THE GIFT OF kZ^'=\Si>'\- 2..^j.5gu'.|.!i Cornell University Ubrary arY819 v-1-2 . ...ho, tunes and storms bVP Studies of .weather^, ty^if, , « |be antici- pated. Chart 10 of the morning of the 14th shows the 12 and 24 hour movement of the center. It will be observed that the rapidity of the storm's movement, which was at the rate of about 54 miles per hour, did not permit a warming up of the air in the east quadrants of the low area, and upon its arrival in the southwest the isotherms still ran southeastward across its line of advance, with freezing temperature as far south as northern Florida. Although a sudden and marked rise in temperature might be expected to precede, and very low temperature to follow, the passage of the storm over the east-central districts, the ^ colder air which appeared in its front the morning of the " 14th rather favored a loss of strength. The report of the fol- lowing morning. Chart 11, shows that the low area flattened or divided against the cold area, and that the separate low areas reunited off the middle Atlantic coast. This storm is an ex- cellent example of the influence of low temperature upon a storm's movement. It was opposed by much lower teinpera- ture in its front until it reached the southwest, and in endeav- oring to skirt the cold area, nearly perished for lack of warmth, which is one of the sustaining elements of a storm. It fol- lowed the path of least barometric resistanceV and, encounter- ing but slight opposition in that respect, traveled at a high rate of speed. Owing to its great velocity, low temperature, which is unfavorable to precipitation, preceded and attended its passage, and no precipitation occurred save in the eastern districts where sharp temperature changes and gradients in a moist atmosphere produced a heavy fall of snow throughout the Atlantic coast States north of Florida. A typical storm of the Saskatchewan Valley type appeared over Alberta, January 20, 1892, and reached the Canadian Maritime Provinces January 23, traveling over the most fre- quented track of storms of this class at an average velocity of about 39 miles per hour. The daily progress of this storm and the conditions which attended its passage are shown on Charts 12 to, 15. The morning reports of the 20th, Chart 12, present conditions which prevailed just before the full devel- opment of this storm within the region of observation. At that time a storm occupied the Lake Superior region, and the trend of the isobars and isotherms favored a rapid eastward movement of the Alberta storm. By the morning of the 21st, Chart 13, the Lake Superior storm had moved eastward to a position north of Lake Huron, and the northwest low area had moved eastward over the Saskatchewan Valley a corre- sponding distance. Between the low areas the pressure had risen. The appearance of a crest of high pressure between low areas moving over the northern districts can, as a rule, be expected, and the lower temperature and higher pressure can, as in this case, be relied upon to rapidly give way and follow in the wake of the eastern low, allowing the western low area to advance at a normal velocity. In the present in- stance the high area disappeared during the 31st, and the northwest storm moved eastward over Manitoba, and by the morning of the 22d, Chart 14, had reached Lake Superior. Its after course to the Maritime Provinces was unobstructed, and its passage was unattended with noteworthy features. Storms of this type, which follow the most frequented path over the northern Lake region, are seldom attended by pre- cipitation until they reach the Lake region, and during their passage thence eastward the rain area is usually confined to the Great Lakes, New York, and New England. Neither are they, as a rule, attended by cold waves, save in the case of the slower moving storms of marked intensity which produce high temperatures in the east quadrants and are folk)wed by a strong sweep of northerly winds. The storms which pursue a more southern course are, however, often attended by areas of precipitation which extend to the Gulf States and by cold waves which rgach the southern limit of the barometric trough. NORTH PACIFIC COAST STOKMS. These storms present characteristics similar to those noted in connection with the Saskatchewan Valley type. A majority of the storms of both types doubtless spring from the extteme eastern limit of the permanent winter low area of the North Pacific Ocean. The North Pacific type of storms as herein classified, however, strikes the American coast farther south, and a reference to Chart 3 will show that their tracks are more widely distributed and, on the whole, run farther south than those of the Saskatchewan Valley type. Of the 20 storms of this class traced for the last ten years, 10 reached the Atlantic coast, all, save one, passing to sea north of the fortieth parallel. In this connection it is interesting to note that this result compares closely with figures found in Bulletin A of the Weather Bureau. During the ten years covered by that report an average of 1.8 storm per month appeared on the north Pacific coast and traversed the North American continent in January. The tables found in the Bulletin show that the storms that appear on the north Pacific coast of the United States in winter possess greater vitality than any other class of storms traced over the Northern Hemisphere, and that in a 10-year period 18 storms from that region traversed successively the North American conti- nent and the North Atlantic Ocean during the three winter months. These storms usually cross the continent in about three days, at an average velocity of 86 to 40 miles per hour. After passing east of the Rocky Mountains they assume the charac- teristics noted in connection with storms of the Saskatchewan Valley type. The storms that pass well to the southward of the forty-fifth parallel carry precipitation to the Gulf States, and the cold-wave areas, depending upon the storm's intensity and previously existing temperatures, usually cover districts included within the low barometer troughs which are swept by the northerly winds of the storm's west quadrants. NOBTHEAST HOOKY MOUNTAIN SLOPE STOKMS. The northeast Rocky Mountain slope type of storms. Chart 4, also belongs to this general class, but, owing either' to un- usually rapid movements, which carry them across the moun- tains between reports, or to slight intensity, they do not ap- pear as fully developed storms until they reach the eastern slope. SOUTHWEST STOKMS. Probably the most important winter storms of the eastern . half of the United States are those which first appear in the Southwestern States. About one-half of the storms that reach the Atlantic coast belong to this and kindred southern types, and the rain and temperature change areas are more extended, ,general, and pronounced than in any other class of storms that traverse districts lying east of the one-hundredth meridian. Storms of this type almost invariably move north- eastward, and, unless the conditions are complicated by north- west low areas, they reach the Atlantic coast within forty-eighl hours. Storms of average intensity which appear over the" interior of the west Gulf States or on the extreme southern Rocky Mountain slope usually cross the Ohio Valley and the Lake region, attended by general rain or snow over the eastern half of the country, and those which advance from the im- mediate Gulf coast or from over the west part of the Gulf of Mexico move almost due northeast, producing areas of general rain over the Southern and Atlantic States, the upper and middle Ohio Valley, and the eastern Lake region, and are often attended by cold waves and dangerous gales in those districts. . During the last ten years 21 storms have appeared in the southwest in January, and in all but Oae instance they reached the Atlantic. About one-half of this number crossed the Lake region and passed thghce over or north of the St. Lawrence Valley; the others traversed the Atlantic coast States. The general features and characteristics of these storms can best be shown by discussing subtypes of the gen- eral type referred to. A storm of the class that first appears on the extreme south- ern Rocky Mountain slope advanced from Texas to southera Lake Michigan from the morning of January 8 to the morn- ing of January 9; 1889, crossed Lake Huron and reached a position far to the northeast of Georgian Bay by the morning of the 10th, and passed eastward over Labrador during the nth. This storm was faintly outlined over the southern Rocky Mountain region on the 7th, and by the morning of the 8th had reached the position shown on Chart 16. By the evening of the 7th rain had set in over the southwest, and by the time this storm had become well marked over Texas the rain area had extended over the Mississippi and lower Ohio valleys. A study of Chart 16 shows that the pressure and tem- perature distribution favors an almost due northeast movement of the center of disturbance. The path of least barometric pressure resistance lies in that direction, and the northward loop of the isotherms over the Mississippi and lower Ohio valleys shows the region of increasing temperature toward which the storm is likely to move with the favorable pressure conditions presented. As is usual in storms of this class, the rain area which covers the Mississippi Valleycan be ex- pected to reach the Atlantic coast within twemty-four hours. It has been observed, however, that when a storm is central! m iexas about thirty hours are required for the rain area to overspread New England. The high area over the northern plateau region will move rapidly southeastward and replace- the low area in the southwest, causing a decided fall in tem- perature withm the cyclonic area which appears on this chart ; m fact, a cold wave can safely be anticipated within the storm area say within the area covered or surrounded by the isobar_ of 29.70. Chart 17 shows the movement of the storm during the succeeding twenty-four hours. A notable teature IS the rapid deepening of the barometric depression, lie other conditions are as outlined in remarks relating to Chart 16. The high area has moved rapidly southeastward, the temperature is some 20° lower in the southwest, and the rain area has extended to the Atlantic coast south of New England, and will extend over that district during the next tew hours. As the storm center will now move well to the northward of the St. Lawrence Valley, and the sweep of the northerly winds will not be strong, save over extreme northern districts, temperature falls sufficient to constitute a cold wave can not be expected within the next twenty-four hours, and rapidly rising barometer in the south and southwest point to a rapid clearing of the weather in the central and southern districts. The report of the following morning, Chart 18, shows the storm well beyond our region of observation, and its further influence will be in the form of diminishing west- erly gales along the middle Atlantic and New England coasts and over the eastern Lake region. The second class of this type of storms first appears near the immediate west Gulf coast, and moves northeastward, par- allel with, and generally somewhat to the westward of the Appalachian range, and crosses eastern New York and New England, the average time of transit being about forty-eight hours. A good example of a storm of this class appeared on the Louisiana coast the morning of January 26, 1889, and the conditions which obtained at that time are exhibited by Chart 19. As rain had been falling over the southern and southeastern districts during the preceding two days, the actual rain area controlled or occasioned by this storm can not well be determined. A study of these storms has shown, however, that when fair weather prevails over the central and eastern districts at the time of the first appearance of the low area on the Gulf coast, the rain area spreads rapidly from the Gulf and covers districts east of the Mississippi and south of the Great Lakes within twenty-four hours and extends over New England within thirty-six hours. When this storm appeared on the Louisiana coast the morn- ing of the 26th, a trough of low pressure extended thence over the eastern Lake region and the St. Lawrence Valley, and the isotherms looped northward over the Ohio Valley and the lower Lakes. These conditions plainly indicate the direc- tion of the storm's advance, and, as neither pressure gradients nor low temperature oppose the advance of the center, it may be assumed that the movement will be rapid. The rain area, which Covers districts south of the Ohio River, may be ex- pected to cover New England within twenty-four hours, and the barometer will rise^ rapidly with a decided fall in temper- ature in the Southwestern States, the fall constituting, a cold wave in the west Gulf States: Chart 20 shows the progress made by the center of -disturbance during the succeeding twenty-four hours. The storm has increased in intensity and rain has fallen throughout the central valleys and in the Atlantic coast States. The conditions presented the morning of the 27th seem, at first glance, unfavorable to a normal northeast advance of the storm center. It will be observed that a well-marked area of high pressure has appeared over the Canadian Maritime Provinces, with a decided fall in tem- perature in that region. Under certain conditions high areas of this class retard and even force back storms advancing from the interior. When this occurs the northeast high area is usually, one of gradual growth and apparently well an- chored, and the barometer is relatively low in the north-cen- 'tral districts with a gentle barometric gradient and high temperature for the season in that direction. W^^^i?' ^? i° the present instance, the western and northwestern districts are covered by an unbroken high area of great magnitude, with much lower temperature to the west and northwest of the low area, the center of disturbance can scarcely recurve in that direction, and may be expected to increase in inten- sity, warm up the cold area in the northeast by a strong in- draught of ocean air, and force a passage along the usual path. So far as January storms are concerned the tracks plotted on Chart 2 show that during the last ten years the southwest type of storms have, without an exception, con- tinued a northeast course after leaving the Gulf States. The morning report of the 28th, Chart 21, shows that this storm increased greatly in strength and advanced in a direct line toward the Canadian Maritime Provinces, either forcing eastward or dissipating the high area which occupied that region the preceding morning. In the meantime the center of the high area in the west remained nearly stationary and the pressure in the Northwest and over the western Lake re- gion decreased instead of increased, as might have been ex- pected, and increased but slightly in the southwest. The effect of these minus pressure changes in the west and north- west was to delay the clearing of the weather which generally closely follows the passage of these storms, and the cloud and rain area lingered over the Ohio and middle Mississippi valleys, and a decided fall in temperature occurred only in the cleared region which covered the interior of the South- western States. A representative storm of the third class of low areas of this type (that is, those storms which appear farthest south and cross the west part of the Gulf of Mexico and pass thence northeastward over the east Gulf, south and middle Atlantic States) first appeared near the mouth of the Rio Grande River the morning of January 23, 1891, advanced to central Alabama by the morning of the 24th, moved thence north- eastward off the nliddle Atlantic coast during the early morn- ing of the 25th, and disappeared in the direction of Nova Scotia during the latter-named date, traversing the territory lying between the mouth of the Rio Grande and a point off the southeast New England coast in forty-eight hours. Chart 22 shows the first indications of the presence of this storm off the mouth of the Rio Grande River. The exact location of the center can only be surmised, but a calculation based upon the well known persistency and uniformly rapid move- ment of this class of storms, and the absence of a pressure gradient to the northeastward, would give the storm a move- ment to the middle Gulf States within twenty-four hours, and as precipitation is one of the first well-marked features of these disturbances, rain could be expected over a large area of the Southern States. Within twenty-four hours the center of this storm had advanced to Alabama, and Chart 23, of the morning of the 24th, presents no material obstacle calculated to prevent the storm from continuing a northeast course. The area of high j)ressure on the middle Atlantic coast has shifted to that position from the south Atlantic coast and is moving; it will not, therefore, oppose the advance of the storm, more especially as the temperature is high and the direction of the isotherms is northeast from the storm center. As the barometer has risen rapidly in the rear of the storm, forming a high area in the southwest, a rapid rise of pressure and rapidly clearing weather will follow closely in the wake of the storm. Like storms that traverse the Atlantic coast States from the Gulf, this low area was unattended by marked changes in temperature in the Atlantic coast districts. The storms that appear over the east Gulf and the east Gulf States in January generally belong to the type herein considered, and do not, therefore, call for individual mention. GENERAL KEMAEKS. From the foregoing charts and remarks it would appear that the storms of January belong to three, and possibly to but two, general types which may be subdivided into a limited number of classes. We have seen that fully one-half of our January storms advance from the Saskatchewan Valley and the north Pacific coast, and that of these types the storms of the first-named type are the most numerous. Many, if not all, of the Saskatchewan type are of Pacific coast origin, and the two types can, therefore, be properly combined and termed the north Pacific type, the difference being merely that the storms traced on Chart 10 reach the coast farther south than those of the Saskatchewan Valley type. A large proportion of these storms doubtless originate near the American coast, and do not advance from the Bering Sea permanent winter low area. The plotted tracks of storm in Weather Bulletin A show that at least four - fifths of the storms that appeared on the Pacific coast north of the mouth of the Columbia River during a period of ten years first ap- peared near the coast, and did not actually travel eastward from the north Pacific or Bering Sea low area. The Bering Sea low area loops far to the eastward and reaches the Alaska coast in the neighborhood of Sitka in January, and this cir- cumstance, taken in connection with the fact that the cold Arctic current flowing southward through Bering Sea Straits and the warm Pacific drift current meet south of the Alaska Peninsula, presents conditions which doubtless ' largely con- tribute to the development of storms off the Alaska coast south of the Alaska Peninsula. The three branches representing the average paths of the north Pacific type of storms are' shown on Chart 25. The north, or Saskatchewan branch, and the north Pacific branch converge and meet in the St. Lawrence Valley, and the north- east Rocky Mountain branch swings slightly to the southward of the north Pacific branch over the Northwestern States, and crosses and passes to the northward of the Saskatchewan Val- ley branch northeast of Georgian Bay. A result of the more southern path of the north Pacific and northeast Rocky Mountain slope storms is to carry precipitation, and marked temperature change areas farther south, and these storms are more liable to be attended by secondary developments still farther to the southward, thereby causing general rain or snow over a great extent of country. The second principal type, which embraces storms that first appear in the southwest, is also divided into three branches, all of which run almost due northeast. Storms of this class doubtless develop in the lee of the southern Rocky Mountains in the United States and to the eastward of the mountain ranges of Mexico, and an important element of their origin is found in the meeting over those regions of the warm, moist, easterly winds, which blow off the Gulf of Mexico, and which are really the western edge of the North Atlantic trade winds and the cold, dry, northwest to north winds which sweep southeastward and southward along the eastern Rocky Mountain slope. As before stated, the storms of this principal type are the most important that traverse the eastern half of the United States in January. They are attended by widespread and abundant precipitation and de- cided temperature changes, and are the most methodical storms as regards their direction and velocity of movement that appear within the region of observation. COLD WAVES. A discussion of winter storms and weather would be incom- plete without a reference to cold waves. The conditions pro- ducing and attending these phenomena are so complicated, however, that even a general discussion of the subject, calcu- lated to prove instructive to forecasters, is attended by marked difficulties. The resolving into types of the innumerable com- binations presented in connection with the development and appearance of cold waves is an extremely difficult if not an impossible task, and the scope of this paper will admit of only a general discussion of their more prominent characteristics, and of a few remarks touching upon recognized conditions favorable to their entry into and progress over the United States. The visible mechanism of a cold wave embraces the cyclonic and anticyclonic areas which traverse the United States from west to east. The low areas warm up the surface air by the southerly winds in their east quadrants, and the cold, dry, northerly winds in their west quadrants that usher in the succeeding high area from the British Northwest Ter- ritory, occasion a marked fall in temperature which is termed a cold wave. It is evident, therefore, that generally speaking, the region covered by a cold wave must be successively sub- jected to the wind circulation of the east and west quadrants of a well-marked low area. It is also evident that the cold waves of the several sections are practically dependent upon the passage of low areas followed closely by unbroken high areas. Thus far the mechanism seems simple and easily under- stood, and if the movement and strength of the high and low areas could be accurately foreseen, the forecasting of cold waves would be one of the simplest instead of one of the most difficult features of weather forecasting. As a matter of fact each of the many districts of the United States pre- sents geographical and topographical features calculated to modify or intensify approaching cold waves. The probable intensity of a cold wave must be calculated for the varying conditions peculiar to each of the districts, and in many in- stances for conditions peculiar to localities. Unlike warm waves, which often produce in the central and northern dis- tricts temperatures higher than those noted in more southern latitudes to the windward, cold waves are not attended in the central and southern districts by temperatures lower than those noted to the north and west. On the contrary, the cold waves diminish in intensity as they sweep south and east, so far as the degree of actual cold is concerned, although the temperature may be relatively lower with reference to the normal temperature. As cold waves approach the moist re- gions of the Great Lakes and the Gulf and Atlantic coasts, conditions must be very marked to insure their overspreading those districts. For, as cold waves follow general storms, and as areas of precipitation, and even of cloudiness, are gener- ally fatal to the advance of a cold wave, the forecaster should be very certain that the weather will clear up in a district before ordering cold-wave signals for that district. Herein lies the difiiculty of verifying cold-wave signals in the coast and Gulf regions ; for the weather is often slow to clear up in the Gulf and Atlantic coast States, and in ad- dition, there sometimes appears to be a slight foehn effect produced in districts to the leeward of the Appalachian range of mountains ; this, however, has not been proven. In the Southern States cold waves can rarely be successfully fore- casted unless a well-defined low area crosses that region, fol- lowed by a well-marked and unbroken high area which has occasioned a decidedly cold wave in districts to the west or northwest. Twenty-four to thirty-six hours are usually re- quired for a cold wave to advance from Texas to the south Atlantic coast. In January the cold waves of the central and northern districts attend the passage of the general type of storms that pass eastward from the north Pacific coast and the Saskatchewan Valley. As these cold waves drop down from the British Northwest Territory in the rear of and im- mediately follow the storms of this type, they assume cor- responding velocities. The average time for a cold wave to advance from the British Northwest Territory to the middle Atlantic and New England States, would, therefore, be sixty to seventy-two hours. But as the storms vary in velocity, so would the time required for a cold wave to sweep the north- ern regions vary. In all cases the velocity of cold waves must be governed by the velocity of the low areas and of the succeeding high areas, and their intensity upon the observed temperature distribution, and the intensity of the low and high areas which promote, sustain, and propel them. 'a f |1| ii'H|imlln I "■es ..as - s '^ ■ M iH ^ ?1 CD 13 O 01 . f o UJ s H &3 S 2 'S3 "C ^ " - ■<■ < 1-— 1 u- .5^ (^ u a Q U- 5 j= . >»■ -'- ~ £•:■ :op \ § L '"iis^r^s^'-^-:." a GO .SO lliiill MIIIIW llilll Z h 1 Z. !l oJ 8T8I moil a9nol3T[3 to i'jBiJ 9a«9"= Kvode anil ioeW V^yH naa lo anobfiDoI ban esJfib wode esiusii sdT High Areas of the North Pacific Coast Region and their Associated Phenomena. Chart VI. September. Chart vm. November. ^3^.^v CHART L MIDDLE ATLANTIC AND NEW ENQtLAND STATES, JANUARY, OH.A.RT 2. January 6, 1886—7 a. m. -^°.. 'Ju.^r-n- ^S^^S^SjW^ -^-r' OHABT 3. fr~ — : ■" January 7, 1886—7 a. mr i%|0" ..,*: f Severe norther, with heavy snow, in 1; Nebraska, Kansas, and Indian Ter- 1 ritory ou the 7th. OHABT 4. January 8, 1386—7 a. m. Heavy snow on the 8th in the upper Mississippi and Ohio valleys, and severe norther and intense cold wav! in Texas; also snow in the eveninK in the Gulf States and At- lantic Coast States. CHART 6. January 9, 1836—7 a. m. Severe cold wave continued over the ^ /, Soathern States and heavy snow A' .xX) and gales over the middle Atlantic ' rfOJ' and New England States on the -^ 9th. OHABT a. January 10, 1886—7 a. m. lik-Severe cold contlnned In the South- ern States. In Florida damage to fruit, fish, and oysters, $1,000,000 to $2,000,000. ^03 CHART 7. January 19, 1889—8 a., m. CHART 8.. January 20, 1889-^=8 a. m. OTTART 9. January 21, 3,889^8 a. m. OHABT 10. January 23, 1891--8 a. ra. OHAET n. January 24, 1891—8 a. vd. Heavy rain extended over the mid- dle Atlantic States on the 24th and changed to snow at night. — ^,/lf, 1_ OHABT 12. January 25, 1891—8 a. m. Heavy snow m New England bn the 25th, and heavy snow and clearing 111 middle Atlantic States. ,56'' /I-/*'- .g^- Chart is. January 6, 1892—8 a. m. Ealn area extended over the south Atlantic States on the 5tb. CHART 14. January 6, 1892—8 Et. m. Eeary snow in the middle Atlantic and New England States on the 6th. , so: -N OH ART 15. January 4, 189S~8 a. m. CHART 16, January 5, 1893—8 a. to. CHART 17 January 28, 1894—8 p. m. Rain began along the Gnlf coast and snow in the npper Mississippi Val- ley and uppOT Laie Region on the OHABT 18. January 29, 1894—8 a. m. CHART 10. January 29, 1894—8 p. m. ^■^Jl 30.2 Heavy snow over the Middle Atlaa- tic and New England States on ■"■ - 39tli. CHART 20. SOUTHEASTERN STATE^. CHART 21. January 17, 1893—8 a. m. :^.ll# ^^^ H^f^T ;3v|'-?'>"'^y^.'' ■'•rM"'^^ i Heavy rain extended over the west Gulf States, and at night snow tell as far south as Mobile on the 17th. CHART 22. January 18.. 1893 — 8 a. rii. Heavy snow In Sooth Atlantic and interior of east Gulf States on the 18th. CHART 23. January 18, 1893—8 a. xa. -^M. Heavy snow In the South Atlantio States on the 19th. , ' ^ ^.^fvWEJT CHART 2 en tOi \v\)(^^ OHABT 28, UPPER MESSISSIPPI VALLEY'. ,/s« December 16, I860— B &■ m. Heavy snow on the irth in the Mid- dle Atlantic States, and in New England on the 18th. CHART 69. December 25, 1800—8 a. m. CHART 70. December 26, 1892—8 a. m. Heavy snow In the Carolinas and eastern Virginia on the 27th. ^^^Mf:^'T^- CHART 71. NORTHWESTERN STATES. DECEMBER. CHART 72. December. 2, 1891—8 a. m.