. Table 7 gives the larger records of absorption during indi- vidual storms, selected from the data in table 3. All records corresponding to depths of two inches or more in the upper two feet are included except in one or two instances where the data seemed questionable. In addition, records corresponding to depths of less than two inches are included for one type of sur- face covering where the absorption under the other type amounted to or exceeded two inches. The percentages of mois- ture present before the rain began and after the rain ceased, as well as the total precipitation during each storm period, are also included. The distribution of the rainfall can be seen by re- ferring to figures 4 to 8, inclusive. The maximum values of ab- sorption given in the table are set in bold face type. The min- imum values have no special significance. It will be noticed that in several instances the absorption was greater than the rainfall. The reason for this is that the place where the samples were taken is located near the foot of a steep hill in the direct path of the surface runoff from the hill- side. As some runoff occurred during each of the storms re- corded in table 7, except the one of September 27 to 29, 1916, the amount of water available to replenish the soil moisture was actually greater than the rainfall. The discrepancy in case of the storm noted is probably due to the difficulties encountered in securing samples representative of average conditions. It is interesting to note that while runoff occurred on the RAINFALL AND RUNOFF 69 plats having the bare surface in all but one of the storms listed in table 7, and on the plats having the sod covering in the greater number of the storms, in no case did the ground become satu- rated. The greatest amount of moisture found in the soil after the rain ceased was only 18 per cent in the case of the sod cov- ering, measured after the storm of August 28 to September 2, 1918 ; and only 16.0 per cent in the case of the bare soil, meas- ured after the storm of October 11 to 19, 1917. Inspection of table 7 shows that the moisture absorbed by the sod is generally greater than that absorbed by the bare soil, as would be expected. There are only two exceptions to this, and in these instances the differences are so small as to come within the limits of possible errors. In one case, that of August 4 to 8, 1916, the values constitute the maxima of the entire record, amounting to 3.92 inches for the sod and to 4.26 inches for the bare soil. This stonn, with a total rainfall of 4.56 inches spread rather uniformly over five days, began when the soil was dryer than at any other time during the period covered by the data. In the other case, that of August 12 to 23, 1918, the values are the next largest, amounting to 3.53 inches for the soil under the sod and to 3.57 inches for the soil under the bare surface. The greater part of the total precipitation of 2.30 inches which fell during this period, occurred in three separate showers on three different days, August 12, 14, and 17. These are the only two instances where the absorption exceeded three inches; and it is interesting to note that both of them occurred during the month of August. All of the storm periods given in table 7 occurred in the summer or fall, during the months of June to November, inclusive. As shown in the preceding discussions, it is only during the summer months that the soil becomes dry enough to absorb such large amounts. The values of 2.34 inches for the soil under the sod and 2.03 inches for the soil under the bare surface, obtained during the period from September 5 to 6, 1916, were caused by a total rainfall of 4.12 inches which fell in the afternoon and evening of the 5th and in the morning of the 6th. The total period of time in which the rain fell was less than 18 hours. Probably 90 per cent of the total precipitation occurred during the 6 hours from 3 to 9 p. m., on the 5th. At the Dayton Weather Bureau station the maximum intensities were 0.43 inches in 5 minutes, 0.79 inches in 10 minutes, and 1.75 inches in 30 minutes. The values of absorption given in table 7 are believed to 70 MIAMI CONSERVANCY DISTRICT be less than the amounts that actually occurred during the given storms, the reason being that while samples were always taken within a few hours or a day after the rain ceased quite frequently samples had not been taken for several days before the rain began. In calculating the values given in the table no allowances were made for the evaporation which must have oc- curred between the time the samples were taken and the time the rain began. Table 8.— Soil Absorption at Moraine Park During Winter Storms storm Period Rainfall Inches Under Sod Under Bare Soil Moisture in Soil Absorp- tion* Inches Moisture in Soil Absorp- tion* Inches Before Rain % After Rain % Before Rain % After Rain % Dec. 12-17, 1915 Apr. 20-22, 1916 Mar. 13-14, 1917 Dec. 9-13,1918 Mar. 5-10,1919 Mar. 15-18, 1919 Mar. 26-27, 1919 Apr. 9-11,1919 1.75 1.08 1.44 1,59 1.09 2.97 0,85 1.53 15.6 17.7 19.8 17,4 19,9 20.4 18.6 18.6 17.3 19.8 19.0 20.0 21.0 20.6 20.6 19.4 0.65 0.80 -0.30 1.00 0.42 0.07 0.77 0.30 16.7 13.7 17.8 16.9 18.0 17.2 17.7 18.0 16.3 15.3 20.3 19.3 19.6 18,4 19.2 18.3 -0.15 0.61 0.96 0.92 0.61 0.46 0.57 0.11 *In upper two feet of soil. In order to show how much moisture is absorbed by the soil during winter rains a few of the larger records of absorption during individual winter storms, selected from the data in table 3, are assembled in table 8. The percentages of moisture pres- ent before the rain began and after the rain ceased and the total precipitation during each storm period are included, as in table 7. Only storms occurring during the months of Decem- ber to April, inclusive, at times when the ground was not frozen, are considered. Maximum values are indicated as before. The negative values appearing in table 8 are probably due to errors in observation as it is hardly likely that the soil was drier after the rain than it was before. It will be noticed that in only one case, that of the soil under the sod in December, 1918, was the absorption as much as an inch. There does not seem to be much difference in the winter between the absorp- tion under the sod and under the bare soil surface. The ground was frozen during the greater number of storms which occurred in the months of January and February, and consequently such records were not included in table 8. RAINFALL AND RUNOFF 71 It will be observed that in the greater number of storms listed, in table 8 the soil was nearly saturated when the rain be- gan and was practically saturated when the rain ceased. It is interesting to point out that although the ground was saturated by the storms of March 5 to 10 and 26 to 27, 1919, no trace of runoff could be observed on any of the plats. • In these in- stances the rates of precipitation must have been less than the rate at which the water could percolate through the two-foot layer of surface soil. SURFACE RUNOFF While the records in table 3 cannot be used to solve all prob- lems connected with surface runoff, they do furnish some inter- esting information. They enable us to study the conditions un- der which surface runoff begins, the rates at which moisture can be absorbed by the soil, the relation of the total surface run- off to the total rainfall, during storm periods as well as during the year, and the amount of water that percolates through the surface soil to maintain the ground water flow of the streams. They also enable us to study the effect of variations in the slope of the ground, in the nature of the surface covering, in the amount of moisture in the soil when the rain begins, and in the character, intensity, and duration of the precipitation. Surface Slope and Surface Covering In order to study the variations in surface runoff caused by variations in surface slope and in surface covering, the larger runoff records of table 3 are assembled in table 9. Runoff meas- urements for all plats are included for all dates on which the observations show a total runoff of an inch or more on at least one of the plats. The differences in runoff due to variations in surface slope and to variations in surface covering have been calculated and are given in the last four columns of the table. The observed total rainfall is also included. Maximum values of the various quantities are set in bold face type as in preceding tables. It should be pointed out that the amounts of rainfall and runoff, here given, did not occur on the date of observation noted in the first column but occurred during the time between the date noted and the preceding date on which observations were made. An idea of the probable distribution of the runoff 72 MIAMI CONSERVANCY DISTRICT O o ;3 e « o a Kc Ot-COTHrHCOCOi— IOOT-COt>0 iX)rHU3CT500tr-CO©'!*{Mrtt- I— IrHrHrHr- 1>— (1— IrHCOr— iCMrHi— < CC-^rJct*E>O'^5D'^rH't:-E>t>l>00CX)0000a> rj^-j'-^'^'-'THi-Hr-liHi-Hi-HTHT-l CX3COt>'X)OOT-(010TH(riCO ^O lO i-H i-H CO c^i i-H c^ oq c*ci (M o o o ■— • +J ^ 3 ^ 3 ■a -a o o 0) o W H H <5 o o ft o „ m H i cS o ^ o cu >. s < cu a s H o hj hj 03 1— * in > *'"' w .'^ « O o OJ ij ,c >< -4-^ <: m ^ H 0) .« K ^ w g H 3 t3 H iJ -M fu :« 1— 1 S H 0) ^ J3 4-> H cS d o 1— 1 fc< (U !h a! u -H a 3. OJ m OJ ,C H RAINFALL AND RUNOFF 99 runs, are given immediately following the last run of the series. Columns are also included giving the experiment number, the run number, the date, the time between successive runs on the same day, or on consecutive days, in hours and minutes, the ratio of total runoff to total rainfall in per cent, the average rate of sprinkling in inches per hour before runoff began, the time in minutes before runoff began, the total precipitation be- fore runoff began, and the condition of the plats when the exper- iments were started. For runs 1 to 8, inclusive, and runs 23 and 24, given in table 16, where the water was applied with one sprinkling can, the time of sprinkling is the net time, not including the intervals re- quired to fill the can. The rate of runoff was computed by using the total net time in which runoff occurred, and the rate of reten- tion was obtained by assuming the time in which the retention took place to be the same as the total sprinkling time. Con- sequently the differences between the rates of precipitation and runoff are not exactly equal to the rates of retention. It might have been better to have computed the rates of runoff and re- tention in a slightly different manner but since these experiments were more or less of a preliminary nature the data has not been recalculated. The sprinkling was continuous during all of the remaining runs of table 16 and during all runs of table 17. For these runs slightly different methods of compilation were used. In calcu- lating the average rate of runoff for a given run the short pe- riods of time at the beginning and ending of the run, in which the rate of runoff was changing greatly, were not. considered. The rate of retention was obtained by simply subtracting the rate of runoff from the rate of rainfall. In all instances, the total time given for a certain experiment, or series of runs, is the actual time from the beginning of the first run to the end of the last run, including nights as well as other intervening periods, rather than the total of the sprinkling times for the runs included in the experiment. For runs 1 to 8, inclusive, in table 16, a second set of totals is given, in which are included the actual rainfall and runoff quantities which occurred during the evening of June 2. In table 18 the total quantities for the different experiments are brought together so that they may be studied collectively. The total time from the beginning of the first run to the end of the last run in hours, the total sprinkling time in hours, the total 100 MIAMI CONSERVANCY DISTRICT B o 61 9 T3 s '3 GQ ■55 g I EcSS rf< CO ►^ W 11 ■«.2.S « ffi o O v3 t,** « & S i&g •Ss.s-§ (§ iS Q 1 IlilBi'S.SS «!■' ^B o S ■ lO'^i-Ci-IOOt-Oc- NIMOClOOOONgN T|i|M >-HOOO TO ooiH-<#-^eooot- tHi-HOOOO t-H Oli-HOOOOOO O^-Hi— lOOt-NrHi-H "*(Mi-lOOOOCO «_>» III lis 0) caT3 " h 0)00 — ■ "5 o COOONOOO U3i-IU3iOt> c CO CO CO i~^ o (MOJCOtHOOO O kO 00 CO O CO O) t- CO CO tH i-( o o 0000000 CD CO -^ O 05 T*-<0 N CO 10 t- 00 U5 CJ t-05^ocqoo«oo- •HOOOOO OCO'^ 00004 OOOCCOIAACO t-05C0T-HC0iHOC^00CHOO (M (N N CO CO CO CO t- 1- 1- 1- 1- 1- O^ O ^-H M CO -^ 10 1— 4 1-H T-H 1-H »— ( tH OS e ijuauiuadxa RAINFALL AND RUNOFF 101 & ■sae S CO -^ CO N G^ 0000000 CO 00 CO »-lOO o tr-t-O"^O00OCa «Ot-Q0t-CO"^i-Ht- eg 0'^ 00000000 oco-^ooa-^Noo -*'cocotj<,-hoo?o lOioooocgcgcoas NOOi-HCOt-t- - 05 tHt-Ht-IC^t-HOON OSOOT-H»-t I I I I 10 rtOt-OJO 00 00 00 0& O^ O) A N Cj] eji Cji Cj] eq N t- 1- 1- 1- 1- 1- 1- eot-oooio^N I f :^uaaiu3dx3 S5a.S I I 00 CO -^US I (MCMNCO Z jnain -uadxa a ffi O ^ C ^ •a 0) 60 O.S ■Sc5 .S ""tJ u g c I ° 0) « 8 §|.S l-H&tt-'l-l «<»» I I « 102 • MIAMI CONSERVANCY DISTRICT B » a CO i s i-H Eh Soil dry and hard; moisture content probably about 10 per cent, when Run 1 was begun. Soil unusually dry and hard; moisture con- tent about 7 per cent; a few cracks in surface within plat and along bound- aries, when Run 8 was begun. > Soil dry and loose; spaded to a depth of 6 inches and thor- oughly raked; mois- ture content about 10 per cent, when Run 11 was b^;uD. 1 m lfl.l lO 0000 CO 1-HCO r^ ^ ca 00000 coejeo odo CO OS 00 10 U3 •*o.-ioo rtOOOO 141 3 0U500 W5 . . . -OO 010 U3 t-NO 0U3000 c4<-icocoia » t- 10 -<1" Tf • •" lO(MCO«)00 • ■ osot- COOO] oe»ioo3t- 3S CO 00010 COCOCT— 10 • ■ CO CO CO COCOi-<^0 1 ■5 H^ a Ol 10 OS 05 !0 i-l N 01 t- CD »H 10 PJ OS 00 CO CO CO OS OS OS *-t »H OS CD ■ ■^iHOUSrtO ■ ^NN-HOOO NMN tHN^HOOO ■ i-t mcocooNi-ico t-U3 t- 010^ ■* T|l CO CO OS >o • oq •* 00 10 CO CO ■ tHOOOOOO .-100 NtlOOOO • 1 s.s fr4 N 00 N 05 00 e0 00 0> 01 U> N CO 00 00 CO lO Tf< t- No»u5eq cot-oot- NOSNOOOOSIO COkOU3"^CO -^ 1 1 1^1 QO CO -^ t- rt CO rt 10 U3»OCOOt-»-HOCO >OCO-rto->*co>o NCOt CO C^IO-^COOO^X C-Oe»COIN(MOO -HOOOOOOTt eaoocQ •>»(NrtOOO0S 14 « 00000000 ooooevi-5|i(Noo §§§§ OOOOOCOCO OOOIN-*A oOTfojoocot-cNm eq OS 01 1- M t- 1- OS •-HOOi-HtHOOCN -^cooo t-osoot- tHOOWS 00 00 Thus 05 t- ■•- 00 CO 1-H 1-H CD U3 X I-l T-l CVI *-< 1^ Tf< ■(N coos 00 :7l^?7oo •eviot-.-i ■i-IO ■OS t-TT i-( :^"P7 1 : •IM00.-IO 1 Q eo OS OS 030000 tH iH T— t i-H 00 00 00 00 00 00 00 OS OS OS 000 »-H 1-H »-l i-H »H rH * 1 1 1 1 1 1 ■ 000000000000 • ^ t. 4^ (N •H N CO Tji «5 to t- 1 oooso 1 cs — 1 N CO ■<* 10 CD 1 *C 4J - g ^uaoiiiadxg 6 ^nainijadxg 9 ^uarauadxa RAINFALL AND RUNOFF 103 ■^tHa ooooooo (M O 00 O t- -<* i-( CO Tj4 CO 03 05 to -^ COCOCOt-IOOO N CO CO Ol C- t- O i-H O50000t-00(M.-HO (M00t-i lO t- o t- 1- •>* ,-1 -ON •CO • ■ eg CO <-H(N ■* -^ "O ooooooo ss IW-1 rHTHCCN 0 CO ■* m S JSHH rH OO 1-1 1- CO t- 0» 1 ■* oi iM ea CO N eq 1 rH ^ " B l-S O<0OO ocoooooo ^ Is t- so 00 00 OOOOOOrfOOO e Tj< t- lO «5 (M (M to T-4 f-( »-H «— ( 1— 4 T— ) T— i T— I 1— ( llll CO Eh OO^OOIA 02 -tf 1-) CD OI U3 OJ tH 'l- WOiON H 00 00 U5 xn Oi 00 -^ CD I> CO Cq m <1 ^cqoo) 35000>0-*lO . S J" -
  • ■ ri a c a '_;'o S O CO a m 0^ 03 Q d f^ <- . . . o 2 ^t»PQ bit h f3 _^— c rH N CO 1* — 1 1^ CO 0) a) oj^ ,5.2 ,£2 !« =s si ca 1 i-Pi^hJoi OiPLiPHCLif^C^tl, 3 ^ H "S^i -f-. r-l 0) b e B V ■—■ ■M o ^^ '■S Jo >>>>>>es 5 ^ »-• »-• ,3 +5-4-5 c3b.hU 9> 0) 1 •2 ? <■> X ^^ ^^ - .^ * 73 73 n ■g-^-SilT* S S S MTSTS-a o o JS 0) C 13 c h h ^ «5 OS ca ca UtXI m •— 1 10 tC DD ■>*co 3,43^ JS ••i+i > CD C«^ c3 a ^ftftD. CO, CO GO OQ DQ c c M C3 d C3 OJ Q) OJ 0) S 0) 0) & S S R'43 Cd cd 1 R ra d c3 wcQwra SS Ifcl lOOOt-O t-TO ■-HNCOW ^— ( tH c 3, ^£•3 " oooo oo B i4 lOOOO 0O(M H-| (MtDON N(M «-■ & erage nkling Inches Hour us-ooa> 05m ;d 00 CO 00 «>io ^« a_e g, ec(NNO CO CO »-HCCCCCO coco NNN (MIM 1 o . . . « c -^ -^ -^ 4-5 +5 a 3 " " u o a »?ooo OO B J ">"> .SS2 £ D3 m F-t F-4 S_o_o_o o o L^ ti d ti "S.Tl CD at E-iH Run umber T-HNCO-* toin ^ i 106 MIAMI CONSmVANCY DISTRICT rainfall, retention, and runoff, in inches, the ratio of total run- off to total rainfall in per cent, and the rate of retention during the last run in inches per hour, are given for each experiment, experiments being arranged chronologically and numbered con- secutively. In table 19 the average rate of sprinkling in inches per hour, the time in minutes, and the total precipitation in inches, be- fore runoff began, are given for the special runs made in order to study the conditions causing runoff to begin. Run numbers, dates, locations, and surface conditions are also noted. Figures 13 to 21, inclusive, show graphically the data ob- tained in the various experiments. Figures 13 to 15 show the results obtained on the bare soil plats at Moraine Park, and fig- ures 16 to 21 show those obtained on the Taylorsville plats. The data secured on the sod plat at Moraine Park is so unusual, as will be explained later, that it has not been platted. The rain- fall, retention, and runoff are shown by means of mass curves, separate curves being drawn for each day's observations. The rates of rainfall, retention, and runoff are, of course, shown by the slopes of the lines. The various runs are numbered, as in tables 16 and 17. Intervals between runs made on the same day are indicated by the horizontal portions of the curves. The comparatively steep parts of the retention curves at the begin- nings of the runs, and also the small peaks or humps near the ends of the runs, are due to surface storage. RAINFALL, RETENTION, AND RUNOFF Reference to table 18 shows that with one or two exceptions the total quantities of water applied in the various experiments were larger than any actual rainfalls on record in, or near, the Miami Valley. With the exception of experiments 1, 2, and 11, the rainfall varied from 12.00 inches in about 5 hours to 17.83 inches in about 30 hours; the total runoff, from 7.95 to 13.70 inches; and the total retention, from 2.10 to 9.88 inches. In experiment 1 the total rainfall was 8.70 inches in about 2 days. Experiment 2 was run on the level sod plat at Moraine Park where the soil conditions were unusual. In experiment 11 the rate of application was purposely kept low in order to study the retention for less intense storms. As noted in tables 16 and 17 the soil in all instances, was comparatively dry when the experiments were started. The ground at Taylorsville was a little drier in October, when experi- RAINFALL AND RUNOFF 107 SO 100 ISO Time Jn Minutes Z50 PIG. 13.— EXPERIMENT 1, MADE ON THE LEVEL BARE SOIL PLAT AT MORAINE PARK, JUNE 1 TO 3, 1920. The curves show the total rainfall, runoff, and retention up to any instant. 108 MIAMI CONSERVANCY DISTRICT 50 100 J50 ZOO ZSO Time in M/nofes 300 350 400 FIG. 14.— EXPERIMENT 3, MADE ON THE LEVEL BARE SOIL PLAT AT MORAINE PARK, JULY 22 TO 23, 1920. The curves show the total rainfall, runoff, and retention up to any instant. RAINFALL AND RUNOFF 109 zao 250 Time in Minutes FIG. 15.— EXPERIMENT 4, MADE ON THE SLOPING BARE SOIL PLAT AT MORAINE PARK, JULY 28 TO 29, 1920. The curves show the total rainfall, runoff, and retention up to any instant. no MIAMI CONSERVANCY DISTRICT 50 /SO ZOO Z50 Time in Minutes 300 350 400 FIG. 16.— EXPERIMENT 5, MADE ON PLAT 1 AT THE TAYLORS- VILLE DAM, AUGUST 9 TO 10, 1920. The curves show the total rainfall, runoff, and retention up to any instant. RAINFALL AND RUNOFF 111 Jff 100 J50 ZOO ZBO T/'me in Minutes 300 3B0 400 PIG. 17.— EXPERIMENT 6, MADE ON PLAT 2 AT THE TAYLORS- VILLE DAM, AUGUST 11 TO 12, 1920. The curves show the total rainfall, runoff, and retention up to any instant. 112 MIAMI CONSERVANCY DISTRICT 150 ZOO ?50 Time in Minutes 300 550 400 FIG. 18.— EXPERIMENT 7, MADE ON PLAT 3 AT THE TAYLORS- VILLE DAM, AUGUST 16 TO 17, 1920. The curves show the total rainfall, runoff, and retention up to any instant. RAINFALL AND RUNOFF 113 SO too /3C 200 ?B0 Ti'me in M/77ufei> FIG. 19. — EXPERIMENT 8, MADE ON PLAT 4 AT THE TAYLORS- VILLE DAM, AUGUST 18 TO 19, 1920. The curves show the total rainfall, runoff, and retention up to any instant. 114 MIAMI CONSERVANCY DISTRICT 30 100 /SO ZOO Time //? Mi'nvtes 250 300 350 FIG. 20.— EXPERIMENT 9, MADE ON PLAT 1 AT THE TAYLORS- VILLE DAM, OCTOBER 19, 1920. The curves show the total rainfall, runoff, and retention up to any instant. RAINFALL AND RUNOFF lis m 200 Z50 ZOO 350 Time in Minutes PIG. 21.— EXPERIMENTS 10 AND 11, MADE ON PLATS 4 AND 3 AT THE TAYLORSVILLE DAM, OCTOBER 21 AND 22, 1920. The curves show the total rainfall, runoff, and retention up to any instant. 116 MIAMI CONSERVANCY DISTRICT ments 9, 10, and 11 were made, than it was when the earlier ex- periments were begun. Consequently the data in table 18 is comparable with conditions occurring during intense summer thunderstorms, or cloudbursts, rather than to those existing dur- ing large winter storms. It is interesting to note that although the total precipitation was unusually large in practically all of the experiments, the rates of retention when the experiments were finished were still comparatively large. This would indicate that the soil was capable of taking up considerable additional moisture. The lowest rate of retention given in table 18, 0.09 inches per hour on the level bare soil plat at Moraine Park, would correspond to a total absorption of 0.90 inches in 10 hours or 2.16 inches in 24 hours if continued for such periods of time. It should be noted that these rates of retention are lower, in most instances, than they would have been if the rates of precipitation had been higher, since it will be shown later that the rate of retention generally increases somewhat, if only slightly, as the rate of precipitation increases. The values of retention include soil absorption, percolation, and evaporation. Since the experiments were run on compara- tively warm summer days when some wind was blowing, it is possible that the evaporation may have amounted to as much as 0.05 of an inch an hour in certain instances, although it is not believed that it ever could have exceeded this appreciably. At Moraine Park, owing to the shallow surface soil, retention must have been mostly percolation after the first run of each exper- iment. Moraine Park Sod Experiment 2, made on the level sod plat at Moraine Park, shows an unusually large retention. Of the total precipitation, 11.66 inches in 1.09 hours, only 1.29 inches ran off, leaving a retention of 10.37 inches. By referring to table 16 it will be seen that a rate of precipitation of 4.17 inches per hour was ap- plied for 30 minutes with no runoff whatever ; that a rate of 14.4 inches per hour was then applied for 12.2 minutes before runoff began ; that this rate was continued for about 13 minutes, obtaining an average rate of runoff of only 0.4 inches per hour; and that a rate of 20.4 inches per hour was then applied for about 10 minutes, causing a rate of runoff of only 7.4 inches per hour, or less than half of the rate of application. In the last run the RAINFALL AND RUNOFF 117 water was applied with a garden hose, the rate being determined by tank measurements immediately before and after the run. In the first two runs the sprinkling was somewhat intermittent, since only one sprinkling can was available. The times given above are the net sprinkling times, not including the intervals ■during which the can was being filled. These extremely unusual results are due to the unusual soil conditions existing at this plat. As explained in chapter III the soil is loose and is filled with roots down to the depth where the material is mostly sand and gravel. Consequently the water can percolate downward at very great rates. The data obtained by sprinkling agrees with the natural rain- fall and runoff records given in the preceding chapter. It was there shown that runoff on this plat occurred very infrequently ; and that when it did occur it was generally due to melting snow or to rain at times when the ground surface was frozen. Moraine Park Bare Soil Experiments 1, 3, and 4 were made on the bare soil plats at Moraine Park, numbers 1 and 3 being made on the level plat and number 4 on the sloping plat. The soil in the sloping plat when experiment 4 was begun was in about the same condition as the soil in the level plat when experiment 1 was started. It was comparatively dry and loose in both cases. When experiment 3 was started, however, the soil in plat 1, although dry, was hard and packed, and consequently in a much more impervious condi- tion. This was caused by the trampling of a herd of cattle which was turned into the field in which the plats are located, imme- diately after experiment 1 was made, while the soil was still in a saturated condition. The sprinkling was intermittent during experiment 1 due to only one can being available, but was con- tinuous during experiments 3 and 4. By referring to table 18 it will be seen that the total reten- tion during experiment 1 was considerably greater than the to- tal during experiment 3, although the quantity of water applied in the latter instance was about 70 per cent greater than in the former. This, of course, would be expected due to the differ- ence in soil conditions and in sprinkling methods. For experi- ment 4 the retention was 4.73 inches or only slightly more than for experiment 1, although the total precipitation was practi- cally twice as great. The relatively smaller retention during ex- 118 MIAMI CONSERVANCY DISTRICT periment 4 was due to the slope of the ground as well as to the difference in sprinkling methods. The total runoff was 4.44 inches, or about 51 per cent of the total rainfall, during experiment 1; 12.30 inches, or about 83 per cent of the total rainfall, during experiment 3 ; and 12.07 inches, or about 72 per cent of the rainfall, during experiment 4. Taylorsville Plats Experiments 5 to 11, inclusive, were run on the plats at Tay- lorsville; numbers 5 and 9 on plat 1, number 6 on plat 2, num- bers 7 and 11 on plat 3, and numbers 8 and 10 on plat 4. Plat 2 was spaded and raked before experiment 6 was begun, as pre- viously noted. The other plats were in their natural condition. Sprinkling was continuous during all experiments. Experiments 5 to 8, inclusive, were practically the same as regards time of sprinkling and total quantity of water applied, except that the total precipitation during number 6 was from 2 to 3 inches greater than during the others. Experiments 9 and 10 were similar, each being run in about the same time and at about the same rainfall intensity. Experiment 11 was run in about the same time as numbers 9 and 10 but the water was ap- plied at a less intense rate. Reference to experiments 5 to 8, inclusive, in table 18 shows that the retention on plat 2, in experiment 6, where the soil had been loosened, was more than twice as great as on plat 1, in ex- periment 5, where the soil was similar in composition and tex- ture but had not been spaded ; and about four times as great as on plats 3 and 4, experiments 7 and 8, where the soil was mostly clay and had not been loosened. The total retention on plat 2 was 9.88 inches or about 55 per cent of the rainfall ; as against 4.15 inches or about 28 per cent of the rainfall, on plat 1, and about 2.45 inches or about 16 per cent of the rainfall, on plats 3 and 4. The total runoff amounted to 7.95 inches or about 45 per cent of the rainfall in experiment 6 ; to 10.65 inches or about 72 per cent of the rainfall in experiment 5 ; and to about 13.15 inches or about 84 per cent of the rainfall in experiments 7 and 8. In experiment 9, made on plat 1 in October when the soil was slightly drier than in August, the retention amounted to 3i35 inches, or about 28 per cent of the precipitation. During runs 1 to 3 of experiment 5, which are comparable with experiment 9, the retention amounted to 2.30 inches or about 23 per cent of the rainfall. Experiments 10 and 11 also show greater values RAINFALL AND RUNOFF U9 of retention than were obtained in the parts of experiments 8 and 7 which are comparable. It is interesting to note that the retention during experiment 11 was slightly greater than during experiment 10 although the total precipitation in the latter was twice as great. This is due to the slightly greater sprinkling time during experiment 11. The rate of retention for plats 3 and 4 varies only slightly with the rate of rainfall, as will be shown later. Conditions during experiments 10 and 11 are shown in figure 21. RAINFALL AND. RUNOFF RATES ON SATURATED SOILS The data in tables 16 and 17 is essentially data on rainfall, re- tention, and runoff on saturated soils. Although the soil was comparatively dry when the experiments were started in all cases, it soon became saturated to such a depth that runoff be- gan. By the end of the first run the soil was generally saturated to such a depth that the rate of runoff caused by a given rate of precipitation was practically constant. This was true for all experiments except number 2, made on the sod plat at Moraine Park, where the soil conditions were unusual, as previously men- tioned; and numbers 10 and 11, made on the hill plats at Tay- lorsville, in October, when the soil was somewhat drier than in the earlier experiments. At Taylorsville, the actual depths of saturation probably varied somewhat during the later runs of the other experiments, increasing as the work progressed. How- ever, these variations may, for the present, be neglected. At Moraine Park the sand and gravel deposits, underlying the 2- foot layer of surface soil, afforded ready drainage and thus prevented the extension of saturated conditions below this depth. Moraine Park Bare Soil The rates of rainfall, retention, and runoff for the bare soil plats at Moraine Park, given in table 16, are shown graphically in figure 22, rates of rainfall being platted as ordinates against rates of runoff as abscissas. The rates of retention are repre- sented by the horizontal or vertical distances from the platted points to the 45° line, curve D, drawn through the origin. This 45° line represents the limiting conditions of runoff. A point would fall on this line only when the runoff rate was equal to the rainfall rate. Points representing runs where the soil was not 120 MIAMI CONSERVANCY DISTRICT saturated have been identified by placing near them the run numbers. Other points have not been numbered. B D / 2 3 /?«/€ of Runoff inJnches per Hour FIG. 22.— RATES OF RAINFALL AND RUNOFF ON BARE SOIL PLATS AT MORAINE PARK. Runs in which the soil was not saturated are numbered. Experiments 1 and 3 were made on the level plat and experiment 4 on the sloping plat. Curve D is a 45 degree line. For each experiment a line has been dl-awn so as to balance all points except those where the soil was not saturated. Conse- quently these lines show the variations in the rate of runoff caused by variations in the rate of rainfall, during the summer and fall, after the soil has become saturated by a fairly heavy and intense precipitation. Variations in the rate of retention due to variations in the rate of rainfall are shown by the differences between the lines RAINFALL AND RUNOFF 121 representing the observations and the 45° line. Differences in the relation between rainfall and runoff for saturated soils due to variations in soil texture and in surface conditions are shown by the differences in the slope and in the location of curves A, B, and C. Curve A averages the various runs of experiment 1, made on the level plat, curve B averages those of experiment 3, also made on the level plat, and curve C averages those of experiment 4, made on the sloping plat. It will be noticed that the rates of runoff corresponding to given rates of rainfall were consider- ably lower in experiment 1 than they were in experiment 3, due to the different method of sprinkling as well as to the different condition of the soil ; also that they were considerably lower in experiment 1 than they were in experiment 4, due to the differ- ent method of sprinkling and to the different slopes of the sur- faces. In all three curves the rate of retention increases as the rate of rainfall increases. In curves B and C this is due entirely to the head caused by the greater depth of water on the ground, this increased head producing an appreciable effect because of the comparatively shallow depth of the surface soil. While the effect of varying head must have been fully as great in curve A as in curves B and C, the method of sprinkling and of calculat- ing the rates of runoff also had some effect. If the sprinkling had been continuous in experiment 1, or if it had been possible to eliminate surface storage effects in calculating the rates of run- off, curve A would probably have fallen in some position inter- mediate between its present position and that of curve B, as at A'. Probably the most interesting* thing brought out by figure 22 is that the relation between rates of rainfall and runoff may be represented by straight lines; that is, that the relation repre- sented by any one of these lines may be expressed by the straight line equation where y is the rate of rainfall, x is the rate of runoff, b is the intercept on the y axis, and s is the slope of the line. It will be noticed that the value of b is about 0.20 inches per hour for curve A, 0.05 for curve B, and 0.24 for curve C. These are the rates of precipitation that can be maintained indefinitely on the Moraine Park bare soil plats, during the summer and fall when the soil is saturated, without any runoff whatever occur- 122 MIAMI CONSERVANCY DISTRICT ing ; that is, for these rates or lower rates, the water can perco- late downward as fast as the rain falls. Since the rate of percolation decreases with a decreasing tem- perature the value of b may be slightly smaller during: the win- ter and spring. Allen Hazen,* speaking of friction losses in sand and gravel, says "I have found that the friction also varies with the temperature, being twice as great at the freezing point as at summer heat, both for coarse and fine sands." It is possi- ble however, that in a shallow surface soil as at Moraine Park the loosening due to freezing and thawing may counteract to a certain extent the effect of the decrease in temperature. Since curves A and C represent soil conditions comparable with those existing when the natural rainfall and runoff data was collected, the values of 6 determined from the sprinkling ex- periments furnish a satisfactory check on the conclusion reached in the preceding chapter; namely, that water can percolate through the surface soil on the bare soil plats, when the ground is saturated, at a rate as great as 0.25 of an inch an hour. Taylorsville Plats Figure 23 shows the Taylorsville data, contained in table 17, platted in the same manner as in figure 22. Curve D is the 45° line as before. Curve E averages the data taken on plat 1, curve F averages that taken on plat 2, and curve G averages that taken on plats 3 and 4. The conditions on plats 3 and 4 are practically identical, so that it is not necessary to draw a line for each plat. In platting the points, however, different symbols were used for the two plats, so that the agreement of the data may be seen. Sprinkling was continuous during all runs. The increased slope of curve F over that of curve E shows the increased retention obtained by spading, or loosening, the soil. It will be noticed that the rate of retention increases con- siderably as the rate of precipitation increases, in the case of plat 2 ; but that it is practically constant in the case of plat 1. The slightly higher rates of runoff shown by curve G over those shown by curve E are due to the slightly greater impermeability of the clay soil at the hill plats. The rate of retention at the hill plats is similar to that at plat 1, in that it increases only *Some Physical Properties of Sands and Gravels, by Allen Hazen, AZrXor^ pt'e^sT^^^^^^^^ ^**^*°"' ^--^husetts, Twenty-fourth RAINFALL AND RUNOFF 123 slightly with the increasing rate of precipitation. It is compara- tively small throughout the range covered by the data. Here, also, the relations between rates of rainfall and runoff for saturated soil may be shown by straight lines. The value of b is seen to be 0.30 inches per hour for curves E and F and about 0.20 inches per hour for curve G, meaning that precipitation can ^ G D / 2 3 Rate of Runoff in /nches per/iour FIG. 23. — RATES OF RAINFALL AND RUNOFF AT THE TAYLORS- VILLE DAM. Runs in which the soil was not saturated are numbered. Note the in- creased retention obtained by spading the soil. Curve D is a 45 degree line. occur at these rates, during the summer and fall when the sur- face soil is saturated, without any runoff taking place. It should be noted that these values are practically the same as those ob- tained at Moraine Park notwithstanding the difference in sub- soil. For winter and spring conditions, however, the values of 124 MIAMI CONSERVANCY DISTRICT b at Taylorsville may be lower than at Moraine Park, due to the greater depth of saturation and the lack of adequate soil drainage. Average Relations In figure 24 the various curves of figures 22 and 23 have been brought together. Points have not been Shown since they would C rEi/G. 4 jn / x W/ .Si i ^ {/, y ^ f ■ /a Y 1 1 ■S! / / /^ ^ } ■f^ ///, f t n / 2 3 4 Rate of Runoff in //jc/jes per Hour FIG. 24. — AVERAGE RELATIONS BETWEEN RATES OF RAINFALL AND RUNOFF, SOIL SATURATED. The two dotted lines represent the average relations shown by the two groups of full lines. For identification of full lines see fig^ires 22 and 23. only confuse the diagrams. It will be noticed that the lines fall into two separate groups; first, curves A and F, and, second, curves B, C, E, and G. Curve H has been added to represent the RAINFALL AND RUNOFF 125 average relation between rainfall and runoff rates shown by the lines of the second group, and curve I has similarly been dravra to represent the average relation shown by those of the first group. The relation shown by curve H may be expressed by the for- mula y=zl.07x-\-0.20 (1) where y is the rate of rainfall and x the rate of runoff, both ex- pressed in inches per hour. Since y is the cause and x the effect it seems more advantageous to change equation 1 to the form x = 0.93 (y—.0.20) (2) It will be noticed that the coefficient of the quantity (y -r- 0.20) is not greatly different from unity ; or in other words that curve H is nearly parallel to curve D. This means that the rate of retention increases only slowly with an increasing rate of precipitation. For practical purposes this variation in the rate of retention may be neglected. Taking the value of the retention as 0.30 inches per hour, a value corresponding to a rate of pre- cipitation of 1.50 inches per hour by curve H, the relation be- tween rates of rainfall and runoff may be expressed by the equa- tion x = y — 0.30 (3) The line representing equation 3 would be parallel to curve D and a constant distance above, equivalent to 0.30 inches per hour. Consequently, during the summer and fall when the ground is saturated, rates of runoff from soils similar to those represent- ed by curves B, C, E, and G may be estimated by simply deduct- ing 0.30 inches per hour from the rates of rainfall. Curve I, representing the average relation for curves A and F, may be expressed by the equation y = 1.60x + 0.25 (4) or, solving for x, x = 0.62 (y — 0.25) . (5) The retention would then be expressed by the equation y — X = 0.60x + 0.25 (6) This time the rate of retention increases appreciably as the rate of rainfall increases; and, consequently, it will not be ad- visable to replace the slope coefficient 1.60 by unity. Equations 3, 5, and 6 are applicable during the summer and fall; on areas similar to those where the experiments were made, after the surface soil has become saturated by a precipitation of three or four inches falling in one of two days. The data 126 MIAMI CONSERVANCY DISTRICT given in the preceding chapter indicates that equation 3 is appli- cable on the sloping bare soil at Moraine Park during the winter and spring after a half an inch or an inch has fallen, although it probably will give rates of runoff slightly too low at such times. The rates of percolation during the winter and spring on plats 1, 3, and 4 at Taylorsville, after an inch has fallen, may differ somewhat from the 0.30 of an inch an hour shown by equation 3, no data being available for these plats for such seasons of the year. The rate of percolation on plat 2 probably decreases as the soil becomes packed by the winter and spring rains. The average relation between rates of rainfall and runoff for the Miami Valley during the summer and fall when the sur- face soil is saturated, probably lies somewhere between those shown by equations 3 and 5. It is, of course, very difficult to estimate the average soil conditions over a large drainage area. However, considering that practically throughout the valley the amount of vegetation exceeds that on plats 1, 3, and 4 at Taylors- ville, and also that the greater part of the land is cultivated, it seems fairly certain that the average rate of retention would exceed that used in determining equation 3. It does not seem possible though, that the retention could amount to as much as that obtained on plat 2 at Taylorsville, where the soil had been thoroughly spaded and raked just before the experiment was made. RAINFALL AND RUNOFF RATES, SOIL NOT SATURATED When the ground is not saturated the relation between rates of rainfall and runoff varies greatly with the amount of mois- ture in the soil, as well as with the soil texture and the surface conditions. On a given pl^t the drier the soil the greater will be the rate of retention and the smaller will be the rate of runoff corresponding to a given rate of rainfall. Experiments 10 and 11, made on plats 4 and 3 at Taylorsville, furnish data on the relation between rates of rainfall and runoff for different amounts of soil moisture. Since the soil and sur- face conditions are the same at these two plats, and since the amount of moisture present when experiments 10 and 11 were started was the same, it is possible to select portions of these experiments in which the amount of soil moisture present was the same for both plats. The rates of rainfall and runoff can then be calculated for these portions for both experiments and RAINFALL AND RUNOFF 127 the difference in the runoflf rate for a given portion will be due entirely to the difference in the rainfall rate. The rates may then be platted as in figures 22 and 23 and curves may be drawn to represent the relations for the different amounts of soil mois- ture. In figure 25 the circles show the rainfall and runoff rates determined in this manner for the following ranges of retention. From 0.00 to 0.65 inches From 0.65 to 1.30 inches From 1.65 to 2.30 inches From 2.40 to 2.75 inches / 2 3 /fates of Runoff in Inches, per Hour PIG 25 —RATES OF RAINFALL AND RUNOFF AT THE TAYLORS- VILLE DAM WHILE SOIL IS BECOMING SATURATED. Lines have been drawn through points calculated from experiments 10 and 11. 128 MIAMI CONSERVANCY DISTRICT The upper set of circles were determined from experiment 10 and the lower set, from experiment 11. Straight lines have been drawn through the points for each range in retention, curves J, K, L, and M ; and a 45° line, curve D, has been added as in figures 22 to 24. For comparative purposes, curve G of figure 23, representing saturated surface soil conditions in plats 3 and 4, has also been added. Since only two points are avail- able for each line, it is, of course, not known whether they should be straight or curved. They were drawn straight be- cause it has been shown that for saturated soil conditions the curves are straight. It will be noticed that curves J, K, L, and M are spaced rather uniformly, and that they are nearly parallel to curve D, the 45° line. However, if the experiments had been carried further, it is likely that the succeeding lines would have been increas- ingly closer together as they approached curve G, since curve G was based primarily on experiments in which the total sprink- ling time, as well as the amount of moisture in the soil when the observations were started, was greater than in exi)eriments 10 and 11. Curves J, K, L, and M illustrate, for plats 3 and 4, the var- iations in runoff and retention rates caused by variations in rainfall rates and in total retention. Retention rates are shown by the horizontal or vertical distances from the various lines to curve D. The variations in runoff and retention rates due to variations in soil texture may be indicated by showing, on fig- ure 25, points calculated for plat 1 for similar ranges in reten- tion during experiment 9. The soil in plat 1 when experiment 9 was begun contained about the same quantity of moisture as the soil in plats 3 and 4 when experiments 10 and 11 were started. Points calculated for experiment 9 are shown by the triangles in figure 25. It will be noticed that the triangle corresponding to a range in retention from 0.00 to 0.65 inches falls a considerable distance to the left of curve J, thus indicating a considerably greater rate of retention and a correspondingly smaller rate of runoff for plat 1. The succeeding points, however, fall increasingly closer to the curves for plats 3 and 4. The last point, corresponding to a retention from 2.40 to 2.75 inches, is very close to curve M. This means that at Taylorsville the runoff from the clay soils on the hills is appreciably greater than the runoff from the loam RAINFALL AND RUNOFF 129 in the valley, when the soil is dry, but not materially different when the soil is saturated. The effect of cultivation on retention and runoff may be shown on figure ^5 by platting points computed from experiment 6, made on plat 2 at Taylorsville. As previously noted the soil in this plat was spaded and raked before the experiment was started. Points computed for the ranges in retention used in determining curves J, K, L, and M, are shown by the squares in the upper part of the diagram. It will be noticed that the points corresponding to ranges from 0.00 to 0.65 inches and fro;m 0.65 to 1.30 inches, show no runoff whatever. Although th<3 curves in figure 17 show the runoff as beginning when the re- tention had amounted to about an inch, the average rate before the retention reached 1.30 inches was so small that it could not be shown in figure 25. The points corresponding to ranges in retention from 1.65 to 2.30 and from 2.40 to 2.75 inches both fall to the left of the point calculated from experiment 9, plat 1, for the range from 0.00 to 0.65 inches. This illustrates the rel- atively great amount of retention obtained by cultivation. Ref- erence to figure 17 shows that after runoff did begin on plat 2, the rate increased gradually throughout the first run. That the increase was gradual rather than abrupt was due to the pres- ence of air in the soil. The ranges in retention for which the points were computed are noted on the curves in figure 25. Points corresponding to similar ranges during other experiments, or during actual rain- falls, will fall on these lines only when the soil conditions, as regards texture, temperature, and moisture, at the beginning of the precipitation are the same as they were in experiments 10 and 11. Consequently, in order to use curves J, K, L, and M in calculating runoff from rainfall, it will be necessary to esti- mate the condition of the soil when the rainfall begins. This estimate can probably be made closely enough that the runoff rate will be determined with a fair degree of accuracy. For in- stance, if it is estimated that when a rainfall of 2 inches per hour, lasting an hour, began, the condition of the soil was the same as in experiments 10 and 11, the runoff rate of about an inch an hour, shown by the curves in figure 25, is probably ac- curate within 25 per cent or within a quarter of an inch an hour. This uncertainty would decrease in amount as the soil became saturated. While an uncertainty as great as 25 per cent is un- 130 MIAMI CONSERVANCY DISTRICT desirable, it is doubtful if an estimate based on judgment alone would be as accurate as one based on the curves of figure 25. The data secured at Moraine Park is hardly sufficient to warrant the preparation of a diagram such as figure 25. How- ever, the differences in retention and runoff due to variations in soil texture and surface slope may be studied in figures 13, 14, and 15, illustrating experiments 1, 3, and 4. Experiment 1 was made on the level bare soil plat in June, experiment 3 on the same plat in July, after the soil had been trampled and packed by cat- tle, and experiment 4 on the sloping bare soil plat when the soil was comparatively loose. Considering the first run of each ex- periment the retention rates are seen to have been considerably greater, and the runoff rates considerably less, during experi- ment 1 than during experiments 3 and 4. During experiment 4 the retention rates were slightly greater, and the runoff rates slightly less, than during experiment 3. This means that the increase in runoff due to the trampling and packing of the soil was slightly greater than that due to the increase in the slope of the surface. The average rate of retention of 0.90 inches per hour ob- tained on the sloping plat in run 16, for a period of an hour and 15 minutes, checks the conclusion reached in the preceding chap- ter that water can be absorbed by the bare soil at times during the summer when the soil is unusually dry, at a rate as great as 1.00 inch per hour for intervals as long as 30 minutes. CONDITIONS BEFORE RUNOFF BEGINS A knowledge of the conditions necessary before runoff be- gins is valuable in studying rainfall and runoff. During many showers of comparatively short duration no runoff takes place although the intensity of the precipitation may be relatively great. During other showers of longer duration and lesser in- tensity similar conditions exist as regards runoff. In order for runoff to begin it will be necessary for two conditions to be ful- filled. First, the precipitation must occur at a rate greater than the rate at which it can be absorbed by the soil ; and, second, the excess rate must continue long enough to fill the surface storage available by reason of the small depressions in the surface, accu- mulations of dead grass or leaves, growing vegetation, and other factors. The relative importance of these two conditions, of course, varies with the soil and surface characteristics. If the soil is bare and free from depressions, rates of precipitation and RAINFALL AND RUNOFF 131 soil absorption are predominant. If the soil is covered with a heavy sod or a deep deposit of forest litter, surface storage is the determining factor. While it is not possible to differentiate between these two fac- tors in a given instance, it is interesting to discuss their com- bined effect. Referring to run 1 of table 16 made on the level bare soil at Moraine Park when the ground was dry, it is seen that a rate of rainfall of 4.25 inches per hour caused runoff to begin in 2.0 minutes, or after a total of 0.14 inches had fallen. Run 1 of table 19, made on similar soil on the same day, showed that a rate of rainfall of 3.65 inches per hour caused runoff to begin in 2.5 minutes, or after a total of 0.15 inches had fallen. Run 16 of table 16, made on the sloping bare soil plat when the ground was 9ry, showed that a rate of 3.00 inches per hour re- sulted in runoff after 3.5 minutes, or after the total amounted to .18 inches. These results confirm the conclusion reached in the preceding chapter ; namely, that water cannot be absorbed by the bare soil at Moraine Park at any time, no matter how dry it is, at a rate as great as 3.00 inches per hour for periods as long as 5 minutes. The apparent exception to this, indicated by run 9 of table 16, in which a rate of 3.65 inches did not cause runoff until 5.5 minutes, is due to the different condition of the surface, the surface in this instance containing a considerably greater number of small depressions. Runs 1, 17, and 27 of table 17, made in August when the ground was about as dry as in the runs mentioned above, show that the soil at Taylorsville in plats 1, 3, and 4, is about the same as at Moraine Park as regards beginning of runoff. However, run 11, made on plat 2 where the soil had been spaded, shows a great difference. In this case a rate of 3.90 inches per hour did not cause a measurable quantity of runoff for 22 minutes, or until the total precipitation had amounted to 1.43 inches. The following morning, when the soil was practically saturated a rate of 1.85 inches per hour resulted in runoff in 6 minutes, or after 0.18 inches had fallen. Runs 8, 24, and 34 of table 17, made on plats 1, 3, and 4 in October, when the soil was somewhat drier than in August, show slightly greater values of retention preceding runoff, than do runs 1, 17, and 27. Other runs of tables 16 and 17 show the conditions before runoff when the ground is practically satu- rated. Figure 26 shows graphically the data discussed above. Times 132 MIAMI CONSERVANCY DISTRICT in minutes required for runoff to begin are platted as abscissas against the corresponding rates of precipitation as ordinates. Points corre(sponding to all runs in table 19 have been platted, but only those corresponding to the first run of each day have been platted from tables 16 and 17, since runs made on the same day were frequently only a few minutes apart. Different sym- Tz/rre in Minutes. FIG. 26.— INTENSITY AND DURATION OF RAINFALL BEFORE RUN- OFF BEGINS. Data secured at the Moraine Park and Taylorsville plats. The curve represents conditions at plats 3 and 4 at Taylorsville when the soil is dry. bols have been used to indicate the various plats on which the data was secured. Where the ground was wet when the rainfall began the points have been blackened ; where it was dry, they have been left white. In one or two instances the experiment number has been placed near the point in order to indicate a different soil condition. RAINFALL AND RUNOFF 133 Points representing runs 24 and 34 of table 17, and 2, 3, and 4 of table 19, made at the Taylorsville hill plats in October, have been balanced by a line, since for these runs the soil and surface conditions were practically the same. Points secured at Moraine Park do not cover a sufficient range to determine a curve. The amount of the surface storage in a given instance is in- dicated by the height of the hump, or peak, at the end of the re- tention curve, see figures 13 to 21, inclusive, caused by the drain- ing off of the water after the precipitation ceased. The amounts are small in all cases, as would be expected since there was rela- tively little vegetation on any of the plats for which mass curves were platted. The quantities vary from practically nothing to about 0.07 of an inch. CHAPTER v.— MONTHLY, SEASONAL, AND ANNUAL RAINFALL AND RUNOFF INTRODUCTORY General information regarding monthly, seasonal, and annual rainfall and runoff, their distribution throughout the year, their extreme variations, and the normal, monthly, seasonal, and an- nual amounts, are of importance in most hydraulic engineering work. Criticisms are often made of the method of discussing rain- fall and runoff by monthly, seasonal, or annual periods. These, as a rule, are based on the condition that the division date be- tween periods may fall within a time of storm rainfall, or of flood runoff ; or that due to snow accumulations, or ground water stor- age, precipitation during one period may affect the runoff in the following period. These objections, of course, are of more im- portance as regards studies based on the shorter periods. They also are of more importance with respect to studies of the larger drainage areas, inasmuch as flood runoff on the smaller areas is more nearly coincident with storm rainfall. Such criticisms are well founded and should be borne in mind. However, they apply principally to theoretical studies of laws governing runoff rather than to particular engineering prob- lems. Because such studies do not lead to the discovery of the laws of runoff is no reason why they should be wholly discon- tinued. In making studies of seasonal and annual rainfall and runoff the above objections may be partially met by using the "water year" rather than the calendar year, and by a judicious division of the year into seasons, or periods. Rafter, in his studies of rainfall and runoff,* used the water year ending November 30. He divided the year into three periods, namely, the storage pe- riod, including the months from December to May, the growing period, including the months from June to August ; and the re- plenishing period, including the months from September to No- TTr .*'^o® Relation of Rainfall to Runoff, by George W. Rafter, U. S. G. S. Water Supply Paper 80, 1903; also Hydrology of the State of New York, by George W. Rafter, Bulletin 85, New York State Museum. 13+ RAINFALL AND RUNOFF 135 vember. For conditions in the Miami Valley, the year ending September 30, which has been adopted by the Water Resources Branch of the U. S. Geological Survey, seems to be more satis- factory than the year ending November 30. It also seems better to consider the months from October to December as the re- plenishing period, the months from January to April as the stor- age period, and the months from May to September as the grow- ing period. This chapter will take up the studies of monthly, seasonal, and annual rainfall, runoff, percolation, and evaporation, which have been made for some of the drainage areas in the Miami Valley. A method of showing hydrological conditions by means of mass curves will also be described. If the amount of water stored in the ground, or on the ground is the same at the beginning and ending of a period of time, the difference between the total rainfall and the total runoff during this period must be equal to the total evaporation, using the term evaporation to include plant transpiration and evaporation of precipitation intercepted by vegetation as well as direct evapora- tion from soil or water surfaces. Studies of ground water flow indicate that in the Miami Valley variations in the amount of Table 20.— -Stations Used in Studies of Rainfall and Runoff Station Stream Drainage Area Records Available * Sidney Lockington Piqua Tadmor Pleasant Hill . ... West Milton Springfield Springfield Wright Dayton Franklin Germantown Seven Mile Hamilton Hamilton Miami River Loramie Creek. . . Miami River Miami River Stillwater River. . Stillwater River. . Buck Creek Mad River ...... Mad River Miami River Miami River Twin Creek Seven Mile Creek. Four Mile Creek. . Miami River Square Miles 555 255 842 1128 453 600 163 488 652 2525 2785 272 128 178 3672 Years. Inclusive 1915-1919 1916-1919 1912-1919 1915-1919 1917-1919 1915-1919 1915-1919 1915-1919 1915-1919 1894-1919 1917-1919 1915-1919 1915-1919 1915-1919 1911-1919 *Years ending September 30. water in the ground at the end of the water year are relatively small proportions of the yearly evaporation. Consequently, in the following tables and discussions dealing with annual quan- tities the term evaporation is used to mean the difference between the rainfall and runoff. However, in the studies of seasonal and 136 MIAMI CONSERVANCY DISTRICT monthly values the term retention has been used, since in these cases variations in the amount of water in the ground are com- paratively large. COMPILATION OF THE DATA Table 20 gives the gaging stations for which the data was compiled, the streams on which they are located, the areas drained, and the period of years for which records are available, the division into years being made on September 30 instead of on December 31. It will be noted that records for full years prior to 1915 are available only for the Piqua, Dayton, and Ham- ilton stations. The annual rainfall, runoff, retention, and ratio of runoff to rainfall, were tabulated for all records available, for each station in table 20, except Piqua. On account of unreliable gage height data at Piqua for some of the earlier years and for a part of the year 1918, the quantities were tabulated for the years 1915, 1916, 1917, and 1919 only. The annual, seasonal, and monthly rainfall, runoff, retention, ratio of runoff to rainfall, and tem- perature had been compiled for the entire record at the Dayton station just before the 1919 data was compiled. Since the 1919 values do not differ materially from the averages based on the 25-year record the studies have not been revised so as to include the 1919 data, except in the case of table 21. The proportions of annual runoff which appear as surface or flood runoff and as low water or ground water flow, were determined for the Dayton, Wright, West Milton, and Buck Creek stations. The Dayton station was chosen because of its comparatively long record and because it is representative of the average conditions throughout the Miami Valley. The other sta- tions were chosen because a cursory examination of the records, as well as the study of flood runoff given in the following chap- ters, indicated that the surface runoff from their drainage areas varies considerably. Mass curves were drawn only for the drain- age area of Mad River above Wright. The annual, seasonal, and monthly rainfall given in the tab- ulations are averages over the drainage areas above the stations, not the amounts recorded at the stations themselves; and are for the years ending September 30, rather than for the calendar years. For the years 1915 to 1919, inclusive, the annual amounts were determined by planimeter measurements on maps showing lines of equal annual rainfall. The annual, seasonal, and monthly RAINFALL AND RUNOFF 137 I C IS 01 a |2| KM P5" lis »OCTjCaCOU5COC5(Mt--i-HOOCDCO?Oi-i (MCOC^CCiCCC0MCOC0COCOC0Tt o t-cocoo CQOOlt-t-NCOCOCO(NC0CD000000CO«Dt-'^COU5 i-HCOOT-H 00OSCOi-H'^WDCDLCTHCDCDUSrH"«#lO NOoqcococot— oooo-^-^'^TfiLO^ CQCOCOCOCOCOCOOOCQCOCOCOCOCOCO osoo ^ -t^ o "^ lo o CO CDlOOOt-i-HrHOO C*CDOlO»O wiMcoc^OQoaiMoqcouatMN C^rHCOi-HlOi-H-^OlrHOiCD^D OSOCOOOr-fOiOONi— (OOS i-(t-00OC^'^00'XiCg000SU3Tj* S «^ § oi tuo^ o 3 13 OS :3 c O w 138 MIAMI CONSERVANCY DISTRICT rainfall for years prior to 1915, and the seasonal and monthly values for the years 1915 to 1918, where it was only necessary to obtain data for the Dayton and Hamilton stations, were obtained by averaging directly the records of all stations on the given drainage areas. While the latter method does not consider the distribution of stations, comparisons of the results obtained, with those obtained by the planimeter measurements, showed that for such large areas and with so many stations, the results by the shorter method are not appreciably in error. The values of annual, seasonal, and monthly runoff were cal- culated from the daily stream flow records, except in the case of the 1911 and 1912 records at Hamilton. These were obtained from the U. S. Geological Survey water supply papers, proper corrections being made for the flow in the Miami and Erie Canal, which is not included in the government data. In the studies of the relation of temperature to runoff the records at the Dayton co-operative station were utilized. It was not considered necessary to calculate the average temperature over the drainage area, inasmuch as any difference which may exist tends to be constant in amount, algebraically as well as arithmetically, and also tends to be relatively small. The method of estimating the proportions of annual runoff which appear as surface or flood flow and as low water or ground water flow, and the method of drawing mass curves, will be de- scribed later. ANNUAL RAINFALL AND RUNOFF Records for Years 1915-1919 Table 21 gives the annual rainfall, runoff, ratio of runoff to rainfall, and evaporation for all stations at which stream flow records are being compiled, for the years 1915 to 1919, inclusive. The average values, although very uncertain due to the shortness of the period, are also included. Studies based on the 25-year record at the Dayton station, discussed later, show that no one of these years was greatly different from normal. In order that the average values should be comparable throughout, missing records at the Lockington, Piqua, Pleasant Hill, and Franklin stations were estimated from the data at adjacent stations and were mcluded in the calculation of averages. An inspection of the table shows that the runoff in the Miami Valley is, on the average, about one-third of the rainfall The runoff from the Buck Creek drainage area seems to be somewhat /-^-- IRAINFALL ANB nUNOFF 139 less than in the other parts of the Mad River Valley. The total runoflE from the Mad River drainage area, as shown by the rec- ords at the Wright station, is the same as the total runoflf from the StillwAter River Valley, as shown by the West Milton rec- ords. The runoff in the upper Miami Valley seems to be prac- tically the same as the runoff in the Mad and Stillwater drain- age areas. The runoff in the Twin, Seven Mile, and Four Mile Creek areas, southwest of Dayton, seems to be higher than in the other parts of the valley. However, the records at these stations are somewhat more uncertain than those at the other stafdons, due to the greater difficulties "encountered in obtaining the stream flow data ; and it is doubtful if the runoff is actually much different from that of the other parts of the Miami Val- ley, The records at Hamilton, which are very satisfactory, seem to hear out this conclusion since they agree substantially with the Dayton records. The runoff during the year 1916 was comparatively high, and the evaporation comparatively low, due to the large amount ef storm rainfall that fell during the months of January, Feb- ruary, and March, when the available surface and ground stor-. age was a minimum and the evaporation rate insignificant. Records Above Hamilton '''" Table 22 gives the annual rainfall, runoff, evaporation, and ratio of runoff to rainfall, for the Hamilton station. Averages of the various quantities and' departures from the averages are also given ; and the maximum and minimum records are set in bold face type. The rainfall records are accurate throughout. The maximum error for a single year probably does not exceed two per cent. The runoff and evaporation records are believed to be fairly accurate for all years except 1912. The runoff of 15.6 inches given for 1912 is believed to be considerably too low, inasmuch as the record for Dayton, for the same year is 23.1 inches. The Dayton record is probably too high. There seems to be no rea- son why the amounts at these two stations should be so differ- ent. The records for the years 1915 to 1919, inclusive, agree very well, as previously mentioned. The rainfall during the year- 1912 was about the same as in 1913 but was much more uni- formly distributed throughout the year. Consequently the run- off would be expected to be greater than normal but less than in 1913. Probably the average of the two records, 19.3 inches. 140 MIAMI CONSERVANCY DISTRICT E a > S ii > 3 a c < a H 9 ^ 9 rri OOC^1000i-(0> .-( 00 00 rH W O 00 CO o t- CO t- 00 (M O 1— I "^ 00 CO -^ Cvl Oq i-l(M ffO 00 -«:I< Oi 10 t- as CO CO c 00 OS bi 1-H 1-H i-H i-H f-H 1-1 i-H rH t-H gf Cft OS Oi OS OS 05 05 OS OS 5 i-H tH ,-H ,-H ,-H ,-H ^ r-( ^ < RAINFALL AND RUNOFF 141 is about the true value. Assuming this figure to be correct the evaporation for 1912 would be reduced from 27.8 inches to 24.1 inches. Owing to the insufficiency of the stream flow records, the runoff for the year 1913 has been assumed to be the same as at Dayton, an assumption probably not much in error since the runoff at Dayton was well determined. While the period of record, only nine years, is too short to give very satisfactory information, the data seems to show that the runoff is the most variable quantity of the three and that the rainfall is the least variable. The average departure of the an- nual runoff from the mean value is seen to be 26.1 per cent ; the average departure of the annual evaporation, 12.5 per cent ; and the average departure of the annual rainfall, only 7.9 per cent. A considerably greater value for the average departure of an- nual rainfall was obtained for the Miami Valley above Dayton, where records for 25 years are available, as will be discussed later. The maximum annual rainfall is seen to be 1.12 times the mean and 1.33 times the minimum; the maximum annual runoff, 1.73 times the mean and 2.54 times the minimum ; and the maximum annual evaporation, 1.23 times the mean and 1.62 times the minimum. The average ratio of runoff to rainfall is seen to be 36.3 per cent, or slightly greater than one-third. The maximum value occurred during the year 1913, probably due to the memorable flood of that year. Although the ratio was unusually low in 1915, when the evaporation was a maximum due to the large amount of storm rainfall occurring during the summer months, the ac- tual minimum value occurred in 1918, amounting to only 25.3 per cent. Records Above Dayton Table 23 gives the annual rainfall, runoff, evaporation, tem- perature, and ratio of runoff to rainfall, for the years 1894 to 1918, inclusive, at the Dayton station, 25 years in all. Averages of the various quantities and departures from the averages are given as in table 22 ; and the maximum and minimum values are set in bold face type. The ratios of the maximum quantities to the mean and minimum quantities are also included. The rainfall and temperature records are fairly satisfactory throughout. The runoff and evaporation records are more re- liable during the years 1905 to 1918, inclusive, than they are dur- 142 MIAMI CONSERVANCY DISTRICT Sals Q .S > o EN OS'-i(MCOt-OOOCO»-((N^CC ^C0^C0U5CO^C0»-HCOOi'-HC^Ii~l CO CO N CO O '^ CO N CO T-H O "«#"*(NCC'-^OOi-(COCOOO(M'X>00 00«OOCT>C50 00?OOOOOOOCDt- WaSr-tOOt-WCOWOOqCC'<#"^CMO'^U3C^t-«DTtt>t>COOO ^1+1+1+1+1++ l+l+l+l I l+l 1+ B^ < a, s "H Q G ►^ > a a a '3 a a a < I 3 - MO) ■^ g (S « T}OOOOOOt*Oii-(OOOiO OOCOrHUD^DiOlOlCOOlOO-^OOSOfMNOit-IOOt^t-CO-^ iOOWOaS'^00'^t-'^«DT-i'OOOOOOOOCO-^OOOi-(CCC^HOOii-(COOqi— ICD -rHOit-OOO 00Q0THt-C0 00'^W00(£>OO(MTj*a5Ot-t--^OO»-trHTj• ft B< <) coM(Moocot-;DSst-- ^ i-l T-(t-H ,_(,_( T-H ,-H ,-H ^H CNI i-H t-H O^'>'-|°0t-t-COIX10O«3i-lTH(MMt-rtrtOli-l'^CO>-((M'*-* T|imoO(N-^c3>tDU5m(MTOt-05t-t-COU5C003-*00(N03T-(05 r-(r-l rHrt ^ ^ r-l rt rH N N r-( i-H rt CCOOOOr-ll005T)(COt>OrH-*C^t-0>OOtOt-C- + 1 + +++I +++I+++I++ (»OOWQ010COt-Oq(NrtU505tOCOt-005-*C>qt-THC)0-ONCOCO(N-H.-IU5^rHlO O pq < o H <; « o > H Q 12; » >>s» vft > s^ I I- fT r Pvaporafion Departure fromAverage. m % \ % ^ ^ ^ ^ J ^^ .% ^ ^ Runoff Departure from Av&rage. in % 6 a>< 'b' ,1 at (V ;3 CM f^ o H CSj I T O O § ^ ^ § ^ ^ '^•^ ^ ^^? Rainra/f Departure from Average /n ^ ^ o -a T3 si O a 3 3 en IS -2 g" ft CI c 150 MIAMI CONSERVANCY DISTRICT nates against the annual temperature departures as abscissas. No definite relation seems to be shown by any of these diagrams. The variations in annual temperature are so small that what- ever effect they may have on rainfall, runoff, or evaporation are not of sufficient magnitude to become noticeable. SEASONAL RAINFALL AND RUNOFF Table 24 gives the seasonal rainfall, runoff, retention, tem- perature, and ratio of runoff to rainfall for the drainage area of the Miami River above Dajrton. The year is divided into three seasons, or periods, as they are generally termed ; the re- plenishing, storage, and growing periods. The replenishing pe- riod includes the months of October, November, and December; the storage period, the months of January, February, March, and April; and the growing period, the months of May, June, July, August, and September. Averages of the various quantities for the twenty-five years of record are included in the table ; and the maximum and min- imum values are set in bold face type. The ratios of the maxi- mum to the mean and minimum quantities are also given. The data is shown graphically in figure 32. The mean values given near the bottom of table 24 show that on the average the rainfall is about 7.69 inches, or about 21 per cent of the mean annual, during the replenishing period ; about 12.23 inches, or 33 per cent of the mean annual, during the stor- age period; and about 17.13 inches, or 46 per cent of the mean annual, during the growing season. The average runoff appears to be about 1.69 inches, or 14 per cent of the mean annual, dur- ing the replenishing period ; about 7.22 inches, or 61 per cent of the mean annual, during the storage period; and about 2.96 inches, or 25 per cent of the mean annual, during the growing period. The average retention appears to be about 6.00 inches, or 24 per cent of the mean annual, during the replenishing pe- riod ; about 5.01 inches, or 20 per cent of the mean annual, dur- ing the storage period; and about 14.17 inches, or 56 per cent of the mean annual, during the growing season. The mean tem- perature is 43.1 degrees Fahrenheit, during the replenishing period, 38.2 degrees during the storage period, and 70.3 degrees during the growing period, the mean annual being 52.76 degrees, as previously mentioned. It is interesting to note the compara- tively high retention and low runoff during the growing season RAINFALL AND RUNOFF 151 »-^(Ml-^OT-t<^3^^^Ha50500o:n-^c-oooooc'-l T-i.-iiOr-it-oj.-irHas'^TtooT-iTitmoocoioooOT-HroTtiiM'* 1-H tH ^ Ca ^ T7< T-H 1-H T-t rH T-H rH COTt-COOTjlt-TtTO,-l^a^(MOO(MO-^ 05100 I = rH C^i-HT-H.-li-Hi-HTHrHi-Hi-H.-Hr-tTHi-Hi-HrHrHTHi-lrHi-H'.-Hi-H »-| COOOCOTHCq"^OOt-THOlOOOOCOC(>3CO(N(NCO»OOOaiC Tf CO CO t- Tl< Oi ■** CO CO ':D O -^ O ^ t- GO CO CO lO C^ t- Oi C— r rH W T-H i-H rH tH i-H 1-1 rH T-H (M i-H CC00000aiO'^00C0 ^c-oosot-t-cDcocicouDt-asOiOOiOiNOiCoaioot-cD •^co-^co-^cocococococococococo-^co-^cococococococc ■3 c fl « * 3"rt . ■rHCM 00 ■ CO rHr-. O t- OS CO CT» '^ W3 tH N -^ lO (M OO O 00 00 -^ 05 t* ITS CD Ti< lO (X> U3 rH .5 3 "rt ^ t>"^C0rH05C^CD-^CSiaiW-^OOCDrHt000OI>C0l0t>CI:>"^aiiOOOOiOCDOOU3'«:t< iOCqtMCOlOCr-Ot-COt-OiCOrHt-OsOiOOC^OO^-^OOrHOCO Olt-OSOOOOrHOO'^OOCOrHOlCDCOlOOS'^asi-HOiCOCDt-OSOO (U-^iq CDlOlOCDCDlOCOUD-'^lOOOCOCO'^COOOO-'^OtNlOOlrHiOLO lOrH t-00-^OCOOrHCOOOaiOrHa:0100-<*C^CDiHO*COCDir3li3rH cDooo(M(Mt-o:ioocooo(M(MiOiMoococooQoasoou:>t> 05t-00-^t-OOOOCOCOOOOlOi'<*U3»OOTHCOMrHt-CO 0) fr lOOOOCXNCO'^M'>*0005CDairHCDlOW3050a'X'CDlOC5ai« COtMOl-^CDWlCCDOLOairHtMCOrH-^fMOCO'^WCO'^NCO ■ •^ g C w (M'^OOCC'^OO^-^OOCDTtLOOCOlCOOlOrHCOtMCOOOOOOOOO o - ■ OOOOWlOCOOlOOit-fMOWCOCDOOOJi-HO^t-rHrHLOCOCD T-H rH lOrHC^IrHrH (M rH i-H C^ (M Ct-iocj-^o^oou5aio:)Oo:»corHiocoo rHlOOCOOO^'^COrHC-OOO'^OONCOaiCOCsl'^'OCOCgCD t:-C£)OaCgOSt-C£>CD»jOCDCDlOCDCDW(NCOlOUS--^COt-WlO"«:t< OUOO CO rH-^ Ph 1-1 csqcDt-i>-iotr-OiiocooococnooocgrHt>c:iff005(M'^oO"^co oot-ooiocooocr-rHWiooou^tr-OrHoscorHCOooooaioooai eg rHCO rH rHcgcg cg^ lo ocgtr-o^-^oscDCTjc^iOr-trHcgaiOsocDooMooioiocot-co ocoooooir-cgcg'^t-coaicooo-^C'^HOrH'^rHcgoitr-oio 00 t- Oi -<:J* O O t- t- lO Oi CO U3 Cr- 00 OO CO Ol O ^H lO t- t- OO CO U5 t- iH CO bH c g ■^iocDt>ooa50rHcgcoT(*iocotr-ooOTOrHcgco"<*u5cot-oo OiOlOiCSOiOlOOOOOOOOOOrHrHrHrHrHrHrHrHrH OOOOOOOOOOOO(7iOSO^0iOiOiOiOiOsaiCi s tl) 1 h( si 1 a; X » c a •S OS >■ a) ^■^ M O o < t- to U5 oo 00 1- TO <-i rt CO (M 00 1- 00 CO th «> u; ;g; 00 CO eq lo w m EzlSSJ lua cbONS]o!i!^rt'^=o.MrHT*THOCDmeD'^COCOCOCO'<*'«*COTl400COCO CO CD t> W 00 t- CO N CO CO 05 t- 1-H CO 00 ^ 05 '«:f< Oi CO ^ CO t- ^ T-( ?D H^ °2 ^ oos^coootO'-iascDWOi-HOco-^cowDooooNusi-HosO o^'-|^ COOCCOOJ'-llM'-l'^'-'CJ"3AC0t— (CJOlOOOCOOCOlOCDOCOOl— 1^-^ i-HC^05air-iiOOiOCIOCD0 0010 0»-(t-lO»-l»-IOCDO"«*0-^ ^ i-( GO CO CO '«* CO N CO IM CO (M -^ CD '«* '^ CO C^ X 03 C5 OS Oi CO -«*CO'^CsIlOCOtOCooocooi-m3coco'^ ooooc lOCD^ '<*rHCOCO'^(MC0CO(NC0C0WC^{M'^U3'^^C0C0O5"^"**"^Tt< aaocDco'^i»Oiaa'«*coot-oo5ioe*loo{MOsO"**cot>cDt- O^OOt-NNOSOt-CDOOOt-COlOCOOW'^TflOt-iftCDCO 1-H (M (N CO (M ^ rH !M rH CO -^ CO 1-H N CO •fl* CO -^ -^ "«* CO i-l N CO CO W iH CO C0CD-^CDt-CD0ilONT-(O'^t-05OS00t>C0T-H^lOt-C0i-l»rt C0C0"^i-H00W(M0000Ot-O"«H0000C0OCg00U9C^C0t-t-l0 oco 00 t- CcD ^^t--^ W3 inir- w OS as CD CO CO CO lo CD 05 -^ lo t* 05 Tj* N CO o '^ eg Cd 00 nmo ■ ■ "^ C0Or-l(Mi-Hr-(C0i-(O"^(MTHO©Tt tr- O eg -^ i-H CO CD Tj) 00 O CO us CO CV| O CO OS CO CO LO o ©■* * iH CO eg o o o '^ eg eg I-H eg eg iH CO eg eg © CO © lo od eg CO eg I-H '«* egcgrn be f-i '^lCCDt-OOOSOrHC^aCO"^U3COt-OOOSO»-HCdCOTft-(OOt-C00'^;D00W00Tll'^i-CO'-HCOCOW"^COff^t-OSi-H"^"^OOCOC^(M-^CO'^CO i-H 1-1 M 'Ht-^-^t--Cqt-'<*'^CN'<#T-^'*'^C30COC^COON{N "^^.QQ -^lOCOt-OOOiO^-KMCO-^lCCDt-OOaiOi-fWCO^if^COC-CO ^ TO 00 00 00 00 OS OS OS OS Oi as Oi OS oa OS OS as OS Oi Oi OS OS OS m I i_.-j^^^^,-ii— l,-^1-^1-^T-l1— ItHt— iT-ti— it-HT-*i-('-('-H'-ii— IT— < 158 MIAMI CONSERVANCY DISTRICT OS s B < •5 B O ^ TjC0M0iOTfC00i T-t (MWfNtM i-HiM ffOi-HMi-H ■^CCOCDOOt-00-^O(NTHC0l0i-H^HU3O»— lt>0«0»-HOOO CKI i-H i-H rH(M .-H « t-H T-H ^0JOC0Oi-H00iOl>00CCO(N'^COr-(CqNC^C»3-^"^COCOCOi-ITj-iOW^(NOSCOlOCOi-li-li-llO4 a W^ ^)5S^2S252r:;5^2?'^"3'=ot-ooosorH(Mco-^iocDt-oo i—li-li— |T-Hi-(r-ti-Hi-• § o o a 9 -2 cs ^ ^ «'0 a C^-^COO'^ ^ rHCO CCiOO t- C^3 OCO ?DO 00lO(M (MO M U3 t> O CN C0^C000C0t*C0C0C<|WC »-H »-H r-I N (M i-H T-H tH rH i-H QQ r-t C^I tJ* U5 i-H CC CO CO CO Tj* (M CM CDOCDt-OONt-t-OTHOOCOt-OOCOt-WIMOOO'^OOCOOOOO T-Hoooooat-Oi{Moacoi-(t>i>-^i-t(Moooi-HCO'i-H00O00t>b-0SO00C0Ot-000S(MW CO"^COU3CD-^CD^r-IO"^Tt<-rJCt> CO"<*'^t>"*:J'"^e^lCOCOUDOO(MCOWt-"^CDOS10lOCOi-HOOOS'^ C^l rH (M rH tH oooiO■I-^cst-■I-^'^ri^(Mcoccn^-lO'^T^oot>u^t-HOO^^-mt-tM »-li-t-^COlOCOCO(MC^CD'i:}t-OOCl0l00Sl0OlC0Si-Hi-H0:i0Si-OC^t-'^M OS-^WWCDt-C^JrHOOOOCOCOi-Hi-HlCOSCDOSOOOSt-HOSkOfM C0OC005CD»-(0STtO »-H i-H 50 (M 1— I rH 1— I Oa T-H rH »-H (M CO rH ■^ CO 1— I rH "^ rH -«^U3CC>t-OOOiOi-HC^CO'^lOCOt-OOOSOi-l(MCO"«i*U^COt>CO 050i050iOSOiOOOOOOOOOOrHi-HrHi— (i-HrHi-trHi— ( OOOOOOOOOOOOOSOiOiOiOiOiOiOSOSOSOSOSOiOSOSOSOiOSOi COW OCOON''* t-OOCO C~ c 5 00CCW^OtD(NI00t-^-^CD0000(M(NC0t-O'^"^O-^ONM t^O{>J-^OlOt-NlOt-C£)COOt-as-«#t-OOOCOCDOOU5COOi iOCOU3'^CftC^COC^I0001COCDOt-(>t-t:-t-C-t>t-t-tr-t-t-b-t-lr-t-«Ot-Cr-i> (Mt-t-iOiOCMO-^CDOOCqCDOOOOOONO-^-^WO-^t-lOCO 00 to CO c£> 00' -^ cotoo»oioco"^'^ioioe»airtcoi£3t-t-cooicocaCDCOOJC^(MCO(NCDt-«DCDOOOOeON i-HOJCO»-HOO-^CDmOW5lOT-(C]CO05O]COC3O0 C]C0COC0C0C<10aWCCqOOO:iOOCOCDlOCO(M-^Tf^Ct-l>lOU3 00NO^ CO00C0COC0CO(NC0C0G^C0CMNCOCOCO00(M(NMCO0Q0000CD0iO^C0(N0^'<:t*C000C^- ^■^i-HOOi-HO^OSW'^^DlOeOCOCOO'^Ot^TflOTjfOOCDCCOD -Oi-H (M 00 CO OS OO ^O CO 05 Oi Oa Oi CD N CO 00 r-( CO "^ ■ O 1-H (M-^U3CO»-(lOTfWOi— IT-Ht-OCOOOCOOOINCOOOCOOOO 1-1 ■ ■ CS CO OS -i-(COCD ■coco CO ■ • t-lOl-H ■ 1-HCO lO tH i-t 0) CO'^iOCDt-OOOSOrHCOCOTtOOOiOT-iCOCO'^lOCOt-00 OiOSOlOSOlOSOSOOOOOOOOOOrHT-H tHi.— ( rH rH rH T-H f-H ooooooooooooooosososososasosoiososososososoaosososoi a 3 ^ \.s- m u *-« O.S 15 2.S . aS> •a 6^ a o RAINFALL AND RUNOFF 161 In this case the three temperature curves are kept together, the three rainfall curves together, and so on. The mean distribution of the rainfall, runoff, retention, and temperature throughout the year is shovi^n by the group of curves in the center of figure 33, as well as by the average values given at the bottom of the tables. The minimum mean monthly rain- fall occurs during the month of February and amounts to 2.24 inches. If the record were increased so as to correspond to a month of 30 days instead of 28 the amount would still be less than the record for November, the next lowest month. March seems to be the month of heaviest rainfall, the average for this month being 3.80 inches. However, the value of 3.73 inches ob- tained for the months of June and July is practically as great. There appears to be a decrease in rainfall during April, the average amount for this month being only 2.98 inches. It is interesting to note that a similar decrease during this month is shown by the majority of the diagrams for southwestern Ohio, published by the U. S. Weather Bureau, in Volume II of Bulletin W.* The rainfall is generally low during the months of October, November, and December ; and generally high during the months of May, June, and July. The distribution of runoff during the year is slightly differ- ent from the rainfall distribution, inasmuch as it is generally low during the summer months. However, the month of greatest runoff is the same as the month of greatest rainfall, the monthly runoff being a maximum during March, amounting to 2.62 inches. September is the month of lowest runoff, the average for this month being only 0.37 inches. The curve of mean retention follows, in a way, the curve of mean temperature, being high in the summer and low in the winter. The minimum monthly retention occurs in February, the month of minimum rainfall, and amounts to only 0.70 inches. June is the month of maximum retention, the average for this month being 3.11 inches. The curve of the average ratio of runoff to rainfall is just the reverse of the temperature curve, being high in the winter and low in the summer. The maximum ratio of monthly runoff to monthly rainfall occurs in March, being 68.8 per cent. The minimum ratio occurs in August, being 11.8 per cent. A study of the maximum and minimum values, and of the ratios given at the bottom of the tables, shows that the various ♦Summary of Climatological Data East of the Mississippi River, Bul- letin W, Volume II, U. S. Weather Bureau, Washington, D. C, 1912. 162 UIAMI CONSERVANCY DISTRICT 100 15 60 10 FIG. 33.— MAXIMUM, MEAN, AND MINIMUM MONTHLY RAINFALL, RUNOFF, RETENTION, AND TEMPERATURE ABOVE DAYTON. The ratio of the mean monthly runoff to the mean monthly rainfall, expressed as a percentage, has been added to the group of mean values. RAINFALL AND RUNOFF 163 FIG. 34.— MAXIMUM, MEAN, AND MINIMUM MONTHLY RAINFALL, RUNOFF, RETENTION, AND TEMPERATURE ABOVE DAYTON. Curves shown in figure 33 have been arranged so as to show more clearly the differences between the maximum and the minimum values of the various quantities. 164 MIAMI CONSERVANCY DISTRICT quantities are all quite variable, much more variable than the annual or seasonal quantities, as would naturally be expected. The runoff again seems to be considerably more variable than either the rainfall or the retention. The maximum rainfall during any one month of the entire record occurred in March, 1913, due to the great storm of March 24 to 27, and amounted to 10.51 inches. The maximum runoflf for any one month also occurred during March, 1913, and amounted to the same value. The ground water flow at this time was unusually high due to the heavy precipitation of the pre- ceding January. The mihimum rainfall for the entire period occurred in March, 1910, and amounted to only 0.07 of an inch. The minimum runoflf occurred in November, 1902, calendar year 1901, amounting to only 0.14 of an inch. Maximum and minimum values of monthly retention and ratio of runoflf to rainfall, as well as all individual values, are more or less erratic due to the short period of time considered. The runoflf very frequently is not comparable with the rainfall for the same month. Probably the chief value of the data given in tables 27 and 28 is to show this. During the winter months the runoflf is often greater than the precipitation, due either to snow accumulations or to floods in the early part of the month caused by heavy precipitation during the late part of the pre- ceding month. The runoflf during February, 1916, was greater than the rainfall, due to the flood runoflf resulting from the heavy precipitation of January 30 and 31. The runoflf of February, 1918, was greater than the rainfall, due to the melting of the heavy snows which had fallen during January. Other negative values of retention might be similarly explained. The curves of maximum, mean, and minimum monthly tem- perature given in figure 34 need no discussion. However, it is interesting to note that the minimum records for October, De- cember, and January all occurred during the unusually severe winter of 1917 and 1918. SURFACE AND GROUND WATER FLOW Table 30 shows the relation between surface and ground water flow for the drainage areas of the Miami River above Day- ton, of Mad River above Wright, of Buck Creek above Springfield, and of Stillwater River above West Milton. The annual amounts of surface runoflf and of ground water runoflf are given in inches depth over the drainage areas and in percentages of the total RAINFALL AND RUNOFF 165 runoff, for each year for which records are available. The total annual runoff in inches is also included. The maximum and minimum records are indicated as before. The proportions of annual runoff which appear as surface or flood flow and as low water or ground water flow can be deter- mined only approximately. No exact separation is possible. In calculating the data given in table 30, the separation was made on the hydrographs in the following manner: Lines represent- ing the rate of ground water flow were drawn so as to pass through the low points only, as shown in figures 35 to 39, inclu- sive. The endeavor was to draw the lines so that the increased flow of tiles immediately after a flood, that is, the drainage of the surface soil, would be included in the surface or flood runoff rather than in the ground water runoff, since such flow acts more nearly like surface flow than like low water flow. It was also assumed that no percolation occurs during the growing season or before the latter part of the replenishing period, that is, dur- ing the period from about May 1 to about December 1. Having arbitrarily drawn the curve representing the rate of ground water flow, it was, of course, simply a matter of calculation to determine the total amounts of surface and ground water runoff during the year. Reference to table 30 shows that in the Miami Valley above Dayton the surface flow is about two-thirds of the total runoff, and the ground water flow about one-third. In the Buck Creek Valley the surface flow contributes only about 44 per cent of the total and the ground water flow about 56 per cent. In the Mad River Valley above Wright, including the Buck Creek Valley, the surface flow amounts to about 53 per cent of the total and the ground water flow, to about 47 per cent. In the Stillwater River Valley the surface flow constitutes about 79 per cent of the total and the ground water flow, only about 21 per cent. These wide differences in the proportions of surface and ground water flow are the result of variations in geological and soil conditions. In the Mad River Valley there is relatively large underground storage in deep deposits of glacial gravel, while the comparatively loose and shallow surface soil permits rapid percolation. Gravel deposits are less extensive in the Miami Valley above Dayton, and still less frequent in the Stillwater Valley. Over a considerable portion of the latter basin there are but a few feet of residual clay soil overlying the bed rock, which generally is limestone. On these drainage areas surface slope 166 MIAMI CONSERVANCY DISTRICT Table 30.— Surface and Ground Water Runoff in the Miami Valley Year Ending September 30 Total Surface Runoff Ground Water Kuuoff in Inches Inches % of Total Inches % of Total DRAINAGE AREA OF THE MIAMI RIVER ABOVE DAYTON 1894 4.92 1.90 ' 38.6 3.02 61,4 1895 3.72 1.29 34.7 2.43 65.3 1896 8.08 4.47 55.3 3.61 44.7 1897 12.78 8.19 64.1 4.59 35.9 1898 14.70 10.41 70.8 4.29 29.2 1899 9.72 5.32 54.7 4.40 45.3 1900 6.58 3.16 48.0 3.42 52.0 1901 5.65 2.67 47.3 2.98 52.7 1902 3.76 1.56 41.5 2.20 58.5 1903 12.56 7.77 61.9 4.79 38.1 1904 13.09 9.38 71.7 3.71 28.3 1905 7.08 4.46 63.0 2.62 37.0 1906 9.18 5.25 57.2 3.93 42.8 1907 17.16 11.38 66.3 5.78 33.7 1908 17.72 12.52 70.7 5.20 29.3 1909 13.12 8.31 63.3 4.81 36.7 1910 15.13 10.56 69.8 4.57 30.2 1911 13.91 9.18 66.0 4.73 34.0 1912 23.09 16.18 70.1 6.91 29.9 1913 24.36 19.71 80.9 4.65 19.1 1914 8.33 4.95 59.4 3.38 40.6 1915 12.09 8.58 71.0 3.51 29.0 1916 19.25 14.19 73.7 5.06 26,3 1917 11.43 7.41 64.8 4.02 35.2 1918 9.42 6.44 68.4 2.98 31.6 1919 11.15 6.71 60.2 4.44 39.8 Average 11.85 7.77 65.6 4.08 34.4 DRAINAGE AREA OF BUCK CREEK ABOVE SPRINGFIELD 1915 8.34 3.92 47.0 4.42 53.0 1916 14.75 6.73 45.6 8.02 54.4 1917 10.25 4.77 46.5 5.48 53.5 1918 10.10 4.64 45.9 5.46 54.1 1919 11.04 4.10 37 1 6.94 62.9 Average 10.89 4.83 44.3 6.06 55.7 DRAINAGE AREA OF MAD RIVER ABOVE WRIGHT 1915 12.03 6.86 57.0 5.17 43,0 1916 19.39 11.85 61.1 7.54 38,9 1917 13.48 6.75 50.1 6.73 49,9 1918 11.09 5.69 51.3 5.40 48.7 1919 12.89 5.56 43.1 7.33 56.9 Average 13.78 7.34 53.3 6.44 46.7 DRAINAGE AREA OF STILLWATER RIVER ABOVE WEST MILTON 1915 12.63 10.63 84,2 2.00 15.8 1916 17 63 14,30 81.1 3.33 18.9 1917 12.33 8.73 70.8 3.60 29.2 1918 13.30 10.34 77.7 2.96 22.3 1919 12.00 9.48 79.0 2,52 21.0 Average 13 F8 10.70 78 8 2 88 21.2 RAINFALL AND RUNOFF 167 has but little influence on runoff. In fact the surface slopes are steeper over the Mad River drainage area where percolation is great, than on the Stillwater where flood runoff predominates. Considering the Dayton records the surface flow is seen to vary from only 1.29 inches in 1895, the year of minimum annual runoff, to 19.71 inches in 1913, the year of maximum annual run- off. The percentages of the totals for these years were 34.7 and 80.9 respectively, which are the minimum and maximum per- centages. T.he ground water flow varied from 2.20 inches in 1902, the year in which the rainfall during the storage period, when practically all of the percolation occurs, was a minimum, to 6.91 inches in 1912. However, as previously mentioned, the rec- ord for the year 1912 is believed to be too high. Probably the value of 5.78 inches given for the year 1907 represents the true maximum amount. The ground water flow was a maximum per- centage of the total in 1895, the year of minimum surface and minimum total runoff; and a minimum percentage of the total in 1913, the year of maximum surface and maximum total runoff. The minimum and maximum values for the other drainage areas are, of course, very uncertain due to the shortness of the record. However, a study of the averages of the Dayton records for the years 1915 to 1919 inclusive, and of those for the entire period of record, indicates that the averages for the other drain- age areas are not greatly in error. The annual ground water runoff is much less variable than the annual surface flow, as would be expected. The maximum value of the annual surface runoff in the Miami Valley above Dayton, 19.71 inches, is about 15.8 times the minimum value, while the maximum value of the ground water runoff, using the 1912 record, is only about 3.14 times the minimum value. MASS CURVES The hydrology of a drainage area may be shown conveniently by means of mass curves. Such curves have been drawn for the Mad River Valley above Wright for the years 1915 to 1919, in- clusive. They are shown in flgures 35 to 39, one year's records being shown in each figure. Separate curves have been drawn to show the rainfall, ground water runoff, flood runoff, total run- off, retention, soil absorption, percolation, and evaporation. In order to avoid confusion the rainfall curve was arbitrarily started at 10 inches on the scale instead of at 0. Hydrographs showing the rate of discharge and the arbitrary separation of 168 MIAMI CONSERVANCY DISTRICT Jan. I Feb\ Mar. \ Apr. | May \jurte FIG. 35.— HYDROLOGY OF THE MAD RIVER VALLEY ABOVE WRIGHT DURING 1915 RAINFALL AND RUNOFF 169 A^ay June \Ju/y Aug. Iseptl FIG. 36.— HYDROLOGY OF THE MAD RIVER VALLEY ABOVE WRIGHT DURING 1916. 170 MIAMI CONSERVANCY DISTRICT PIG. 37.— HYDROLOGY OP THE MAD RIVER VALLEY ABOVE WRIGHT DURING 1917. RAINFALL AND RUNOFF 171 FIG. 38.-HYDROLOGY OF THE MAD RIVER VALLEY ABOVE WRIGHT DURING 1918. 172 MIAMI CONSERVANCY DISTRICT FIG. 39.— HYDROLOGY OP THE MAD RIVER VALLEY ABOVE WRIGHT DURING 1919. RAINFALL AND RUNOFF 173 flood runoff from ground water runoff are platted in the upper parts of the figures. No explanation of the four curves mentioned first is needed. They were simply drawn from the rainfall and runoff data de- termined as previously described. The retention curve was ob- .tained by subtracting the flood runoff from the rainfall. Hori- zontal lines have been drawn under the peaks, or humps, on the retention curves, thus indicating the surface storage. The larg- er humps, having a comparatively long duration and occurring during the winter months, are due to precipitation in the form of snow. The two larger humps due to this cause have been marked "snow", but the others have not been designated. The sharp peaks of comparatively short duration represent storage on the ground or in the stream channels during flood periods. The retention curve, as modified by the horizontal lines, rep- resents the total of the soil absorption, percolation, and evapora- tion curves, the ground water runoff being maintained by the ground water storage or percolation water. The lines under the humps should really have been drawn so as to slope upward to- ward the right instead of horizontal, since soil absorption and evaporation are continuous, to some extent at least, throughout the storm period. The soil absorption curves, or soil storage curves as they might be termed, were drawn after a careful study of the Mo- raine Park soil moisture records given in chapter III. It was assumed that there is a variation of five inches in the amount of moisture in the soil during the year; that the soil reaches its dryest condition sometime late in the summer, during August or September; that it gradually fills with moisture in the fall, during the months of September, October, November, and De- cember; and that it then remains saturated until late in the spring, when it begins to dry out due to transpiration and in- creased soil evaporation. In drawing the percolation curves it was assumed that no percolation occurs during the summer or early fall months ; that is, that percolation ceases about the time the soil begins to dry out in the spring and does not begin until late in the fall, about the time the surface soil becomes saturated. It was also as- sumed that the percolation curve joins the ground water runoff curve at the time percolation begins. The former assumption is believed to be essentially correct for the Miami Valley except in very unusual instances. The latter assumption, while more 174 MIAMI CONSERVANCY DISTRICT or less arbitrary, does not lead to appreciable error. Of course the percolation curves do not need to touch the ground water runoff curves at the time percolation begins. They could have been drawn a fixed distance above, at this time; that is, the curves shown in the figures could have been arbitrarily raised a certain amount. By drawing the percolation curves in this way the total per- colation during a given winter and spring was made just great enough to maintain the ground water flow until percolation began in the following fall or winter. It is not believed that in the Miami Valley percolation during a given storage period ever affects greatly the ground water flow after percolation begins in the succeeding fall; or, in other words, that the ground water level at the time percolation begins ever varies greatly from year to year. The minimum amount of annual ground water runoff in the Miami Valley above Dayton occurred in 1902, following a year in which the rainfall was only 30.1 inches, or about 7 inches less than normal. In 1914, when the rainfall was 32.3 inches, or about 5 inches less than normal, following a year in which the rainfall was 42.9 inches, or about 6 inches more than normal, the ground water runoff amounted to 3.38 inches, or about 1.18 inches more than in 1902. A part of this 1.18 inches was prob- ably due to the 2 inches greater rainfall in 1914. It does not seem probable that percolation during a particularly wet season ever increases the ground water runoff during the following year by as much as an inch. In drawing the percolation curves during the winter and spring when percolation was taking place, consideration was given to the rainfall distribution and form of occurrence as well as to the temperature and other meteorological conditions. The evaporation curves were determined by subtracting from the retention curves, or from the horizontal lines under the reten- tion curves, the sum of the soil absorption and percolation curves. In doing this points were taken about a month apart as shown on the diagrams. The attempt was to show the general shape of the evaporation curve throughout the year rather than the daily variations. It is only at the points indicated that the evapora- tion is equal to the retention less the sum of the soil absorption and percolation. In order for this relation to hold throughout it would be necessary to throw the small irregularities of the re- tention curves into the evaporation, absorption, or percolation curves. It is probable that during the summer and fall the ir- RAINFALL AND RUNOFF 175 regularities should be thrown into the evaporation and absorp- tion curves ; and that during the winter and spring they should be thrown into the evaporation and percolation curves. More irregularities would be expected in the evaporation curve during the summer than during the winter. Data on transpiration and soil evaporation seems to indicate that during the summer months rates of evaporation as great as a half an inch a day, or even greater, may occur immediately after a heavy rain. Such con- ditions would cause jumps in the evaporation curve somewhat similar to those in the retention curve. Table 31 gives the monthly evaporation, taken from the curves, for the Mad River Valley above Wright for each. year; and also the average amount for each month, based on the five years' records. The average monthly evaporation from a water cfiCokf^- Jan. Feb. Mar. Apr. May June July Aug. Sept. Oct. Nov. Ded Months FIG. 40.— MONTHLY EVAPORATION IN THE MAD RIVER VALLEY ABOVE WRIGHT For comparative purposes a curve has been added showing the monthly evaporation from water surface at Columbus. surface at Columbus, for the years 1907 and 1908, for the months of April to November is also included.* Figure 40 shows graph- ically the data on average monthly evaporation given in table 31. It will be noticed that the evaporation from a water surface is ♦Water Resources of Illinois by A. H. Horton, Report of Rivers and Lakes Commission, State of Illinois, Springfield, Illinois, 1914, page 310. 176 MIAMI CONSERVANCY DISTRICT somewhat higher than the evaporation from the land, especially during the fall months. The evaporation from the land is com- paratively low during the fall because the available supply of moisture has been depleted by the high rates of transpiration and soil evaporation during the growing season. Table 31. -Monthly Evaporation on Drainage Area Above Wright Station, in inches, 1915 to 1919, inclusive Years 'Average Month Average Evaporation* from 1915 I9ie 1917 1918 1919 Water Surface October .... 1.04 1.18 .63 1.20 .65 .94 4.58 November. . .67 .57 .57 .70 .48 .60 2.35 67 60 55 53 .42 .55 January.. . . February . . . March 55 55 55 .52 .40 .51 47 48 45 40 55 .47 .80 .85 .75 .60 1.20 .84 April. 1.20 1.17 1.15 1.60 1.50 1.32 3.05 May 1.70 3.90 3.60 4.15 3.70 3.41 4.';3 June 5.15 6.30 7.20 4.30 4.25 5.44 5.99 July 7.60 4.05 5.65 5.90 7.70 6.18 6.37 August 4.15 2.60 2.65 5.95 4.95 4.06 6.81 September.. 2.60 1.90 .95 2.15 1.80 1.88 6.11 , Total 26 60 24.15 24.70 28.00 27.60 26.21 *Based on records taken at Columbus, 1907-1908. Table 32. — Seasonal Evaporation in Mad River Valley Above Wright, in Inches Season Years Average 1915 1916 1917 1918 1919 Replenishing Period Storage Period 2.38 3.02 21.20 2.35 3.05 18.75 1.75 2.90 20.05 2.43 3.12 22.45 1.55 3.65 22.40 2.09 3.15 20 97 Growing Period Total 26 60 24.15 24.70 28' 00 27.60 26 21 Table 32 gives the seasonal evaporation for the Mad River Valley calculated from the data in table 31. The data indicates that, on the average, about 80 per cent of the annual evaporation occurs in the growing period, during the months of May to Sep- tember, inclusive. This leaves only 20 per cent for the replenish- ing and storage periods, or the seven months from October to April, inclusive. CHAPTER VI.— RAINFALL AND RUNOFF DURING 1913 FLOOD The flood of March, 1913, was not only the most severe of which there is record in this valley, but as regards damage was also the greatest that has occurred in the eastern half of the United States since the days of first settlement, or since floods first began to attract attention. A description of this flood and of the damage it wrought has been published in an earlier report.* It was caused primarily by hard rains which commenced on March 23, and continued with but little interruption until the 27th. Contributing factors were a saturated soil when the rain began, as a result of previous rains, and low temperatures which reduced evaporation to insignificant rates and affected the per- colation of water through the soil. Being the maximum flood on record it was necessary, of course, to make detailed investigations of the rainfall and runoff for use in the design of the flood prevention works. The hy- draulic design of the works has been described in volume VII of the technical reports, f The results of such studies were also needed in determining the benefits and damages resulting from the construction of the retarding basins. This chapter will give the rainfall and runoff date secured, and will discuss the various studies which were made imme- diately following the flood. RAINFALL The daily rainfall over the Miami River drainage area dur- ing the storm of March, 1913, is shown in figure 41. Maps are included showing one-inch isohyetals for the 24-hour periods end- ing at 7 p. m. of March 23, 24, 25, and 26. A map for March 27 has not been reproduced because the precipitation on that day amounted to only about half an inch. Figure 42 shows the total *The Miami Valley and the 1913 Flood, by Arthur E. Morgan, Chief Engineer, Technical Reports, Part I, The Miami Conservancy District, Dayton, Ohio, 1917. tHydraulics of the Miami Flood Control Project, by S. M. Woodward, Technical Reports, Part VII, The Miami Conservancy District, Dayton, Ohio, 1920. 177 178 MIAMI CONSERVANCY DISTRICT EMEN I.SO^^f Rainfall of March 24- KENTON3.60 DENNISDH t.2S 'mammz.io «/ scale of miles Rainfall of March 25 Rainfall of March 26 FIG. 41.— DAILY RAINFALL OVER THE MIAMI VALLEY DURING THE STORM OF MARCH, 1913. The amounts recorded at the various stations are iHdicated by the figures written after the names of the stations. RAINFALL AND RUNOFF 179 Rainfall, March 25 and Z4- Rainfall, March ?3 to 25 /he/. KENTON 8.65 /£IV BKEMEHB.SO-J OS^ ii''" '^ { BELL E FONTAINE Rainfall, March 23 to26lnc/. Rainfall. March 25 to 27 inc./. FIG. 42. — CUMULATED RAINFALL OVER THE MIAMI VALLEY DUR- ING THE STORM OF MARCH, 1913. Rainfall shown on each map is the total from the beginning of the storm up to 7 p. m. of the second date noted under the map. 180 MIAMI CONSERVANCY DISTRICT accumulated precipitation for the periods ending at 7 p. m. of March 24, 25, 26, and 27. The amounts of rainfall recorded at the various places where gages are maintained are shown by the figures placed after the names of the stations. At the river stations, where the rainfall is measured in the morning, the amounts estimated for the 24- hour periods ending at 7 p. m. are enclosed in brackets. Fortunately there were a number of well distributed rain gages in this part of the Ohio Valley in 1913. Reports from about 50 stations were utilized in the preparation of figures 41 and 42, many of which were outside the area shown on the maps, some being in Indiana and Kentucky. Although Dayton was the only regular Weather Bureau station located within the Miami Valley at that time, there were several in nearby cities, as at Cincinnati, Indianapolis, Fort Wayne, Toledo, Sandusky, and Columbus. Only a partial graphical record of rainfall was se- cured at Dayton. Owing to the flooding of the business section of the city, in which the oflice is located, the triple register could not be kept in operation, the clock stopping at 4 :30 p. m. on March 25. Figure 43 shows the distribution of the precipitation, as re- gards time, at the above mentioned regular stations, platted from data published by the U. S. Weather Bureau in Bulletin Z.* The abscissas represent time, and the ordinates, the amount of rainfall in inches per hour, the amounts being shown by horizon- tal lines extending through the hours in which they occurred. Since the actual amounts used in platting the horizontal lines were usually for one-hour periods, the ordinates, in such cases, indicate the total precipitation for each hour as well as the rate. The total rainfall for any Jength of time at a given station is represented by the area under the portion of the curve corre- sponding to that time ; that is, by the product of the rate and its duration. Hourly readings were not available for March 27, or for March 26 at Fort Wayne. The precipitation in the latter in- stance, and probably in the former at some stations, was in the form of snow. Since the precipitation was small in both cases the rates have been computed and platted as having continued *The Floods of 1913 in the Rivers of the Ohio and Lower Mississippi Valleys, by Alfred J. Henry, Meteorologist, Bulletin Z, U. S. Weather Bureau, Washington, D. C, 1913. RAINFALL AND RUNOFF 181 I V) I I .25 25 25 25 25 .50 .Z5 .50 .25 25 t J5 m llii fl.nn AnJ jna ~J\ A r^ Columbus ^ \ 5andu5ky Jb. I To/edo Mh r\ Fort Wayne ^A^ ■ t Z^ Mar.Z3 Jj I Dayton W L. Jl Cincinnati u. rffi m ^ Average __ Mar. 24 K- Al^ M//: 25 M7/: f5 /i/ac 27 PIG. 43.— HOURLY RAINFALL AT STATIONS NEAR THE MIAMI VALLEY DURING THE STORM OF MARCH, 1913. Based on graphical records at the U. S. Weather Bureau stations. 182 MIAMI CONSERVANCY DISTRICT uniformly throughout the 24 hours. The actual rainfall on the 27th probably ended by evening or sooner. It will be noticed that the rain seemed to occur in showers at the various stations, the showers being a little more intense at Cincinnati and Dayton than at the other cities. The times of oc- currence of the principal showers at the different places seem to agree fairly well. Consequently, a curve has been added showing the average hourly amount based on the seven actual records. Although the total precipitation over the Miami Valley above Dayton determined from the isohyetal map in figure 42 amounted to 9.60 inches while the total determined from the curve in figure 43 amounted to only 6.40 inches, this curve is the only available indication of the rainfall distribution above Dayton. It appears that the rainfall began during the morning of the 23rd and continued during the greater part of the day, ending at about the time the daily readings were being taken by the co- operative observers. The average over the Miami Valley above Dayton amounted to 1.20 inches. The precipitation was heaviest in the northern part of the drainage area. The rain began again about midnight and fell almost continuously throughout the 24th, 25th, and 26th. On the 24th the total up to 7 p. m. averaged 2.20 inches above Dayton, the greatest precipitation occurring over the headwaters of Twin Creek, further south than on the pre- ceding day. The greatest rainfall occurred on the 25th, averag- ing 4.11 inches, and registering a maximum of 5.61 inches at Bellefontaine, about 55 miles northeast of Dayton. It was on the morning of this date that the rivers, which had been steadily rising, overtopped the levees in the principal cities. Near the following midnight the highest stages were attained at places between Dayton and Hamilton. On the 26th the average rain- fall amounted to 1.62 inches; on the 27th it amounted to 0.47 inches. The latter was of little consequence, however, as the waters were then everywhere receding rapidly. RUNOFF Unfortunately there were only three river gages in the Miami Valley at the time of the flood. These were the gages maintained by the U. S. Weather Bureau at Piqua and Dayton and by the U. S. Geological Survey at Hamilton. Discharge measurements during flood stages had been secured only at Hamilton. During the summer and fall following the flood, extensive hydrographic surveys were made for the purpose of determining RAINFALL AND RUNOFF 183 the maximum rates of discharge at various places along the streams. These surveys and the methods used in calculating the maximum rates of runoff have been described in an earlier re- port.* The results obtained are shown on the map in figure 44. At each location where measurements were made the quantities are indicated by means of a fraction and a quotient. The num- erator represents the maximum total rate of discharge in second feet at the given place; the denominator, the drainage area in square miles above that place; and the quotient, the maximum rate of runoff in second feet per square mile over the given drain- age area. It will be noticed that the rates of runoff are unusually high at all places. However, in the case of the smaller drainage areas, these rates continued for but small fractions of a day. Figure 45 shows the hydrographs of the 1913 flood at Piqua, Dayton, and Hamilton. These were determined as accurately as possible from the gage readings taken during the flood, from the maximum rates of discharge based on the surveys, and from current meter gagings made during subsequent smaller floods. The rates of runoff in inches per day over the drainage areas are platted as ordinates, and the times, as abscissas. For com- parative purposes the average curve of hourly rainfall obtained in figure 43 has also been included. It will be noticed that the rates of runoff above Piqua were lower than those above Dayton and Hamilton, throughout the entire flood period. The rates above Dayton and Hamilton agree very well. It is believed that the reasons for the lower rates of runoff above Piqua are the less intense precipitation above that station, the storage in the Loramie and Lewistown reser- voirs, which has been estimated to be equivalent to a depth of about a quarter of an inch over the total drainage area above Piqua, and the somewhat less rolling topography in the upper parts of the valley. Considering the steeply rising portions of the hydrographs there seems to have been a difference in time of about 6 hours between the Piqua and Dayton curves, and of about 8 hours be- tween the Dayton and Hamilton curves. It will be noticed that at Piqua the river reached its crest at about 10 o'clock Tuesday morning, March 25, and then remained stationary about four hours. At Dayton, however, the maximum stage was not reached until about midnight Tuesday, although the river was within a ♦Calculation of Flow in Open Channels, by ^.^nE.Houk Technical Reports, Part IV, The Miami Conservancy District, Dayton, Ohio, 191». 184 MIAMI CONSERVANCY DISTRICT 5700 , , hDAYTON 649 / / ( fMIAMISBURB ^257000 . 27Z2 N \ ^♦MIDDLETOWN / DISCHARGE IN SEC:FT DRAINAGE AREA IN SQ.MI RUN-OFF IN SEC.-FT PER SQ. MI.: 257000 384 000 Scale of Miles 10 20 PIG. 44.— MAXIMUM RATES OP RUNOFF IN THE MIAMI VALLEY DURING THE FLOOD OF MARCH, 1913. The figures are written opposite the places at which measurements were made subsequent to the flood, and give tftie runoff in second feet per square mile as mdicated in the lower right hand corner RAINFALL AND RUNOFF 185 RainFa/tin /nches per Z4- Hours, Depfh in inches per 2.4- Hours 186 MIAMI CONSERVANCY DISTRICT foot of its highest stage at noon. At Hamilton the crest was reached at about 3 a. m. Wednesday, March 26. Investigations made within a few weeks after the flood showed that the Still- water River just above Dayton reached its crest at about noon Tuesday; that the Miami River above its junction with the Still- water reached its crest at about 7 p. m. Tuesday evening; and that the Mad River above the Miami reached its crest at about midnight Tuesday. In each case the river remained practically stationary for a few hours before it began to fall. Consequently the highest stage at Dayton was due to the coordination of the highest stages in the Upper Miami and Mad Rivers. If the crest in the Stillwater had been delayed a few hours the stage at Dayton would have been higher than it actually was. Owing to the distribution of the most intense rainfall as re- gards drainage areas as well as regards time, the highest stages in the various streams were caused by the local runoff rather than by the runoff from thfe upper drainage areas. No indica- tions of the occurrence of a definite flood wave were found, ex- cept, possibly, in the case of Mad River. In this instance the investigations seemed to indicate a difference in the time of crest of about 12 hours between Springfleld and Dayton, a distance of about 25 miles. The comparatively slow movement of the crest was due to the great amount of storage in the valley. The Still- water River was at its crest at practically the same time from Covington to Dayton, a distance of about 30 miles; as was also the Miami from DeGraff to Tippecanoe City a distance of about 45 miles. RELATION OF RUNOFF TO RAINFALL Reference to figure 45 shows that while the rainfall curve is comparatively irregular the runoff curves are comparatively smooth. This is because the runoff curves are for large drain- age areas, 842 to 3672 square miles. In such cases the irregu- larities of the rainfall tend to be eliminated by the effects of stor- age on the ground and in the numerous small tributary drains, as well as by the time required for the runoff to reach the main streams. The conditions may be said to be analagous to the op- eration of a retarding basin, the rainfall curve corresponding to the inflow to the basin, and the runoff curves, to the outflow. The runoff curve for a drainage area of a few square miles would un- RAIl^FALL AND RUNOFF 187 doubtedly have shown irregularities corresponding to those in the rainfall curve. It will be noticed that considerable rainfall occurred during the first two days, while in the same period there was compara- tively little runoff. It also appears that when the rainfall ceased rather abruptly, the runoff continued for several days after- wards. This condition was due to the surface storage mentioned above, the water held by the small depressions and irregularities in the surface of the ground, as well as in the stream channels, draining out gradually after the rainfall ceased. In order to consider properly the total surface runoff result- ing from the rainfall of March 23 to 27, it is necessary to keep in mind the weather and ground conditions preceding the storm. January was an unusually wet month, the total precipitation, which was well distributed through the month, amounting to over seven inches. February was drier than usual, the rainfall totaling an inch less than the normal of three inches for that month, and occurring mostly in the last three days. March was wet throughout. From the first to the 21st moderate rains were recorded at all of the gaging stations on about ten days. On the 21st the precipitation throughout the valley averaged nearly a half an inch. It is evident from these conditions that at the beginning of the rain on March 23 the ground was fully saturated, that the ground water flow was greater than usual, and that there was some surface runoff in the streams as a result of the precipi- tation of March 21. The latter factor can be eliminated in the determination of the flood runoff, caused by the storm of March 23 to 27, by totaling the runoff only to the time when the amount of water in the stream channel was the same as when the flood began. This time was estimated to be in the evening of March 31. However, the effect of the ground water flow must be al- lowed for in a different manner, since some runoff was being maintained by underground storage during the entire flood period. From a study of the daily discharges before and after the flood it is estimated that 0.05 of an inch per day should be deducted from the total runoff, in order to obtain the true sur- face, or flood, runoff. As nearly as can be determined this amount would be the same for Piqua, Dayton, and Hamilton. Figure 46 shows mass curves of rainfall, flood runoff, and retention for the flood of March, 1913. Rainfall and retention curves are shown only for the total drainage area above Dayton ; 188 MIAMI CONSERVANCY DISTRICT flood runoff curves are shown for Piqua, Dayton, and Hamilton. The rainfall curve was calculated from the average curve shown in figure 45, arbitrarily raising the latter so that the total pre- cipitation for each day agreed with the value determined from the isohyetal map in figure 41. The flood runoff curves were likewise calculated from the rate curves of figure 45, deducting the ground water flow from the total, as mentioned above. The retention curve is simply the difference between the rainfall curve and the Dayton flood runoff curve. FIG. 46.— RAINFALL, RUNOFF, AND RETENTION DURING THE FLOOD OF MARCH, 1913. Curves show total values of the various quantities up to any instant. The retention represents soil absorption, evaporation, and storage on the ground and in the streams. That soil absorption and evaporation were comparatively unimportant during this flood is indicated by the fact that the curve falls rapidly after the most intense precipitation ceased, reaching a value of about 0.84 inches by the evening of March 31. That this would be true RAINFALL AND RUNOFF 189 was indicated, of course, by the previous rainfall and weather conditions described above. The maximum retention occurred at about 8 a. m., March 25, amounting to 5.50 inches. • If it is estimated that the total soil absorption and evaporation up to this time had amounted to a half an inch, the storage on the ground and in the streams would be about 5 inches. Of this amount it is probable that the proportion held in the main channels and in the adjacent over- flow sections was about 2 inches, and the portion held in the small streams and ponds and on the ground surface, about 3 inches. Table 33 gives the daily rainfall and flood runoff for each day of the flood, for Piqua, Dayton, and Hamilton. The values Table 33.— Daily Rainfall and Runoff Above Piqua, Dayton, During the 1913 Flood, in Inches and Hamilton Date Piqua Dayton Hamilton Rainfall Runoff Rainfall Runoff Rainfall Runoff March 23 1.57 1.77 4.14 1.50 0.40 2. A3 1.41 0.87 0.64 0.46 0.32 0.29 1.20 2.20 4.11 1.62 0.47 ■o!25' 2.12 3.20 1.52 0.79 0.46 0.26 0.16 l.OO 2.40 4.04 1.62 0.51 oiio' 1.11 3.30 1.85 0.87 0.47 0,29 0.22 24 25 26 27 28 29 30 31 Totals for flood 9.38 6.85 73.0 2.f3 9.60 8.76 91.2 84 9.57 8.21 85.8 1.36 Ratio of runoff to rainfall Total retention, in inches of daily rainfall were obtained from the isohyetal maps in figure 41, and the values of daily runoff, from the mass curves of fig- ure 46. The amounts are for the 24 hours ending at 7 p. m. in all cases. Total quantities for the flood period and ratios of total runoff to total rainfall are given at the bottom of the table. ^It will be noticed that the total runoff at Piqua was somewhat smaller than at Dayton and Hamilton. The total at Piqua amounted to 6.85 inches, or to only about 73 per cent of the rain- fall, while the total at Dayton was 8.76 inches, or about 91 per cent of the rainfall, and at Hamilton, 8.21 inches, or about 86 per cent of the rainfall. The total amounts at Dayton and Ham- ilton agree very well. The conditions believed to be responsible for the smaller runoff above Piqua have already been mentioned. CHAPTER VIL— RAINFALL AND RUNOFF DURING FLOODS SINCE MARCH, 1913 Since March, 1913, only floods of nominal size have occurred in the Miami Valley. During the 1913 flood, the maximum stage at Dayton was 29 feet, or 6 feet above the levees. The highest water since that time was during the flood of April, 1920, when the maximum stage was 16.2 feet, or about 7 feet below the tops of the levees. Practically no flooding has occurred within the cities during this period although farm lands near the var- ious streams have been flooded several times. Hqwever, this flooding generally occurred during the winter and early spring months, when but little if any damage was sustained. RAINFALL, RUNOFF, AND RETENTION DURING FLOODS Table 34 gives the total rainfall, runoff, and retention in inches, and the ratio of the total runoff to the total rainfall, in per cent, for the larger floods. The dates of the storm periods are given in the table headings. The dates of the flood periods have not been indicated. In general, however, they began shortly after the storm rainfall began and continued from two to five days after the rainfall ceased, the exact time, of course, varying with the nature of the storm and with the topography and extent of the drainage area. For comparative purposes the sizes of the drainage areas above the various stations, have been included. The topography and geology of the Miami Valley have been dis- cussed briefly in chapter I. A map of the valley showing the gaging stations is given in figure 1, page 17. The values of rainfall included in the table are averages for the drainage areas, for the entire storm periods. They were de- termined by planimeter measurements on isohyetal maps, simi- lar to those in figures 41 and 42, pages 178 and 179. The values of runoff represent the total flood or surface runoff caused by the indicated rainfall. The runoff maintained by the ground water storage has been deducted in each case, as in the determination of the 1913 flood runoff. The values of retention are simply the 190 RAINFALL AND RUNOFF 191 differences between the storm rainfall and the flood runoff. The value for a given storm includes the total amount of moisture taken up by the soil and the total evaporation during the storm period. Storms are arranged chronologically. The data for the larger drainage areas is believed to be more accurate than most similar data which has been published. The valley is well supplied with rainfall stations, as shown in figure 1 ; the rating curves on the principal streams have been well de- veloped; and the main portions of the flood hydrographs were determined by special readings taken every hour or every two hours. While the data for the smaller areas is not so accurate, it is believed that with one or two exceptions the errors are not excessive. DESCRIPTIVE NOTES On account of the condensed form in which the data in table 34 is presented it has not been feasible to include descriptive notes. Since such notes are desirable in any study of the sub- ject, they are given in the following paragraphs. July 7-8, 1915 The rainfall in the Miami Valley during the storm of July 7 to 8, 1915, was most intense in the evening of the 7th, from about nine to about ten o'clock. Practically the entire precipi- tation fell sometime during that evening, only a few hundredths of an inch falling after midnight. Although a half an inch fell on the 11th and 12th, it has not been included in the storm as it caused no surface runoff. The most intense precipitation at Moraine Park has been shovra graphically in figure 9, page 80. The soil moisture experiments carried on at Moraine Park, described in chapter III, show that the soil was not saturated when the rain began, although some rain had fallen each day during the period from July 1 to 5, inclusive, and although the total precipitation during the month of June had been slightly greater than normal. In fact, observations made on the 8th, after the rain had ceased, showed that even then the soil was not saturated to a depth of 6 inches. The streams began rising about 9 p. m. of the 7th. Highest stages were recorded sometime during the 8th in the upper por- tions of the valley ; on the morning of the 9th at Dayton ; and during the evening of the 8th at Hamilton; the crest at Hamilton 192 MIAMI CONSERVANCY DISTRICT a c s Pi • a a c s a H m o 2 ^ CO .2§Si-Sg m ■a(S~-aa?: >o d CM CM CO o o i^-sl ^^ CO CD >^ CO 05 I-» t^ =p §8 »iO (N m 03 ■o Ol o CO oc -* CM Ol «- £ (>< CO (M (N CO ""^ CM CM CO CO «D o c O 1 c c l^rlss 00 rH O 00 t^ »o 00 -0 'I* CO ■* o> CM CO CD CO CO r^ CO o CO CM o ^ CD 00 CM ^ >) ^ P |ggl CD d d d CO d 05 d T-i o d CD d o d d 1 d i s Cd CO ■>* ^ 03 o o CO »o oa ^ ■* l?-5" 05 C<3 c> Ol 00 lo t^ 00 o -* l^ ■* T-H ■3 M S d 1—1 T-H d o o d d 1— 1 - .— t '"* ^~* i S CO »o rq ^_, -1^ OS CO CM CO CI »n CO CO s •i|.sl >o ^ »o CO t~ CO 00 00 t^ o 00 Ol 00 £ « = 1— t ^ ^ *"* ^ .—1 '~* ^^ .— 1 CM 1— 1 ^ T-H U3 d d cri -*> lO d CO r~ d CO T-H d lO (NtD CD 00 t^ i> l> CM CO CO t^ oo 00 ■* l§sl (N t^ tf 1— 1 o CO o lO CO 00 o CO I^ &a> (N >o 00 o o kO t^ 00 lO m 00 St^ 1-H d d >— 1 '"' O) CM CM .— 1 d d d tH i s 00 l:^ CO 00 02 o lO CD t^ CM 00 CO lO |?.^| CO CO rH CO 00 c» (N CO d d CM CM CM i-H CM CM CM >o O -* r^ CS 03 CO wo T~< CO o 00 C U3 a> ^ CO on oa 00 CO o t^ -a 00 o •^ t-- CO o 1-H s d o o CD lO CS CTi o Oi CO CO CM ■^3-^1 o (N .-H (N 00 03 T— t o CM o CM CO CM 05 CM CM T-H CM CO ca lO >o (N « g CO on CM iO CM 00 00 CM •5 s ■^ CO >o p< CM ^- w .— < TK CD s c^ T-4 CD CO s a §■§ 2 V ^ o » aj 1 > a c ^ 1 c s c J o -g •1 o 1 RAINFALL AND RUNOFF 193 t- Ratio of Run- off to Rain- fall in per cent s s 00 CO CO ss CO g3 S CO )< Tfl IN o o 05 eq iS o 1-H t— i-H 1— ( t> a> OS OS o 05 1-H o « S fH N '"' •"' C^ eq '^ 1-* 1-H 1-H CM CM -> cq CM •sisi-sl W S^'sS'S S to S t^ >o r^ l> to S >o to to 0S<» eti.S S «; ■* Co CM a -S 00 CO W) h. <- O m »o »0 o IS o o o o g. d IM 1-H 1-H 1-H d d d 1-H i s ■* s 1^ Ol o »o ffi iO O o 05 00 00 t^ -^ o l^-l o CO o »-H ■*; d d to d to o 00 o 00 d r-H d 1-1 d 1-H d to d i^aJ m (M CO s 1 00 00 to CO CO t>- l> ■■C CO to h* CO (-1 't ' m lO ■* O t- 1-H «^ CO o o 1 c ev: a d CO 1-H 1-H in o d o o 1-H d o- CO «o o CO CM 00 CM to ta 1-H )i Jo-s-s cc (N « ■* Oi 00 CT> O CO CO t- t- "3 '*] ■3 « s 1— 1 — »-H d d d 1-H y-t 1-H d d d 1— 1 i s nr 00 ■^ CO 00 o ■^ CO o: OS Ol OJ IC ^ •S^B* cr s OCi o o rt^owaSs tc 00 1~- tc s CO (N iH CI 1 ll^ o d 5 o o 5 ■a CO d 1-H d 00 d o o d CO d OJ 1-1 d 00 1-H d 1-H d s m g & X •3 S CO CO CO ^ ta T|1 o o o ■* •o s 1 o o o o .c o o o o o o o ■NbI QC s (N 1 ' ^ s 05 1-H 3 S g CO OJ 1-H 00 l-H s CD ^"•'s -i 1-1 O d 1—1 1-H 1—1 »— 1 o o o o o lass^ IC w (N n ^ o s X S S S cq 8S g CM fi-^|S »/: « OC ,-H •^ to 1-H -*. tc ^ ^ ^ > a c ■1 C 1 1 OQ 1 n •3 CQ 1 ■E •1 1 ^ 1 i 1 a IS 1 1 194 MIAMI CONSERVANCY DISTRICT ^ O 1 c fl ^^ ~~" O) ■S«»>-3s2 4 Oi M ^ i CO ■^ t^ CO 00 Iv 00 CO 1-H •* IN tf^'S««S UD IV .3 US CO tv 1—1 IN lb Pfl o ft ^* !>■ W CO '-I to CO CD 03 >0 CO 00 1* IV 00 l§sl 03 t^ ■-i o ■a CT> lo ^ '"' '" o CM CO IM CO O 1-H lO CO IN ci (N 1-H IN i ® o> Tf 00 =3 sa "^ IN >a CO no (§=■"1 « O CO 05 CO a> CO CO CO CO ? (N M ^ O) IN cq o o - 1— 1-H IN CO 1-H 1-1 •ii.sl ■* IV -*l CB tv O o a, o CD o to O) 0> .-H o 00 IN IM O 1-H •* CD 00 -H i ^ t>- tv CO 00 (N tv ^_, CO ^ u: lO lO 00 IV 00 " »-l 00 m d ^ (N CTi OJ 1-H d d CO d CO d d CC c CO d 1-H d 00 d d d •i%sl (M to 1— 1 l^ (N 05 t^ 05 IN t^ lO r~ CT> ■^ o -* CO CO t^ ■:C CC •o tv IN lO ■* «•" s CO (M CO (K (N CD IV «r: OS « n* OJ 1>i< CO lO ■ o M S IM 0» cs (N IN i-H IN N IN (N cq 1-H 1-H 1-H (N o ^o 1 b'S t- ItiMl lO *" 00 1-H s? CD 00 CO •s d CO 00 -* CO s »c '3 09 CO -*" on IScI en tv c l> o W3 ?s 1* (N ■^ c^ CO % o CO 03 tr cr o 1—1 1-H |v 1 *o 1 «*'"! c o c c o o 1—1 o o o d 1-H -H d d i"l lO o c w: Iv ■* 1—1 >o O d -i «5 «: a. c X 1 c c s f- s 1 1 1 1 1 o 1 1 1 c c 'i 1 1 ,2 1 s RAINFALL AND RUNOFF 195 o atio Run- ain- 11 in rcent to CO Ol ^ OJ (N OS >o d t^ lO d 3t.^sas o> 00 t^ CO CO c<; CO CO lO 1^ (^ CO M CO CO 1^ l^ I— 1 00 00 t— 1 o o CO CO CO OS CO CO CO 00 OS CO !N o r-t 00 ^ H on CO CO 00 05 00 CO Q r-- ,_! (M CD t>- ,_^ IM uo OJ CO 00 o T-H ° t^ OS 00 ^ lO CO 00 CO ■*' CO CO C^ CO 1-H c^ r-H (M (N CM CO CO c^ g o ■HbI ,_H a> fl lO o CO UD (M CO U5 (M , 1 »o 00 o CO -^ t^ IJ t^ in ^ O o 1> W 00 CO t^ « s "3 -* lO ■* ■* ■* >o W >o ■* ■* Tjl ■* ■* ■* A CO ^ (T OS di CO t^ t^ nr i-H TjH OS 00 IM ^• Oi >o 1^ '^ >ffl 00 00 (N CO CO t~. s lO CO t^ "5 M »i3 CO Q^ CO o OS CO ^ nr »o o ,_, CO iO cs rH CO en CO t-. ■* o a or OS (N t^ -* CO t^ d e: r-i o d (N i-H o rH I-H s i. s (M CO CO -H 00 Tt* iO CO cr CO OS o ■* o CO |?.s-§ t^ CO t~ o I> CO o lO o •* t^ o « s i-H N (N ^ IN CO d i-H ^ « cd (M CO CO T-t i= 1 CO ■>t ■* 00 CO 00 o Tf 00 OS i-H o ta 00 o (N (M -a< >o t^ r-H c (M IM (M lO t~ ^ " l-H CO CO CO CO CO CO (M CO CC CO CO ■* ■* ■* CO ■mm ir •-H o CO ■* CO ^ la^l 00 1-1 Jo (M CD ^ ■^ CO CO CO CO CO CO o ^ o i cl ^ CO ^ CO CO £2 >o oq d 00 (-< CO CI •-H T— t o Oi M^'SK-S iSnl rr CO CO CO (M -* cr (M CO ■* I:- Ttl rt s CO ^ ^ CO CO o t^ |«"S 1—1 -^ ^^ '-' 1-H --I CO cq (M '-' r-l ^ ^ '"' •"• l^--t oc cc CO s ^ s g s 5 CO IM 00 o § ■3 S, 1 O O c o o o o o O o o '-' 1-H CVI o ■^^J fe CO •Sc CO o s t^ g CO 1 (3 3 S' OS 19* MIAMI CONSERVANCY DISTRICT being caused by heavy runoff from the areas below Dayton. The streams then fell rapidly, reached normal stages by the 14th in all cases. Dec. 24, 1915-Jan. 7, 1916 The flood of the last of December, 1915, and the first of Jan- uary, 1916, was due to rain falling on a partially melted accu- mulation of snow. Precipitation occurred nearly every day dur- ing the period from December 24 to January 7. During the last of December the temperature at Dayton varied from a few de- grees below the freezing point to a few degrees above, so that the precipitation was sometimes snow and sometimes rain. Above Dayton the temperatures were slightly lower. The result was that by the end of the month there was an accumulation of from 6 to 12 inches of snow above Dajrton. This was melted by a rainfall of from one to one and a half inches falling on January 1 and 2. Some flood runoff occurred during the last of December. However, the main part of the flood began on the morning of January 1. The highest stages were reached sometime during the 1st or 2nd, at all places except Hamilton. At this station the crest occurred early in the morning of the 3rd. The surface runoff had drained out completely by the 10th. The rainfall during November and the first three weeks of December was not greatly different from normal; so that the ground was practically saturated when the storm began. As no cold waves had occurred up to that time the ground could not have contained any appreciable amount of frost. January 10-13, 1916 The most intense precipitation during the storm of January 10-13, 1916, occurred on the 12th, although appreciable amounts fell on each of the other dates. On the 10th, the first day of the storm, the average rainfall over the different drainage areas varied from 0.15 inches in the Miami Valley above Piqua, to 0.37 inches in the Buck^reek Valley. On the 11th, the amounts were about the same except that they were slightly heavier below Day- ton. On the 12th, the amounts varied from 0.51 inches in the Buck Creek Valley to 1.14 inches above Germantown. The fol- lowing day, the last day of the storm, the rainfall was compar- atively light below Dayton. However, above Dayton it varied RAINFALL AND RUNOFF 197 from an average of 0.06 inches in the Stillwater Valley to 0.54 inches in the Buck Creek Valley. The precipitation occurred as rain on the 10th, 11th, and 12th, but changed to snow on the 13th, as the temperature fell with the passing of the storm. The streams began rising on the 11th and reached their high- est stages on the 13th. They then fell rapidly reaching normal stages by the 17th or 18th. The weather and soil conditions during the last of Decem- ber, 1915, and the first few days of January, 1916, are indicated by the description of the preceding storm. Comparatively low temperatures from January 6 to 8, inclusive, froze the ground to a depth of from three to five inches, as indicated by observa- tions at Moraine Park. Consequently, the conditions were con- ducive to high rates of runoff. January 26-31, 1916 The storm of January 26 to 31, 1916, followed close after the storm of January 10 to 13. The weather from the 13th to the 19th, inclusive, was cold with temperatures as low as 3 degrees below zero at Dayton. Consequently the soil, which had been saturated by the preceding storm, froze to a depth of several inches. The weather warmed up on the 20th, and from a quar- ter to a half an inch of rain fell each day on the 20th, 21st, and 22nd. From the 23rd to the 26th the weather was warm and fair. Only a few hundredths of an inch fell on the 26th. On the 27th, 28th, and 29th, the precipitation over the different areas varied from 0.06 to 0.34 inches, from 0.18 to 0.45 inches, and from 0.38 to 0.69 inches, respectively. On the 30th and 31st the most intense precipitation occurred. The amounts varied from 0.67 inches above the Buck Creek station to 1.57 inches above West Milton, on the 30th; and from 0.41 inches above Springfield to 0.94 inches above Lockington, on the 31st. The rivers began rising on the 28th, reached their highest stages on January 31 and February 1, and then fell rapidly reach- ing comparatively low stages by February 6 in all cases. The rate of falling was increased somewhat by the comparatively cold weather which followed the storm. March 21-22, 1916 Practically all of the rainfall during the storm of March 21 to 22, 1916, fell on the 22nd, only a few hundredths of an inch 198 MIAMI CONSERVANCY DISTRICT falling on the 21st. The soil was saturated due to the precipita- tion of the preceding part of the winter but was not frozen. Al- though some snow had fallen during the earlier part of the month, it had all melted by the 21st. The rivers began rising on the 21st, reached their highest stages on the 22nd or 23rd, and then fell rather slowly. The flood runoff has been determined for the period beginning on the 21st and ending on the 26th. Flood runoff after the 26th was not included because the following storm began on the 25th. As there was still some surface runoff in the streams on the 26th, caused by the storm of the 21st and 22nd, the values given in the tables may be slightly low for this storm and slightly high for the following one. March 25-28, 1916 Only a few hundredths of an inch fell on the 25th. On the 26th the average precipitation for the different drainage areas varied from 0.28 to 1.07 inches, the greater amounts falling in the lower portions of the valley. On the 27th the values varied from 0.35 to 1.81 inches, the greater amounts on this date fall- ing on the Mad River drainage area. The precipitation on the 28th was comparatively small, the averages for the various ba- sins varying from 0.04 to 0.24 inches, The streams began rising again on the 26th, reached their highest stages on the 27th and 28th, and then fell rapidly. By April 3 the surface runoff had entirely passed the city of Ham- ilton, the lowest station included in the table. May 6-7, 1916 The entire precipitation in this storm fell during the night of May 6 and 7. Although a few hundredths of an inch fell during the period from May 10 to 13, no surface runoff resulted and con- sequently the amounts have not been included in the table. The ground was fairly wet when the rain began, due to the rains of April and of May 2, 3, and 4, but was not fully saturated. Evaporation and transpiration rates were considerably higher than they had been during the preceding months, but had not yet reached their maximum summer values. The streams rose rapidly, reaching their crest stages on the 7th and 8th, and then fell rapidly. By the 13th the stage at Ham- ilton had fallen to a normal value. RAINFALL AND RUNOFF 199 January 3-6, 1917 The most intense precipitation during the storm of January 3 to 6, 1917, fell on the 5th and 6th. The rainfall on the 3rd and 4th was comparatively light, amounting to only a few hundredths of an inch each day. On the 5th and 6th the total amounts var- ied from 1.67 inches in the Miami Valley -above Sidney and in the Buck Creek Valley above Springfield to 2.12 inches in the Twin Creek Valley above Germantown. The soil was practically saturated when the rain began and was not frozen, although some freezing weather had occurred during the preceding month. There was no snow on the ground. The precipitation during the entire storm was in the form of rain. The rivers began rising on the 5th, reached their highest stages on the 5th and 6th, and then fell steadily. The total flood runoff had passed Hamilton by the 10th. March 11-14, 1917 The heaviest rainfall during the storm of March 11 to 14, 1917, fell on the 13th, the amounts on that day varying from 0.95 inches above the Wright, Seven Mile, and Four Mile stations, to 1.49 inches above Pleasant Hill. The rainfall averaged about a half an inch throughout the valley on the 11th, and from 0.05 to 0.37 inches on the 12th, the heaviest precipitation on the lat- ter date falling in the southern part of the valley. On the 14th the rainfall varied from 0.07 inches in the Mad River Valley to 0.38 inches in the Seven Mile basin. The ground was saturated when the rain began due to the precipitation of the preceding winter. The rainfall was greater than normal during January. Although the precipitation was less than normal during February the meteorological conditions were not such as to dry the ground to any appreciable depth. Some rain fell in the valley on each day of March preceding the storm except the 1st, 9th, and 10th. The rivers began rising on the 11th, fell slightly on the 12th, and then began rising again on the 13th, reaching crest stages on the 13th and 14th. The flood runoff had passed Hamilton by the 20th. Light rainfall occurred on the 16th, 17th, and 18th, but no surface runoff was caused and consequently the amounts have not been included in the table. 200 MIAMI CONSERVANCY DISTRICT June 26-29, 1917 Only a few hundredths of an inch fell on June 26, the first day of the storm. On the 27th, the average precipitation on the different drainage areas varied from 0.48 inches above Locking- ton to 1.08 inches above Wright. On the 28th, the amounts var- ied from 0.63 inches above Germantown to 1.13 inches above Springfield. On the 29th, the rainfall varied from 0.03 inches in the Buck Creek Valley to 0.56 inches above Sidney and Tadmor. The ground was in ordinary June condition, being neither unusually dry nor unusually wet. Although considerable pre- cipitation had occurred during the preceding part of the month it had been utilized by the comparatively high evaporation and transpiration rates which occur during this part of the year. The streams began rising on the 28th, reached their highest stages on the 28th an^ 29th, and then fell slowly. The total sur- face runoff had passed Hamilton by the 5th. Although some precipitation occurred on July 2, no appre- ciable surface runoff resulted. Consequently the amounts have not been included in the table. July 12-17, 1917 The storm of July 12 to 17, 1917, followed close after the preceding described storm. The rainfall on July 12 varied from 0.22 inches above Germantown to 0.88 inches above Sidney. On the 13th the amounts varied from 0.17 inches above Sidney to 0.47 inches above Pleasant Hill. On the 14th, the day of heav- iest precipitation, the amounts varied from 0.65 inches in the Four Mile Creek Valley to 1.72 inches above Sidney. On the 15th the precipitation was comparatively light, the amounts being less than a quarter of an inch except in the upper Miami Valley where they varied from 0.29 inches above Tadmor to 0.39 inches above Lockington. The rainfall was light throughout the valley on the 16th. On the 17th, the last day of the storm, the amounts varied from 0.11 inches above Lockington to 1.17 inches in the Four Mile Creek basin. The ground was fairly wet when the rain began, due to previous rainfall, but was not saturated. The rivers began rising on the 13th, reached their maximum stages on the 14th and 15th, and then fell slowly. The total flood runoiaf had passed Hamilton by the 22nd. Dee. 28, 1917, to Feb. 15, 1918 The flood of February, 1918, was caused almost entirely by melting snow. The winter of 1917 and 1918 was noted' for its RAINFALL AND RUNOFF 201 ^verity. The occurrence of heavy snows and severe cold waves began early in December. The greater part of the precipitation between December 3 and February 11 occurred as snow, the temperature being below freezing the greater part of the time, sometimes several degrees below zero. The ground froze to a depth of a few inches during the cold period from December 6 to 18, but thawed out and became sat- urated during the warmer period from the 19th to the 27th. The snows of the early part of December melted during the latter period, the melting being hastened somewhat by light rainfall occurring on the 23rd, 24th, and 25th. The ground then froze again, due to colder weather, and remained frozen until the thawing period of February. During the period from December 28 to January 3, inclusive, the snowfall throughout the valley was equivalent to about 0.65 inches of rain. This snow was partially melted by a rainfall of about an inch occurring during the period from the 5th to the 8th. Freezing weather began again, however, before much sur- face runoff could occur. From January 11 to February 6 the snowfall was equivalent to about 2 inches of rain. The actual amounts varied from 1.60 inches above Lockington to 2.58 inches above the Four Mile Creek station. From February 7 to 15 the precipitation amounted to about a half an inch throughout the valley. This was in the form of rain and was distributed over several days time. Small rises occurred in January. However, the principal flood runoff began February 9. The streams rose rather ir- regularly reaching their highest stages on the 12th or 13th. Ice jams occurred in many places. The flood runoff had drained out by the 18th. May 11-13, 1918 The most intense precipitation during the storm of May 11 to 13, 1918, fell on the 12th and 13th, less than a half an inch falling on the 11th. The soil was in ordinary May condition when the rain began. It was not saturated although the pre- cipitation during the preceding month had been greater than normal. The streams began rising on the 12th, and reached their highest stages on the 13th. They then fell rather uniformly, reaching normal stages by the 18th or 19th. 202 MIAMI CONSERVANCY DISTRICT August 25-31, 1918 The precipitation during the storm of August 25 to 31 waS heaviest in the Mad River Valley although heavy rainfall oc- curred throughout the Miami River drainage area. Only a few hundredths of an inch fell on the 25th. However, the rest of the precipitation was distributed rather uniformly throughout the remainder of the storm period. The soil was neither unusually dry nor unusually wet when the rain began. The streams in the Mad River Valley began rising on the 30th and reached crest stages on the 30th or 31st. But little, if any, flood runoff occurred in the other portions of the valley. By September 3 the total surface runoff had passed Hamilton. March 14-19, 1919 Although the precipitation during February and the first part of March had been less than normal, the ground was prac- tically saturated when the storm of March 14 to 19, 1919, began. The precipitation on the 14th was comparatively light, amount- ing to only a few hundredths of an inch. The heaviest rainfall of the storm occurred on the 15th. The total precipitation on the 14th and 15th, up to 7 p.m., of the latter date, varied from 1.15 inches above Sidney to 2.47 inches above the Four Mile Creek station. On the following day the rainfall was not quite so intense, the amounts varying from 0.77 inches in the Buck Creek Valley to 1.54 inches in the Twin Creek Valley. On the 17th and 18th the rainfall was nearly uniform throughout the Miami Valley, the total amounting to about 0.75 inches on the 17th, and to about 0.25 inches on the 18th. Only a few hundreds of an inch fell on the 19th. The intensities at Moraine Park on the 15th, 16th, and 17th, are shown in figure 10, page 81. The rivers began rising on the 15th, reached their crest stages on the 16th and 17th, and then fell uniformly, reaching normal stages in all parts of the valley by the 24th. April 15-21, 1920 The storm of April 15 to 21, 1920, caused the highest stage that has occurred at Dayton since the great flood of March, 1913. Only a few hundredths of an inch fell on the 15th. On the 16th the precipitation was considerably heavier. The total amounts for the 15th and 16th varied from 0.83 inches in the RAINFALL AND RUNOFF 203 Seven Mile Valley to 1.92 inches above Sidney. Averages vary- ing from 0.25 to 0.50 inches per day over the valley fell on the 17th, 18th, and 19th. On the 20th and 21st the precipitation was again heavy, the daily amounts varying from about a half an inch to about 2 inches in the different parts of the valley. The rainfall was slightly greater on the 20th than on the 21st except in the Mad River Valley. The ground was nearly saturated when the rain began, due to rains of March and the early part of April. The rivers began rising on the 16th, fell on the 17th, 18th, and 19th, and then rose rapidly on the 20th, reaching crest stages on the 20th and 21st. They then fell rapidly, reaching normal stages throughout the valley by the 28th. TOTAL RETENTION A study of the data in table 34 shows that during similar storms the total retention is generally greater in the Mad River drainage area than in the other portions of the Miami Valley. For instance, during the storm of December 24. 1915, to January 7, 1916, the total retention above Springfield, resulting from a precipitation of 3.65 inches, amounted to 2.70 inches; while in the Stillwater Valley above West Milton the retention, due to a precipitation of 3.99 inches, was only 1.00 inch. In the Miami Valley above Tadmor the retention caused by a precipitation of 3.89 inches amounted to 1.01 inches, or practically the same as in the Stillwater Valley. Other storms show similar condi- tions, although the differences are not always so great. The total retention in the Buck Creek Valley is generally a little greater than in the other parts of the Mad River drainage area. As explained in chapter V, the relatively high retention in the Mad River Valley is due to the comparatively loose and shal- low surface soil, underlain by extensive deposits of gravel. The total retention during the summer storms seems to be much greater than during similar winter storms, as would, of course, be expected, due to the relatively higher rates of evap- oration and soil absorption and the greater amounts of avail- able surface and soil storage during the summer. This is well shown by a comparison of the storms of August 25 to 31, 1918, and April 15 to 21, 1920. These storms were very similar as regards duration and intensity. The total precipitation during the August storm was somewhat smaller than during the April storm, especially in the Stillwater and Upper Miami Valleys. 204 MIAMI CONSERVANCY DISTRICT However, in spite of the smaller rainfall, the retention during the August storm was greater than during the September storm in every catchment area, the differences varying from 0.50 inches above Germantown to 2.61 inches above the Four Mile Creek station in all instances except Buck Creek. In this case the difference was only 0.05 inches. The negative values of total retention obtained for the Seven Mile and Four Mile Creek valleys for the storms of January 10 to 13 and 26 to 31, 1916, are probably in error. This may be due to either or both of two causes. The runoff may be too large due to the difficulties encountered in securing accurate records on a flashy stream of this nature ; or the rainfall may be too small due to the occurrence of heavy showers between rain gaging stations. It is believed that the latter reason accounts for the greater parts of the discrepancies. For similar reasons some of the data for the smaller drainage areas for other storms may be in error. As before mentioned the total retention for a given storm period represents the total quantity of water taken up by the soil plus the total quantity evaporated. Rates of soil absorption and evaporation both vary widely due to variations in meteoro- logical, topographical, and geological conditions. Consequently it does not seem feasible to estimate the value of each component during the various storms given in table 34. Soil absorption and evaporation are both relatively high in the summer and relatively low in the winter. The former is probably greater if the rainfall is steady, than it is if the rain falls in separate intense showers. The latter is greater when the rain falls intermittently, especially if the showers are sep- arated by intervals of warm windy weather. In the case of forested areas appreciable amounts of the precipitation are in- tercepted by the foliage of the trees and are evaporated after the rain ceases without ever having reached the ground. While the areas covered by forests in the Miami Valley are relatively small some data on interception has been secured. During the summer and fall of 1919 two rain gages were main- tained by the writer at his residence. One was located in the open and the other under a hackberry tree, about forty feet high, about midway between the trunk and the outer edge of the foli- age. The data secured is given in full in table 35. It will be noticed that the quantities intercepted vary consid- erably. During slow steady rains from 0.10 to 0.15 inches are RAINFALL AND RUNOFF 205 § . eg 1 u bo-e S S 1 -^1^ Is t ' >?^>}>?>'^ S?^'2" feSS ■" >>^^S3 >>""& '"§3fcS)SS>.S'at>o>.fe OT P3 cQ m m J CO cQ H cQ M M M tJ W CO w 03 03 J en 1-4 1« M M CO w m ra H m m 02 t c c •a III (M COt-m-* O (M O t- t- >0 CO O NCg CO t- t- 1-1 ■* to (M Th (M -* rt rt (;D lO 05 CO T-(OOOOi-HOi-tOT-HOiHi-HC^Cegcoco f IBs OOOOOOOOOOOOOOtHOOOOOOOOOOOOOOOt-IOO 'e Sdc^ac°'^^^SE;Scj3eaSc«caocs^=8^SSSSo,jiSe c » i toTl 50 "cooioooo «s SP.oi '^?!'^'\;?«>'™tDC-toi>ix>?!':^oi B '3 (N ^ t- t-t- t- to tOt- S g ■ s . • ■ g • SS|^^j;c«£SSj.-j:SSgjJdSsHp;p;j;fjj4gdScSSdSc 4 g G n J^ t-" lO «5 4 i p Oi05asOS050S030sa50iOiasasOi0505as050S050^0i0305050SOSOiO>OpOO 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 T-l>-(THlOCOCOtD-*-*'-*>0050aO-*lO'-ll050t-OSrtrtt-OCO(MOJ^ 7? 1 I 1^7^77^^"? 1 1 i 1^^77777777 11^7 1 it t-ti-0000000000000000ai0ia5OOOOOOOOOOOOOOj-JrHl0?D«D?D * 206 MIAMI CONSERVANCY DISTRICT intercepted. If the precipitation occurs as showers, separated by clear, windy intervals, the quantities intercepted are consider- ably higher, due to the evaporation between showers. MAXIMUM VALUES OF RETENTION Table 36 gives the maximum values of retention, in inches, during a few of the larger storms. The quantities were obtained from mass curves of rainfall and runoff similar to those shown in figure 46. They are not as definitely determined as they would have been if data on hourly rainfall had been available. Except in a few instances the rainfall data was limited to daily amounts. Had more exact information been at hand the values of maximum retention would probably have been increased slightly. Table 36. — Maximum Retention, in Inches, During Various Floods Gaging Station Dates of Storms July 7-8 1915 January 10-13 1916 January 26-31 1916 March 21-22 1916 March 25-28 1916 March 14-19 1919 Sidney Lockington. . Piqua Tadmor Pleasant Hill. West Milton. Buck Creek . . Springfield . . . Wright Dayton Franklin Germantown . Seven Mile . . , Four Mile . . . , Hamilton 1.94 04 08 1.20 1.03 1.12 1.30 1.41 1.30 1.58 1.91 1.15 0.71 1.07 1.13 1.17 0.60 1.13 1.31 2.88 1.95 2.02 2.09 2.28 1.50 1.15 1.25 1.28 1.36 1.89 1.42 1.38 1.66 1.93 2.96 2.53 2.16 2.32 1.38 0.97 0.59 1.40 1.51 0.64 0.96 1.87 0.65 0.79 0.92 0.87 0.94 0.96 0.78 1.01 1.51 1.64 80 46 47 36 24 1.34 1.43 2.13 1.63 1.91 2 1 1 2 2 2 2 2 2 2 2. 2, 30 81 59 50 24 29 26 1 33 i 19 55 49 f 45" *No flood runoff. The maximum retention occurred sometime prior to the oc- currence of the crest stages. In the case of the July, 1915, storm they occurred about the time the rain ceased. Since practicailly the entire rainfall fell in about an hour in this instance, the maximum values of retention are nearly as great as the total precipitation. Other conditions being the same the values tend to be lower on the smaller drainage areas due to the compara- tively short time required for the runoff to reach the gaging sections. The maximum value of retention, being the maximum dif- ference between the mass curves of rainfall and runoff, repre- RAINFALL AND RUNOFF 207 sents the accumulated soil absorption plus the accumulated evap- oration plus the actual quantity of water held on the ground or in the streams at the time the maximum retention occurs. The accumulated soil absorption plus the accumulated evaporation cannot be definitely determined except for the entire flood per- •iod. However, it is possible, by a proper consideration of the duration of the storm and of the time of occurrence of the maxi- mum retention, to estimate the proportion of the total absorp- tion and evaporation which has occurred up to the time of the maximum retention. These estimated quantities may then be deducted from the values of maximum retention, thus obtaining estimated values of the maximum quantities stored on the ground surface and in the streams. Table 37 gives the estimated values of accumulated soil ab- sorption and evaporation at the time of the maximum retention, and also the resulting estimates of maximum surface storage, for the storms included in table 36. The volumes of water stored in the channels of the streams may be directly calculated whenever data is available regarding water levels, area of cross sections, and distances. Table 38 gives the maximum volumes stored in certain lengths of the principal streams of the Miami Valley during the floods of July, 1917, March, 1919, and March, 1913, calculated in this manner. The flood of July, 1917, was a comparatively small freshet, not exceeding the channel capacity at any place in the valley. The maximum rates of discharge were only from 5 to 8 per cent of the maximum rates which occurred during the 1913 flood. The flood of March, 1919, was somewhat larger. It caused some overflow in practically all parts of the valley outside the cities. The. maximum rates of discharge in this case were from 10 to 20 per cent of the maximum 1913 rates. The volumes given for the March, 1919, flood do not include the storage on the overflowed lands outside of the main chan- nels. However, the volumes given for the March, 1913, flood do include such storage. In the latter instance the volumes on the overflowed areas were much larger than the volumes within the channels themselves. Reference to table 38 shows that the quantities held in the main channels during the floods of July, 1917, and March, 1919, were relatively small. Since these quantities represent the stor- age in the channels when the crest stages occurred, the quantities in the channels when the maximum surface storage took place 208 MIAMI CONSERVANCY DISTRICT o o o to •a a o > a c »-H05T-ICONC0-*-^t-OS(N|.-H01M 2 9 33- 1— IOOtHOOt-Ht— (T-HOO»-H»— ItHt-H -fi-T lOIMOCTi •t-otot-oo-*ot-cot- MNrt IM(M(M(N -COOO^OOCOCOOOt-00-.* ll ssa oooo •oooooooooo '£3 - IM-^O-* •1Ot-itJ, . OOIOO ■ OCOIO t^co 00-^ ■^ S'T'S lOrtcoia ■ t^OOOO t-lO -silo U3 m H 1^- OOOO ■ooooo oo o 03 cnm ■ OS ^ CO 00O5 OOCOO t- c K 1 1-:: pHt>PQa: 1 c c c 3 ft S3 ^ tl p ^1 1 c o RAINFALL AND RUNOFF 209 must have been still smaller. A study of the data in table 37 and 38 shows that the greater parts of the maximum surface storage during the floods included in table 37 must have been on the ground, in the small pools and depressions, and in the channels of the smaller tributary streams, rather than in the main channels. RATES OF RAINFALL AND RUNOFF Table 39 gives the maximum daily rates of rainfall and runoff in inches over the drainage areas, and the absolute max- imum rate of runoff, in inches per 24 hours, for the greater number of the storms included in table 34. The greatest 24- hour runoff is not the maximum amount calculated for the 24 hours ending at 7 p.m., but is the amount for the 24 hours in which the greatest runoff occurred. It was obtained graphi- Table 38. — Storage in River Channels During Various Floods River Length Drainage Area' Square Miles Storage in Inches Flood of July 1917 March 1919 March 1913 Stillwater Miami Mad Miami Total of 3 lengths Total of 4 lengths Pleasant Hill to Dayton . . Loramie Creek to Dayton. Springfield to Dayton. . . . Dayton to Hamilton above Dayton above Hamilton 674 1162 689 3672 2525 3672 0.16 0.17 0.07 0.14 0.14 0.24 0.33 0.21 0.09 0.22 0.21 0.36 1.48 1.64 1.79 1.19 1.64 2 31 *Total above lower end of length. cally after the hydrographs had been platted. However, the maximum 24-hour rainfall is the maximum for the 24 hours ending at 7 a.m. or 7 p.m., generally the latter, rather than the actual maximum. The available rainfall data, was not sufficient to determine the true 24-hour maximum. The ratio of the greatest daily runoff to the greatest daily rainfall, in per cent, is also included. It will be noticed that no very unusual 24-hour rainfalls have ocurred during the period covered by the table. With the exception of the summer storms of July, 1915, and May, 1916, the quantities seldom amount to as much as two inches. The run- off during the summer storms was relatively less than during the winter storms as already mentioned. Consequently the 210 MIAMI CONSERVANCY DISTRICT B u O 3 o a '3 CD Oi i 1 •3 1 Maxi- mum Rate of Runoff in Inches per 24 Hours dddo '■dddddd a « ad Ratio of Maxi- mum Runoff to Maxi- mum Rainfall in per cent W CO (N id |> 00«3 CO 2 ■ ■ ■ ■ OOOOOO'-' OiCCy-ii-t aj r-H T— 1 Maxi- mum 24-hour Runoff in Inches CO 00 en 05 r^.-i t^r^ d .-idd OCOCOCOCO'+^Oi'^COiO {».... ^OOOO.gO^i-HO Maxi- mum 24-hour Rainfall in Inches 1.08 1.37 1.18 1.14 Ir^t^COOOCO W)00OC<>00 iCCOOOt^-i-H |SCO(M(NO i-HOOOi-H^i-Hi-Hi-HtM CD OS 1-1 2 1 ■3 Maxi- mum Rate of Runoff in Inches per 24 Hours CO "^ "^ CO dddo CO CO '^i^ "^ '^^ ddddo • 00I^TtO3O5 •-idddd rH en ooo ^dd-r^ a> f >> 1 •3 (/I Maxi- mum Rate of Runoff in Inches per 24 Hours 0.32 3-1916 0.58 0.46 05 CD t^.-( ^ CO i-H coco-^ dddad ^H o iocj> CO oor^(N ^r^dd Ratio of Maxi- mum Runoff to Maxi- mum Rainfall in per cent 14. ed 9-1 23. 18. N t~CO*(MCOt~ CO ^ CO (N CO ddddd 8g2S?5 ■-iddd Maxi- mum 24-hour Rainfall in Inches O5io-*aio ooroo oco Oa N cq (Ml Drain- age Area Square Miles to >0 (M QOCO co^igico^g^co^^co § 'Is M i 1 a West Milton. . . Buck Creek. . . . Springfield Wrifi-ht. . i Germantown. . . Seven Mile .... Four Mile Hamilton RAINFALL AND RUNOFF H S M 3 C ■* w lliis'^ilsi •6s §1=1 «3SB. lis R +j •a E g| J S'B £; i2 n! *^ "^ w "^ t^ '-' "^ >-< CD «(>. ,60. oooooooooooo.Sc^o a O0itDTf00t^C0O:03THCC«D CO tH CO CO CO CO (N (N (N CO CO TtH to 1 "ii COiC^-^lOuOfMCOtNTtlCOt^ gl>CO OOOOOOOOOOOOPhOO ThiCOC- O O CD lO tH ^»0 -^ lO <:* (N tN >-< (N OCMi-<000^00000000 COiOCOCNCOCql-^i-ii-HC^tN^i-iT-Hi-i sa^r^ l>-Tti«— lO^Oo'^(NlCCO^-H^^Tt^OTt^ lO (N t^ lO ^ '^ g Th CO ■^ -^ ^ r-H t-H (N O'-tooo'ocoooooodoo ■r! fl 2 " 0> so 3^ C C'g r-i(N(N(Mi-i^(MiM(M.-i.-i^OO^ -J. fl oa « g K B -, O ^ S t^-M g OOOOi-iOOOOOOOOOO X 5 2 2 c^ SB •BS§|J 3 3JJ.S.S u SB43 B oooo.Soooooooooo :o. C . cd ^ IK lC»O(M00CCOCO00(NtOl^d'«**co*ocDOoo-^ Maxi- mum 24-hour Runoff in Inches ocaoiot^cot^i-Hcot^cocotNas i-HOOOOOOOOO'-H'-H.-HO Maxi- mum 24-hour Rainfall in Inches r- oOrHOOOTtiiracQOt^cOi-ii-ieoco „„„„„„ .„„„„^^„ 1 at rH 1 Maxi- mum Rate of Runoff in Inches ^24^ Hours oc -* c OOOO'-HOOOOOi-Hi-H.-iO Ratio of Maxi- mum Runoff to Maxi- mum Rainfall in per cent s iOCDiC00050SOOOtCOOiOtOI>-CO Maxi- mum 24-hour Runoff in Inches McDrfCOTtlCC-^cDONi-HCDOOS ooooooooooo^»- :§ oi ■* 00 .-1 r- r- to eo 00 00 .-H ■ "■* (M N »-l CO (M »-t 1-1 i-t »-H r-H ^ . .^ rH ■ o Maxi- mum 24-hour Runoff in Inches a Maxi- mum 24-hour Rainfall in Inches to It- lO Tt< (M (M tt) CO -* CO cn -tON rtrt .-<,-H.-HrH^rHrtrtO -Ort Oi ■s 2 Maxi- mum Rate of Runoff in Inches per 24 Hours o . . . - o o o . oooooo coooo g a go Ratio of Maxi- mum Runoff to Maxi- mum Rainfall in per cent rtooooireioo'StNiiNaiuj'S'S'S^ Maxi- mum 24-hour Runoff in Inches .-iCOtNtNCCN O^^^i-tT-H o O OVh dooooo^odod'^^^d Maxi- mum 24-hour Rainfall in Inches 1^ SSgggg^S^SSgggg ,-hi-h^^OOi-t s m s o o r/l 9 Ed O o :^ CQ s 9 s a o o a C'-'o 0,0101 as- a»-4 0O "PS -5! IN •^Jss s«° *^3«S O3G0OiOiOSOSI^Q0OiOiOit-l>.I>.OS GC W3 tN- 0:1 00 00 CO 01 05 ooa)0io>0i05- 0000002 OS CO (N CO M Tt< 00 CO 1— I ,— I CO 10 "^ * 10 C. 00 00 C5 Oi CO 00 Oi 01 Oi !>. * C^l 03 cOOii--00OiOlOi0000cD00 Oi O O 00 t^ CO * CO i— t ■^ 00 l>- CO I— I CD t-^ CO t-^ I>- 00 GO t-^ 00 OC 00 lO 10 I>- Oi J, » ^»o I (D ^<^^ iCiO(MQ0C0OC000(N»0^(M00Q0(M iO»O-^(Ni0Oc000*O^OiOh-{M^i-iCD ^ (M{M CO P. & o 214 MIAMI CONSERVANCY DISTRICT values of the 24-hour runoff are not unusually large. On the larger drainage areas they do not materially exceed an inch. On the smaller areas they seldom amount to more than an inch and a half. The ratios of the 24-hour runoff to the 24-hour rainfall, of course, vary widely, owing to the different charac- teristics of the storms as well as to the different soil and surface conditions. The maximum rates of runoff in inches per day are, of course, somewhat larger than the average rates for the 24-hour periods. On the larger areas they sometimes amount to about an inch and a half. On the smaller areas they occasionally amount to from three to four inches. Table 40 gives the ratios, in per cent, of the maximum 24- hour rates of discharge to the absolute maximum rates, for various floods. In order to conform with the arrangement in table 34, the storm dates have been given rather than the flood dates. Data for the flood of February, 1918, which was causea by melting snow, has not been included since the hydrographs were very erratic due to ice jams. The values for a given station vary considerably, as would be expected, due principally to the different durations of the most intense precipitation. For the storms where practically the entire precipitation fell in one day, such as those of July 7 to 8, 1915, and May 6 to 7, 1916, the values are generally smaller than where the heavy rainfall continued through two or more days, as in the storms of January 26 to 31, 1916, and March 14 to 19, 1919. The values are, of course, generally smaller for the smaller drainage areas. CHAPTER VIII.— FLOOD FORECASTING One of the outgrowths of the rainfall and runoff investiga- tions described in the preceding chapters has been the develop- ment of the flood forecasting system now maintained by the Mi- ami Conservancy District. Steps toward the inauguration of this system were taken during the summer of 1916, following the freshets of January, February, March, and May of that year. It was recognized then that such a system would be a necessity during the coming construction period, and that the prelimin- ary steps should be taken at once in order that by the time con- struction began a thoroughly established information service might be available to the contractors as one of the assets of the job. Although the United States Weather Bureau was maintain- ing a flood warning service in the Miami Valley at that time, which it had inaugurated several years before, their forecasts were limited to rises in which flood stages were reached or ex- ceeded. Such a service was, of course, wholly inadequate for the purposes of the District. Inasmuch as it would be necessary to locate the construction equipment to a considerable extent in the valleys along the river channels, or actually within the chan- nels, where sudden rises of only two or three feet might cause considerable damage, it was felt that an independent system should be developed by which the necessary forecasts could be made and issued without the interposition of a third party at critical times. The necessity of adopting such a course became still more evident when it was considered that in several in- stances the construction work would be located at places where only four to six hours elapse between the occurrence of the most intense rainfall and the maximum stages. It should be stated here that the most hearty cooperation has always existed between the officials of the Weather Bureau and of the Conservancy District and that this cooperation has re- sulted in a considerable saving in expense to both parties. More- over, the agreement and accuracy of the forecasts issued from the two sources during critical times has resulted in a feeling of 215 216 MIAMI CONSERVANCY DISTRICT confidence and security on the part of the people throughout the entire valley. THE PRESENT SERVICE At the present time forecasts of crest stages and the times of their occurrence are being made for the five dams and for all of the principal cities along the Miami River below Sidney. In the case of the dams and of the cities where construction work is being carried on in the river channels, forecasts are made for rises of all magnitude, even as small as one or two feet. For the other places forecasts are made only when the rises are great enough to cause some damage or to cause apprehension on the part of the public. However, complete information regarding river conditions is always available at the headquarters office for all places, even though the magnitude of the rise is negli- gible. Warnings are issued to the construction engineers or con- tractors at the various places, or to other interested parties. During critical times bulletins regarding river conditions are issued to the public through the local newspapers or by posting in conspicuous places. Information is also widely distributed by telephone. Operators are kept on duty during the entire night so as to furnish desired information at any time. During one freshet two operators were kept busy answering such inquiries from early morning until midnight. These inquiries come not only from the residents of Dayton but also from residents of practically all parts of the valley, from Sidney on the north to the Ohio River on the south. During long continued storms it is necessary to issue several forecasts. Final estimates of maximum stages to be attained cannot be made until the most intense precipitation has occurred and the weather conditions are such that no additional heavy rainfall is expected. At such times the preliminary forecasts are based on the total rainfall occurring up to the time the forecasts are made. These are issued with the information that the rain is expected to continue and that the forecasted stages will be exceeded by amounts depending on the amount of the subsequent rainfall. For weather forecasts the officials of the District rely entirely o nthe work of the U. S. Weather Bureau. RAINFALL AND RUNOFF 217 REPORTS BEING SECURED In order to maintain the above described service it is, of course, necessary to receive numerous reports from all parts of the valley. Cooperative arrangements were made with the U, S. Weather Bureau in June, 1916, by which special rainfall and river reports from their observers are secured at the head- quarters office of the District during storm periods, as well as at the local office of the Weather Bureau. Special reports at such times are also secured from the greater number of the stations maintained by the Miami Conservancy District. These reports are made by telephone or telegraph whenever 0.70 of an inch of rain falls in 24 hours or less, or whenever there is a sudden rise of three or more feet in the river stage. A confirmation of each report is made by mail as soon as the message has been tele- graphed or telephoned. The gaging stations in the Miami Valley are shown in figure 1, page 17. Reports are received from all of the combined river and rainfall stations, from all but one or two of the rainfall stations, and from nearly all of the river stations located above Hamilton. While it may appear to the reader at first that re- ports from so many stations are not necessary, it must be re- membered that reports from all stations at the proper time can- not be expected. Observers have other work to do besides at- tending to their gages. Although the importance of making the reports should be emphasized' to the observers, and although the members of the observers' families should be trained to take readings and send reports, it frequently happens that for some legitimate reason the report is not made. Such occurrences must be expected in a flood forecasting system and must be allowed for by arranging for more reports than are absolutely neces- sary. Too many reports do no harm while too few result in poor predictions. Although the observers are permitted to send these reports by either telegraph or telephone, they are instructed to send them by telephone whenever it is feasible to do so. The advan- tages of receiving the messages by telephone are, first, that the forecaster can question the observers regarding existing rain- fall, river, or weather conditions, thus obtaining desired infor- mation which they might otherwise neglect to furnish even though instructed previously to do so; and, second, that the observers, in reporting by telephone, are necessarily kept on the job until the message is delivered, thus insuring more prompt 218 MIAMI CONSERVANCY DISTRICT delivery. Quite frequently it has happened that reports handed to telegi'aph operators in small towns have been delayed so long in transmission that they arrived too late to be of use. One factor which has been of great assistance in the fore- casting work of the District and which has made possible the forecasting of small rises of one or two feet in the principal streams, is the direct telephone communication maintained be- tween the headquarters office and the division offices at the five dams. Private wires are maintained to each dam. Consequently it is posible to obtain accurate information from trained ob- servers, directly interested in the work, at any time, without the trouble involved in putting through long distance calls by the usual methods. The river observers are instructed to report the latest gage reading and time of observation; the time the rise began and the stage at that time; whether the river is rising, stationary, or falling; the maximum stage and time of occurrence, if the river is falling; and such general information regarding rain- fall, snowfall, runoff, and weather as they may possess. Rain- fall observers are instructed to report the time of beginning of rain ; the time of reading the gage and the amount of the pre- cipitation up to that time ; whether or not it is still raining ; the time of ending of the rain, if it has ended; the amount and character of the precipitation during the preceding week; and such general information regarding the nature of the rainfall, snowfall, runoff, and weather, as they may have observed. FORECASTING METHODS Owing to the comparatively short intervals of time required for the flood runoff to reach the main streams, flood forecast- ing in the Miami Valley is of necessity largely based on the forecaster's judgment and on his familiarity with previous floods. With reports coming in from all parts of the valley and with people clamoring for information simultaneously, there is no opportunity to sit down and quietly analyze the problem or to apply theoretical formulas. The forecaster keeps in touch with the weather and soil conditions throughout the valley at all times so that when the rain begins he knows approximately how much precipitation will be required to fill the surface storage and what proportion of the remainder may be expected to be taken up by the soil. This in- formation in conjuction with the rainfall reports and his knowl- RAINFALL AND RUNOFF 219 edge of similar previous floods enables him to forecast the maxi- mum stages and the times at which they will be reached at the principal stations in the upper portions of the valley as soon as the most intense precipitation has ceased. Preliminary fore- casts are, of course, made as the reports arrive, as previously noted, but maximum stages can not be forecasted until the most -k I. V6 /4 It 10 8 Q6i _ O-A- Z 4 6 8 10 /Z /4 6age Heights at West Milton in Feet FIG. 47.— RELATION BETWEEN CREST STAGES ON THE STILL- WATER RIVER AT PLEASANT HILL AND WEST MILTON. intense precipitation has occurred. Having estimated the max- imum stages to be attained in the upper portions of the valley, crest stages at the lower stations on the main streams above Dayton are forecasted by the aid of crest relation diagrams, proper consideration being given to channel storage and rain- fall intensities below the upper stations. Figure 47 shows the relation between crest stages at the 220 MIAMI CONSERVANCY DISTRICT Pleasant Hill and West Milton stations on the Stillwater River; filgure 48 shows a similar relation for the Springfield and Wright stations on Mad River; and figure 49 shows similar re- lations for the Sidney, Piqua, and Tadmor stations on the upper Miami River. Points vary from the curves somewhat in all instances, due to local variations in rainfall intensity. This is especially noticeable in the case of the Piqua-Tadmor curve, given in figure 49. Points falling to the left of the curve repre- sent storms in which the most intense precipitation occurred /4 Hv V «i!! /2 •^ ^ fO «;: •1 8 fr •K is 6 u> '♦^ -5: •$^ 4 ^ u ^ 2 vo y ■^ ^ /^ ( / y' u n> / / 8 fO IZ 14 Gage Heights at Wriffht in Feet FIG. 48.— RELATION BETWEEN CREST STAGES ON THE MAD RIVER AT SPRINGFIELD AND WRIGHT. over the northwestern portion of the drainage area ; while points falling to the right represent storms where the heavy rain fell mostly on the eastern portions of the valley. Discrepancies due to this cause are less noticeable on the Sidney-Tadmor curve be- cause heavy rainfall falling east of Piqua drains northward to- ward Sidney as well as southward toward Tadmor, thus pro- ducing similar conditions at both places. It is for this reason RAINFALL AND RUNOFF 221 O Q H Q w Q w E-i O P3 > t5 Oage Heights at Sidney and Piqua in Feet to H Q F-l H O iz; pq CO o < % w H I? O m < H (» H m o % H pq t/2 o H <) &4 224 MIAMI CONSERVANCY DISTRICT ^ "^ ^ ^ <^ «o > ffa^e Neights at Dayfon in Fe& t C^J RAINFALL AND RUNOFF 225 above Dayton. Diagrams for use in such instances are shown in figure 51. It frequently happens, however, that the crest stages at points below the mouth of Twin Creek are caused by the flood runoff from areas below Dayton rather than from those above. Of course, this happens more frequently as the distances below Dajrton increase. During the storm of July, 1915, the crest at Hamilton was caused primarily by the flood runoff from the Seven Mile and Four Mile Creek valleys. In this instance the Hamilton crest occurred about 10 hours before the maximum stage was reached at Dayton. The forecasting methods can probably be made plainer by de- scribing a typical case. The flood of January 5, 1917, will be taken for this purpose. On the morning of January 5 the reports showed that the rainfall had averaged about 1.7 inches over the drainage area of the Stillwater River above West Milton, about 1.6 inches over the drainage area of the Miami River above Piqua, and about 1.4 inches over the drainage area of the Mad River above Springfield. For the Stillwater Valley it was esti- mated that 0.7 of an inch would be required to fill the soil and surface storage; and that the remaining inch, which would run off, would reach West Milton at a maximum rate of about a half an inch in 24 hours, or 7800 second feet. The rating table for the channel at West Milton showed that this would correspond to a stage of 8.0 feet. From the available records of previous floods it was estimated that this stage would be reached at about 3 p.m. By similar methods, making proper allowances for dif- ferences in topographical and geological conditions, it was es- timated that a maximum rate of runoff of about 10,400 second feet, corresponding to a stage of about 7.0, would occur at Piqua at about 3 p.m. ; and that a maximum rate of runoff of about 4,600 second feet, corresponding to a stage of about 7.5, woulvi occur at Springfield at about 2 p.m. From the known distances of these stations above Dayton and from the known slopes and velocities of the streams it was next estimated that the crest from the Stillwater River would reach Dayton first, that the crest from the Mad River would ar- rive next, and that the crest from the Miami River would arrive last. Since the quantity of water flowing in the Miami River ■ was about a third greater than that in the Stillwater River and more than twice as great as that in the Mad River it was esti- mated that the maximum stage at Dayton would be reached when the crest flow from the Miami arrived, and that this would occur 226 MIAMI CONSERVANCY DISTRICT at about 5 a.m. on the following morning. As the rainfall hadi been comparatively general over the entire valley it was esti- mated that the runoff from the drainage area below West Mil- ton, Piqua, and Springfield would be more than sufficient to fill the storage space in the channels below the stages corresponding to the expected discharges. However, since the Stillwater and Mad Rivers would both be falling when the crest from the Mi- ami arrived, it was estimated that the maximum discharge at Dayton would be about 21,000 second feet, or slightly less than the combined maximum discharges at West Milton, Piqua, and Springfield. This quantity corresponded to a stage of 8.9 feet and therefore a stage of 9.0 feet was predicted. From crest relations during previous floods, it was then es- timated that a stage of 10.0 feet would be reached at Miamis- burg at 6 a.m., January 6 ; and that a stage of 9.0 feet would be reached at Franklin at 7 a.m. of the same day. Forecasts were not being made for the dams at that time as construction work had not been started. At Hamilton the crest was caused by Seven Mile and Four Mile Creeks and was nearly reached at the time the forecasts were being made. The following table compares the forecasted and actual con- ditions and also shows the time interval in hours between the time the forecasts were made and the time the crests were reached. Table 41. — Forecasted and Actual Conditions During the Flood of January 5, 1917 Station Forecast at 10 a. m. January 5 Actual Advance Warning in Hours Stage Time Date Stage Time Date West Milton Piqua 8.0 7.0 7.5 9.0 10.0 9.0 3 p.m. 3 p.m. 2 p.m. 5 a.m. 6 a.m. 7 a.m. Jan. 5 Jan. 5 Jan. 5 Jan. 6 Jan. 6 Jan. 6 6.5 5.9 7.3 9.1 9.5 9.2 5 p.m. 5 p.m. 1 p.m. 1 a.m. 7 a.m. 9 a.m. Jan. 5 Jan. 5 Jan. 5 Jan. 6 Jan. 6 Jan. 6 7 7 3 15 21 23 Springfield Dayton Miamisburg Franklin The above described example is probably typical of the fore- casting work in this valley. The predictions are not always so accurate for Dayton and Franklin; neither are they always so inaccurate for West Milton and Piqua. As a general rule, how- ever, the forecasts for the locations on the Miami below the Stillwater and Mad Rivers are more certain than those for the upper stations. APPENDIX BIBLIOGRAPHY The following bibliography contains references to the more valuable articles consulted in the preparation of this report. The literature on rainfall, runoff, evaporation, and related subjects is so voluminous that only the more important can be noted. As a rule references to articles dealing solely with rainfall, or stream flow, or the methods of measurement of either, have not been included. Some of the publications referred to have already been cited in the text. The references are grouped according to the principal subject matter of the articles. No article is referred to more than once although it may contain valuable data relating to more than one subject. The more general works on hydrology and kindred subjects are given in the first group. General Works Hydrology, The Fundamental Basis of Hydraulic Engineering, by Daniel W. Mead. McGraw-Hill Book Company, Inc., New York, 1919. The elements of Hydrology, by Adolph F. Meyer. John Wiley and Sons., Inc., New York, 1917. Hydrology of New York State, by George W. Rafter. Bull. 85, New York State Museum, Albany, New York, 1905. Hydrology of the Panama Canal, by Caleb M. Saville. Trans. Am. Soc. C. E., Vol. 76, page 871, 1913. The Flow of Streams and the Factors that Modify it, with Special Reference to Wisconsin Conditions, by Daniel W. Mead. University of Wisconsin, Madison, Wisconsin, Bull. 425, 1911. Waterworks Handbook, by Flinn, Weston, and Bogert. McGraw-Hill Book Company, Inc., New York, 1916. Public Water Supplies, by Turneaure and Russell. John Wiley and Sons, Inc., New York. American Sewerage Practice, Vol. I, by Metcalf and Eddy. McGraw- Hill Book Company, Inc., New York. The Control of Water, by P. A. M. Parker. D. Van Nostrand Com- pany, New York, 1913. River Discharge, by Hoyt and Grover. John Wiley and Sons, Inc., Now York. Irrigation Pocket Book, by R. B. Buckley. Spon and Chamberlain, New York. 227 228 MIAMI CONSERVANCY DISTRICT The Soil, Its Nature, Relations, and Fundamental Principles of Man- agement, by F. H. King. The Macmillan Company, New York. Physics of Agriculture, by F. H. King. Published by the author, now deceased, at Madison, Wisconsin. Rainfall and Runoff The Relation of Rainfall to Runoff, by George W. Rafter. U. S. Geological Survey, Washington, D. C, W. S. Paper No. 80, 1903. Derivation of Runoff from Rainfall Data, by J. D. Justin, Trans. Am. Soc. C. E., Vol. 77, page 346, 1914. Computing Runoff from Rainfall and other Physical Data, by Adolph F. Meyer, Trans. Am. Soc. C. E., Vol. 79, page 1056, 1915. Relation of Rainfall to Runoff in California, by J. B. Lippincott and S. G. Bennett. Engineering News, June 5, 1902, page 467. Report on Water Supply, Water Power, the Flow of Steams, and Attendant Phenomena, by C. C. Vermeule. Vol. Ill of the Final Report of the State Geologist, Geological Survey of New Jersey, Trenton, New Jersey, 1894. Forests and Water Supply, by C. C. Vermeule. Annual Report of the State Geologist, 1899, page 137. Geological Survey of New Jersey, Tren- ton, New Jersey. California Hydrography, by J. P. Lippincott. U. S. Geological Survey, Washington, D. C, W. S. Paper No. 81, 1903. Variations in Precipitation as Affecting Water Works Engineering, by C. P. Birkinbine. Journal of the American Waterworks Association, Vol. 3, No. 1, March, 1916. Rainfall Causing Flood of Sept., 1899, in the Elbe Basin, Bohemia. Report of Austrian Hydrographic Bureau, 1899.' Gives daily rainfall and runoff data for flood period. Relation of Runoff to Rainfall in Certain Great Floods. Handbuch der Ingenieur Wissenschaften, Teil III, Band I. Interception Rainfall Interception, by Robert E. Horton. Monthly Weather Review, U. S. Department of Agriculture, Washington, D. C, September, 1919, page 603. Contains results of elaborate experiments on precipitation intercepted by different trees and different agricultural crops. Ebermayer's Experiments on Forest Meteorology, translated from Ebermayer's original work and converted into English units by Robert E. Horton. Thirty-second Annual Report of the Michigan Engineering So- ciety, 1911. Gives valuable data on interception, evaporation from water and soil surfaces, and percolation. Evaporation from Water Surfaces Colorado Climatology, by Robert E. Trimble. Agricultural Experiment Station, Colorado Agricultural College, Fort Collins, Colo., Bull. 245, 1918. Gives monthly records of evaporation from a free water surface at Fort Collins for the thirty-one years from 1887 to 1917, inclusive. RAINFALL AND RUNOFF 229 Water Resources of Illinois, by A. H. Horton. Report of Rivers and Lakes Commission, State of Illinois, Springfield, Illinois, 1914. Contains records of monthly evaporation from free water surfaces at several places in the United States. Evaporation from the Surfaces of Water and River-Bed Materials, by R. B. Sleight. Journal of Agricultural Research, U. S. Department of Agriculture, Washington, D. C, July 30, 1917. Describes experiments made at Denver, Colorado. A New Evaporation Formula Developed, by Robert E. Horton. Engi- neering News-Record, April 26, 1917, page 196. Formula includes a new and logical wind correction factor by which the increase in evaporation due to an increase in wind velocity decreases as the wind velocity increases, the effect of the wind becoming negligible at about 20 miles per hour. California Evaporation Records, by Edwin Duryea, Jr. Engineering News, February 29, 1912, page 380. Gives evaporation records secured at Lake Tahoe and at several stations in the Santa Clara Valley. Records of Evaporation Obtained at 23 Different Stations in Various Parts of the United States, Engineering News, June 16, 1910, page 694. Summary of U. S. Weather Bureau evaporation records. Evaporation and Seepage from Irrigation Reservoirs, by Kenneth A. Heron. Engineering News, August 12, 1915, page 294. Gives evaporation records taken near Modesto, California. Evaporation from the Salton Sea, by C. E. Grunsky. Engineering News, August 13, 1908, page 163. An interesting study of the evapora- tion from the Salton Sea, giving actual records. Depth of Evaporation in the United States, Engineering News, January 5, 1889, page 8. Reprinted from an article by Professor T. Russell in the Monthly Weather Review. Discusses the use and accuracy of the Piche evaporimeter. A Study of the Depth of Annual Evaporation from Lake Conchos, Mexico, by Edwin Duryea and H. L. Haehl. Trans. Am. Soc. C. E., Vol. 80, page 1829, 1916. Evaporation Observations in the United States, by H. H. Kimball. Engineering News, April 6, 1905, page 353. Evaporation, by Desmond Fitzgerald. Trans. Am. Soc. C. E., Vol. 15, page 581, 1886. An Annotated Bibliography of Evaporation, by Grace J. Livingston, Monthly Weather Review, U. S. Department of Agriculture, Washington, D. C, 1908-09. Evaporation from Snow and Ice Some Field Experiments on Evaporation from Snow Surfaces, by F. S. Baker. Monthly Weather Review, U. S. Department of Agriculture, Washington, D. C, July, 1917, page 363. Contains valuable data on snow evaporation in the mountains of central Utah. Evaporation from Snow and Errors of Rain Gage when used to Catch Snowfall, by Robert E. Horton. Monthly Weather Review, U. S. Depart- ment of Agriculture, Washington, D .C, February, 1914, page 99. Con- tains data on evaporation from snow surfaces at Albany, New York. Water Resources of the Penobscot River Basin, Maine, by H. K. Barrows, and C. C. Babb, U. S. Geological Survey, Washington, D. C, 230 MIAMI CONSERVANCY DISTRICT W. S. Paper No. 279, 1912. Contains data on evaporation from ice surfaces in Maine. Condensation upon and Evaporation from a Snow Surface, by B. Rolf. Monthly Weather Review, U. S. Department of Agriculture, .Washington, D. C, September, 1915, page 466. See also Science Abstracts, Sec. A, September 25, 1915. Articles give brief description of experiments carried on in Swedish Lapland. An Intensive Study of the Water Resources of a part of Owens Valley, California, by Charles H. Lee. U. S. Geological Survey, Washington, D. C, W. S. Paper 294, 1912. Gives data on evaporation from snow surfaces in the San Bernardino Mountains, also results of experiments on evaporation from soil^and water surfaces in Owens Valley. , The Disappearance of Snow in the High Sierra Nevada of California, by A. J. Henry. Monthly Weather Review, U. S. Department of Agri- culture, Washington, D. C, March, 1916, page 150. Soil Evaporation and Transpiration Factors influencing Evaporation and Transpiration, by John A. Wid- stoe. Utah Agricultural College Experiment Station, Logan, Utah, Bull. 105, 1909. Describes extensive experiments on four different soils. Factors affecting the Evaporation of Moisture from the Soil, by F. S. Harris and J. S. Robinson. Journal of Agricultural Research, U. S. Department of Agriculture, Washington, D. C, December 4, 1916. De- scribes experiments made at the Utah Agricultural Experiment Station and gives an interesting curve, determined experimentally, showing the effeict of wind velocity on the rate of soil evaporation. The curve agrees substantially with the one used by Horton in developing his evaporation formula. Evaporation from Irrigated Soils, by Samuel Fortier and S. H. Beckett. Office of Experiment Stations, U. S. Department of Agriculture, Wash- ington, D. C, Bull. 248, 1912. Gives detailed descriptions and results of experiments carried on in several western states during the years 1908 to 1910, inclusive. Evaporation from Irrigated Soils, by Samuel Fortier. Engineering News, Sept. 5, 1912, page 432. Describes experiments mentioned in above reference. Evaporation Losses in Irrigation and Water Requirements of CropS; by Samuel Fortier, Office of Experiment Stations, U. S. Department of Agri- culture, Washington, D. C, Bull. 177, 1907. Describes experiments on evaporation from soil and water surfaces in California. See also Engi- neering News, Sept. 19, 1907, page 304. Experiments in Evaporation, by C. B. Ridgeway. Wyoming Agricul- tural Experiment Station, Laramie, Wyoming, Bull. 52, 1902. Describes experiments on soil evaporation made at Laramie. Brief abstract in Engineering News, Sept. 11, 1902, page 187. - Daily Transpiration during the Normal Growth Period arid its Cor- relation with the Weather, by Lyman J. Briggs and H. L. Shantz. Journal of Agricultural Research, U. S. Department of Agriculture, Washington, D. C, October 23, 1916. Gives results of extensive experiments carried on with different crops at Akron, Colorado. RAINFALL AND RUNOFF 231 The Water Requirements of Plants, by Lyman J. Briggs and H. L. Shantz. Bureau of Plant Industry, U. S. Department of Agriculture Washington, D. C, Bull. 284, 1913. Gives results of numerous determina- tions of water requirements of plants in pot culture at Akron, Colorado, and Amarillo and Dalhart, Texas. The Water Requirements of Plants, by Lyman J. Briggs and H. L Shantz. Bureau of Plant Industry, U. S. Department of Agriculture, Washington, D. C, Bull. 285, 1913. An interesting review of the liter- ature on the above subject. The Measurement of Soil Evaporation under Arid Conditions, by Charles H. Lee. Engineering News, October 12, 1911, page 428. De- scribes experiments on evaporation from soil and water surfaces in Owens Valley, California. The Determination of Safe Yield of Underground Reservoirs of the Closed Basin Type, by Charles H. Lee. Trans. Am. Soc. C. E., Vol. 78, page 148, 1915. Discusses evaporation; transpiration, percolation, and related subjects, and describes experiments treated in preceding reference. The Determination of the Duty of Water by Analytical Experiment, by W. C. Hammatt. Proc. Am. Soc. C. E., Feb., 1918, page 307. Dis- cusses evaporation, transpiration, soil moisture, percolation, and so forth, and describes experimental work. The Duty of Water in the Pacific Northwest, by J. C. Stevens. Proc. Am. Soc. C. E., March, 1920, page 461. Discusses evaporation, percolation, and surface waste. Method of Estimating the Amount of Evaporation from Water and Soil Surfaces in the Livermore Valley of California. Engineering and Con- tracting, April 30 and May 7, 1913, pages 506 and 523. Soil Moisture Movement and Distribution of Moisture in the Soil, by F. S. Harris and H. W. Turpin. Journal of Agricultural Research, U. S. Department of Agriculture, Washington, D. C, July 16, 1917. Describes field and labora- tory studies made at Logan, Utah. Effect of Temperature on Movement of Water Vapor and Capillary Moisture in Soils, by G. J. Bouyoucos. Journal of Agricultural Research, U. S. Department of Agriculture, Washington, D. C, Oct. 25, 1915. Describes laboratory experiments carried on at Michigan Agricultural Experiment Station. Water Penetration in the Gumbo Soils of the Belle Fourche Reclama- tion Project, by O. R. Matthews. U. S. Department of Agriculture, Wash- ington, D. C, Bull. 447, 1916. Describes experiments on the rate and depth of penetration. The Movement of Water in Irrigated Soils, by H. A. Widstoe and W. W. McLaughlin. Utah Agricultural College Experiment Station, Logan, Utah, Bull. 115, 1912. Describes extensive experiments made on the Greenville farm near Logan. Soil Moisture Studies under Dry Farming, by F. S. Harris and J. W. Jones. Utah Agricultural College Experiment Station, Logan, Utah, Bull. 158, 1917. Gives data on depth of penetration, amount of water stored in surface soil, and reduction of soil moisture by plant growth. 232 MIAMI CONSERVANCY DISTRICT Soil Moisture Studies under Irrigation, by F. S. Harris and A. P. Bracken. Utah Agricultural College Experiment Station, Logan, Utah, Bull. 159, 1917. Scope similar to that of Bull. 158, see preceding reference. Studies on the Movement of Soil Moisture, by Edgar Buckingham. Bureau of Soils, U. S. Department of Agriculture, Washington, D. C, Bull. 38, 1907. Gives experimental data showing differences in soil evaporation under arid and humid conditions, also a theoretical discussion of soil moisture movements. Investigations of Soil Management, by F. H. King. Bureau of Soils, U. S. Department of Agriculture, Washington, D. C, Bull. 26, 19lo5. Gives detailed data on soil moisture variations under eight soils located at Goldsboro, North Carolina, Upper Marlboro, .Maryland, Lancaster, Pennsylvania, and Janesville, Wisconsin. Distribution of Water in the Soil in Furrow Irrigation, by R. H.' Loughridge and Samuel Fortier. Office of Experiment Stations, U. S. Department of Agriculture, Washington, D. C, Bull. 203, 1908. Describes experiments made in citrus orchards in southern California. Natural Vegetation as an Indicator of the Capabilities of Land for Crop Production in the Great Plains Area, by H. L. Shantz. Bureau of Plant Industry, U. S. Department of Agriculture, Washington, D. C, Bull. 201, 1911. Gives experimental data on soil moisture variations at Akron, Colorado. Moisture Content and Physical Condition of Soils, by Frank K. Cameron and Francis E. Gallagher. Bureau of Soils, U. S. Department of Agriculture, Washington, D. C, Bull. 50, 1908. Gives experimental data. The Wilting Coefficient for Different Plants and its Indirect Determ- ination, by Lyman J. Briggs and H. L. Shantz. Bureau of Plant Industry, U. S. Department of Agriculture, Washington, D. C, Bull. 230, 1912. Gives experimental data. Relation of Movement of Water in a Soil to its Hygroscopicity and Initial Moistness, by Frederick J. Alway and Guy R. McDole. Journal of Agricultural Research, U. S. Department of Agriculture, Washington, D. C, August 20, 1917. Gives experimental data taken at the Nebraska Agricultural Experiment Station. Relation of the Water Retaining Capacity of a Soil to its Hygroscopic Coefficient, by Frederick J. Alway and Guy R. McDole. Journal of Agricultural Research, U. S. Department of Agriculture, Washington, D. C, April 9, 1917. Describes experiments carried on at the Nebraska Agricultural Experiment Station. Percolation Stream Flow and Percolation Water, by Samuel Hall. Journal of the Institution of Water Engineers of Great Britain, abstracted in Engi- neering and Contracting, Oct. 29, 1919, page 499. On Evaporation and Percolation, by Charles Greaves. Proc. Inst. C. E. 1875-76, Vol. 45, page 19. Gives results of experiments on percola- tion and evaporation from soil and water surfaces. Percolation and Evaporation, by J. H. Gilbert. Proc. Inst. C. E., Vol. 45, page 56; Vol. 105, page 36. RAINFALL AND RUNOFF 233 Some Physical Properties of Sands and Gravels, by Allen Hazen. Mass. State Board of Health, Boston, Mass., Twenty-fourth Annual Report, page 553, 1892. Gives data on variation in percolation rates caused by temperature changes. Floods Floods of 1913 in the Ohio and Lower Mississippi Valleys, by A. J. Henry, U. S. Weather Bureau, Washington, D. C., Bull. Z, 1913. The Ohio Valley Flood of March-April, 1913,. by A. H. Horton and H. J. Jackson. U. S. Geological Survey, Washington, D. C, W. S. Paper 334. The Miami Valley and the 1913 Flood, by A. E. Morgan. The Miami Conservancy District, Dayton, Ohio, Technical Report, Part I, 1917. Floods in the East Gulf and South Atlantic States, July, 1916, by A. J. Henry. Monthly Weather Review, U. S. Department of Agriculture, Washington, D. C, August, 1916, page 466. Destructive Floods in the United States in 1903, by E. C. Murphy, U. S. Geological Survey, Washington, D. C, W. S. Paper No. 96, 1904. Destructive Floods in the United States in 1904, by E. C. Murphy, U. S. Geological Survey, Washington, D. C, W. S. Paper No. 147, 1905. Destructive Floods in the United States in 1.905, with a discussion of flood discharge and frequency and an index to flood literature, by E. C. Murphy. U. S. Geological Survey, Washington, D. C., W. S. Paper No. 162, 1906. The Rivers and Floods of the Sacramento and San Joaquin Watersheds, by N. R. Taylor. U. S. Weather Bureau, Washington, D. C., Bull. 43, 1913. Southern California Floods of January, 1916, by H. D. McGlashan and F. C. Ebert. U. S. Geological Survey, Washington, D. C, W. S. Paper No. 426, 1917. Floods and Flood Protection. Carnegie Library of Pittsburgh, Month- ly Bulletin, July, 1908. A detailed bibliography. Flood Flows, by W. E. Fuller. Trans. Am. Soc. C. E., Vol. 77, page 564, 1914. The Flood of March, 1907, in the Sacramento and San Joaquin River Basins, California, by W. B. Clapp, E. C. Murphy, and W. F. Martin. Trans. Am. Soc. C. E., Vol. 61, page 281, 1908. A Study of the Southern River Floods of May and June, 1901. Engi- neering News, Aug. 7, 1902, page 102. The Floods of the Mississippi River, by Wm. Starling. Engineering News, April 22, 1897, page 242. The Mississippi Flood of 1897, by Wm. Starling. Engineering News, July 1, 1897, page 2. The Floods in the Spring of 1903 in the Mississippi Watershed, by H. C. Prankenfield. U. S. Weather Bureau, Washington, D. C, Bull. M. 1903. Flood Forecasting River Stage Forecasts for the Arkansas River, Dardanelle to Pine Bluff, Ark., by Herman W. Smith. Monthly Weather Review, U. S. Department of Agriculture, Washington, D. C, March, 1916, page 143. 234 MIAMI CONSERVANCY DISTRICT Freshets in the Savannah River and the Forecasting of High Water at Augusta, Ga., by Eugene D. Emigh. Monthly Weather Review, U. S. Department of Agriculture, Washington, D. C, January, 1914, page 46. Precepts for Forecasting River Stages on the Chattahoochee and Flint Rivers of Georgia, by C. F. Von Herrmann. Monthly Weather Review, U. S. Department of Agriculture, Washington, D. C, July, 1919, page 475. Flood Crests on the Ohio and Mississippi and their movement, by A. J. Henry. Monthly Weather Review, U. S. Department of Agriculture, Washington, D. C, November, 1920, page 651. Treatise on Flood Prediction, and on the Hydrology of the Seine, De Prandeau, 1884. Reviewed in Annales des Ponts et Chaussees, 1884, II. Prediction of Floods in the Central Loire. Annales des Ponts et Chaussees, Oct., 1890. Prediction of High Water on the Elbe in Bohemia, Holtz. Annales des Ponts et Chaussees, April, 1891. Predicting Floods in Rivers. A review of the method employed by the government hydraulic engineer of Queensland in foretelling floods on the Brisbane River. Engineering Record, Sept. 16, 1899. Flood Forcasts. Rev. Tech., Feb. 10, 1899. The Flood Warning Service on the Danube and its Tributaries in Upper Austria. Oest. Wochenschrift, des Offent Baudienst, Jan. 3, 1903. The Prediction of the Height of Water in the Elbe and Moldau Rivers in Bohemia. Oest. Wochenschrift, des Offent Baudienst, Dec. 7, 1901. The Forecasting of Floods in the Yonne at Auxerre after Rains in the Morvan Mountains, P. Breuille, Annales des Ponts et Chaussees, I, 1911.