TN295 No. 8964 ^■iy' . , L ' 9 w , C • . » o ^,v cy * ^> :iP .^J^^*. ^^ 0"° ^^ -^v-^^' ■"^ii ^-* ay t .^^ ^ .0 ..s • • /^ "^ * » 0^ ,o«o, -^Q a\ ^'% >1 ♦'i'rfZ^' "t- -*''^°*^ o- %° f ♦ *■. o V K 1 ° *? . *i/^% V ^* -^^ ^ V ^! .^^'V •v ^ iOvv * ^^' o""^ -^O .^^°^ V^o^ ^oV^ '^d' .^^°^ v<,./%^^^;/...;v^'> v^-^oo v"^*y v^^v <^ '^oV^ * -J 4> ^0 -^-K. ^* o«o o, • t*, a'^ »'jA^>(f/l,'o ''-^^ .^^ ^ IC 8964 Bureau of Mines Information Circular/1984 Ground Penetrating Radar A Review of Its Application in the Mining Industry By W. E. Pittman, Jr., R. H. Church, W. E. Webb, and J. T. McLendon UNITED STATES DEPARTMENT OF THE INTERIOR ((iPnMMiU. ^AJUtMA^fh^i^ Information Circular/8964 Ground Penetrating Radar A Review of Its Application in the Mining Industry By W. E. Pittman, Jr., R. H. Church, W. E. Webb, and J. T. McLendon UNITED STATES DEPARTMENT OF THE INTERIOR William P. Clark, Secretary BUREAU OF MINES Robert C. Horton, Director Research at the Tuscaloosa Research Center is carried out under a memorandum of agreement between the Bureau of Mines, U.S. Department of the Interior, and the University of Alabama. Library of Congress Cataloging in Publication Data: T ^0 Ground penetrating radar. (Information circular / United States Department of the Interior, Bu- reau of Mines ; 8964) Bibliography: p. 16-23. Supt. of Docs, no.: I 28.27:8964. 1. Mine safety— Equipment and supplies. 2. Ground penetrating ra- dar. I. Pittman, Walter E. II, Series: Information circular (United States. Bureau of Mines) ; 8964. TN295.U4 622s [622'.8] 83-600319 ^6° CONTENTS \ji Page (V ^ Abstract 1 \ Introduction 1 ~- Origins of GFR concepts 2 "" Electromagnetic transmission through the earth 3 Electrical characteristics of earth and rock « 4 Electronic characteristics of coal 5 Applications of GFR 6 Radar measurements of Ice and snow thickness 6 Radar measurements of salt domes 7 Ar cheologlcal and geotechnlcal applications 7 Subhlghway measurements 8 Microwave measurements of coal seams 8 GPR systems 9 Video pulse radar applied to GPR 9 Short pulse GPR 10 Borehole radar 10 Synthetic pulse radar 11 Strata control radar 12 GPR antennas 14 Interpretation of GPR data 14 Conclus Ions 15 Bibliography 16 ILLUSTRATIONS 1 . Block diagram of a simple short pulse radar system 2 2 . Block diagram of a synthetic pulse radar system 3 3. Radar control console of GPR used during underground testing 13 4. Antenna deployment against mine roof 14 to o -2 UNIT OF MEASURE ABBREVIATIONS USED IN THIS REPORT A ampere lb/in2 pound per square inch cm centimeter m meter dB decibel MHz megahertz dB/ft decibel per foot mi/h mile per hour op degree Fahrenheit mm millimeter ft feet m/s meter per second GHz gigahertz MW megawatt Hz hertz ns nanosecond in inch pet percent kHz kilohertz V volt km kilometer W watt kW kilowatt GROUND PENETRATING RADAR A Review of Its Application in the Mining Industry By W. E. Pittman, Jr./ R. H. Church, ^ W. E. Webb,^ and J. T. McLendon ABSTRACT The Bureau of Mines, as part of its Health and Safety Technology Pro- gram, is conducting research on the use of ground penetrating radar (GPR) for mine hazard detection. GPR offers a possibility of mapping immediate below-surf ace geologic conditions, including faulting and other anomalies, thereby exposing potential safety hazards to miners. This report summarizes a literature review of the current status of GPR research in and outside the Bureau. INTRODUCTION The presence of unsuspected geologic anomalies in coal-bearing strata poses a threat to the safety of coal miners during mining. Clay veins can create structural weaknesses and cause the collapse of mine open- ings. Uncharted well casings, sand channels, sulfur balls, "horse- backs," "kettlebottoms ," and mine voids are known to be hazardous features. While coal mining will remain a dangerous industry, it is obvious that advance knowledge of what lies around the mine opening and ahead of the coal face could save many lives and also improve the effi- ciency of coal production. Many techniques have been proposed for premining investigation and some have proved valuable. GPR is one of those holding promise of becoming an important method of detecting near-face hazards. As part of its continuing effort to advance mine safety technology, the Bureau of Mines has been conducting research on various types of GPR's, for use from the surface, from the underground working face, or from a borehole. This report reviews the historical progress of GPR's and the current status of research to apply them in the mining industry, ^Technical information specialist, Tuscaloosa Research Center, Bureau of Mines, Tuscaloosa, Ala.; professor of the history of science, Mississippi University for Women, Columbus, Miss. ^Mining engineer, Tuscaloosa Research Center. -^Physical scientist, Tuscaloosa Research Center; professor of electrical engineer- ing. University of Alabama, Tuscaloosa, Ala, ORIGINS OF GPR CONCEPTS GPR probably owes most of its begin- nings to military research. In particu- lar, the U.S. Army has maintained a long term interest in GPR as a rapid means of detecting land mines and subsurface tun- nels. In 1956, Cook separately proposed and demonstrated the use of an airborne radar to measure the thickness of float- ing sea ice (33-34) .^ He recognized that there were two interfaces (air-ice, ice- seawater) , each of which would give clear radar returns. The thickness of the ice could be determined from the different times of arrival at the receiver of the two echoes and knowledge of the electro- magnetic wave's transit speed in ice. Frequencies between 75 and 200 MHz were used to achieve penetration of the ice, using a single cycle or "pulse" of the radiofrequency (RF) carrier for good dis- tance resolution of the two echoes. This has been the primary concept adopted for use in most GPR's. delay, phase, polarization, and propaga- tion direction. The return signals are normally displayed and recorded as ampli- tude with respect to delay time. The frequencies used are usually in the 30- MHz to 2-GHz range (8^, 23). Another form of GPR is the video pulse radar. It transmits a very short pulse with a bandwidth that is very wide. This video pulse output spectrum spreads from essentially direct current to beyond 3 GHz. The signal scattered by the target can then be sampled over a very broad frequency range, and information about the target can be derived from each sam- ple. The extremely short pulse also al- lows for "time isolation" ; the trans- mitted pulse magnitude can fall to a very small value before the reflected pulse returns. This avoids signal interference (23, 25). However, several variations of the sin- gle or short pulse radar systems have been proposed and tested for use in GPR. Most of the successful ones have been of the short pulse type (fig. 1). In GPR work, compared with standard radar use, the distances involved are very short. Hence, the times of travel of an electro- magnetic radar pulse and its reflection from a target are correspondingly short. To avoid interference between outgoing pulses and the returning reflected pulses, it is necessary to keep the pulses extremely short. The problem is complicated by the fact that certain fre- quencies can achieve better depth pene- tration, and these optimum frequencies vary according to the nature and type of the soil or rock. The most common type of short pulse GPR uses a short pulse of known waveform whose return signal is received in the time domain and can provide information about the target from its amplitude, time ^Underlined numbers in parentheses re- fer to items in the bibliography at the end of this report. Transmitter 1 r Receiver Transmit antenna Transmit electronics Receive antenna Receive electronics _l Controller Timing control J "I Amplifiers Filters r Tape recorder Facsimile display FIGURE 1. - Block diagram of a simple short pulse radar system. Synthetic pulse radar (fig. 2) trans- mits a single frequency at a time. The returning signal is measured for its am- plitude and phase, and the transmitter frequency is then stepped up to the next value, and so on. From the return data, a pulse can be reconstructed by taking the inverse Fourier transform of the re- ceived signal ( 146 ) . A frequency modulated-continuous wave (FM-CW) GPR system sweeps through a band of frequencies that it transmits and then compares the instantaneous output frequency with the frequency of the sig- nal reflected from the target. The dif- ference in frequencies is a function of the distance to the target and the di- electric constant (or constants) of the material in the wave path (8, 46-49). Each of these GPR systems has advan- tages and disadvantages. Each system al- so involves the use of digital circuitry to analyze the returning signal from the target. Workable models of each system have been developed and tested as de- scribed later in the text. ELECTROMAGNETIC TRANSMISSION THROUGH THE EARTH Studies on the transmission of electro- magnetic waves through the earth date back to the 1920' s. These were undertak- en for geophysical research and to devel- op through-the-earth mine communications , and many of the studies were done by the Bureau of Mines (72-74, 76 , 123 ). By the 1960's, theoretical models of electromag- netic wave propagation in the earth had been developed, owing largely to the work of James R. Wait and others. In a series of seminal studies ( 129- 137 ) , Wait analyzed the theoretical response of the uniform earth to vari- ous forms of electromagnetic excita- tion and derived mathematical solutions to describe it. He, along with F. C. Tape drive Control unit Control micro- processor A/D conversion Synthe- sizer ^ F.O. trans Offset synthesizer I MHz if reference Sample and bold Sample and hold Transmitter I__j 1^ Power _ jV — V amplifier I F.O. X' ' cable ^ F.O. RCVR. Quadrature hybrid mixer I r 90' mixer U ! ANT XMIT-if reference Receiver 1st mixer Amplifier ANT FIGURE 2. - Block diagram of a synthetic pulse radar system. Frischknecht and others, later extended the analyses to a two (or more) layered earth (59, 127-128 , 137 ). These cases utilized detectable interfaces within the earth. Investigators in the 1950' s and 1960's also studied cases involving vary- ing frequencies, as well as transmissions from loops, dipoles, and infinitely long wires in various locations and orienta- tions (23-J^, 60, 62, 125-126 ). Much of this theoretical work, which underlies most through-the-earth electromagnetic radiation studies, would later prove to have only limited applicability to GPR (27, 61, 64, 67, 80, 111-113 ). The fre- quencies studied were so much lower than those used in actual GPR that different electronic phenomena predominated. A coal seam presented a special case. Wait pointed out that electromagnetic waves could propagate laterally in re- sistive layers ( 138 ) . A coal seam can be considered a resistive slab bound by con- ductive rock. In such a model, the domi- nant mode of transmission of electro- magnetic energy has low attenuation, a result that many observers had already noted empirically. Not all RF energy penetrated deeply into the earth. Clough (29) pointed out that an electromagnetic wave impinging upon the earth at the critical angle given by Snell's law will be refracted and will give rise to lateral waves. These will usually be of low amplitude and lost in other signals, but they could predominate in some modes. This might be particularly true of GPR in contact with the ground surface or a coal face. The physical basis of electromagnetic surface waves was studied by Lytle, Miller, and Lager (^9^), who described the physical phenomena that occurred and derived the mathematics relating them. A method of determining the distance to a buried target was described by Clay, Greischor, and Kan (28^) in 1974. Using matched filter detection, a technique al- ready common in other electronic applica- tions, the investigators were able to de- termine the distance to conducting layers in the earth. In this method, a set of filters are constructed to match the waveforms of signals that have traveled to different depths within the earth and have been reflected. ELECTRICAL CHARACTERISTICS OF EARTH AND ROCK Concurrently, there was much research being done to determine the basic elec- trical characteristics of soil and rock. The mechanisms by which solids propagate electromagnetic energy were discussed at length by Von Hippel ( 124 ) and Meakins (90) . Collett and Katsube also reviewed electrical characteristics of rock (con- ductivity and permittivity, or alternate- ly, loss tangent), how they are deter- mined, and the measuring systems and techniques used to determine these char- acteristics in the frequency range 10^ to 10^ Hz (30). Ward and Eraser ( 142 ) also described the conduction of electricity in rocks. Methods of measuring rock electrical characteristics were reported by Yatsyshin, Zhuk, Salamatov, and Yako- vitskaya ( 147 ) and Lytle (85) , who in- cluded a good overview of the subject as well, and Khalafalla and Viner (78) , who described methods of dielectrometry ap- plicable to the 10^- to 10^-Hz range. There are many other valuable published reports (6^, 41, 66^, 79, 103 , 118 , 122 ). They vary as to the type of soil or rock studied, frequencies used, method of mea- surement, whether the measurements were in situ or in the laboratory, and other factors. Watt, Mathews, and Maxwell pub- lished an early study of Earth crust electrical characteristics ( 144 ) , while Griffin and Marovelli determined the di- electric constants and attenuation fac- tors for six basic rock types in the frequency range 20 to 100 MHz (63) . Hoekstra and Delaney (70) studied the di- electric constants of soils at ultrahigh frequency (UHF) and microwave frequency ranges. Saint-Amant and Strangway also reported on the dielectric constants of various rocks and soils ( 1 10 ) in a de- tailed investigation of powdered and sol- id dry rocks in the frequency range 50 kHz to 2 MHz. They found that both the dielectric constant and the loss tangent increase with increasing frequency and increasing temperature. At high enough frequency (over 1 MHz), the loss tangent approaches a constant value. A valuable study by Hipp ( 68 ) related electromagnet- ic propagation parameters to functions of frequency, soil density, and soil mois- ture. The last has particular applica- bility to coal. In a related study, Campbell and Ulrichs ( 21 ) pointed out the geologic implications that might be de- rived from the electrical characteristics of lunar rocks. ELECTRONIC CHARACTERISTICS OF COAL Other studies have a more direct bear- ing upon GPR (2, 4-, 50). Cook, in 1970, reported his investigations of bituminous coal, in which he tried to determine the transparency of coal to very high fre- quency (VHF) radiation (32). Two fresh bituminous coal samples were tested at 1, 5, 25, and 100 MHz in a measuring circuit utilizing a capacitance test cell and an RF bridge. In situ tests were also car- ried out. He found that at 60-dB signal attenuation he could obtain a 50-m pene- tration with a 100-MHz signal and a 1- to 10-km penetration distance with a 1-MHz signal. The presence of clay or pyrite veins in the signal path strongly af- fected the penetration and prevented du- plication of these results. Cook then undertook the systematic study of a rep- resentative range of rocks (38 kinds) to determine their transparency to radar frequency electromagnetic energy (35) . Using short, broadband pulses at 1, 5, 25, and 100 MHz, he measured the trans- mission penetration with a circuit using an RF impedance bridge and a parallel- plate capacitance test cell. He found that low-loss propagation was possible in certain granites, limestones, coals, and dry concrete, and that (then) exist- ing VHF mining radar could explore effec- tively for distances of up to several hundreds of feet. Shorter but still use- ful distances were possible for other coals , gypsums , oil shales , dry sand- stone, high-grade tar sands, and schists. For most shales, clays, and fine-grained soils, probing distances were generally restricted to less than 10 ft. Signifi- cantly, RF losses increased by as much as a factor of 12 when the samples were wetted. The controlling factor determin- ing rock transparency to radar evidently was the uncombined moisture content of the rock. Balanis, Jeffrey, and Yoon (_5) measured the microwave transmission characteris- tics of coal samples. Using two coal samples, one fine and one granular, each at two "typical" moisture levels (10 and 15 pet) , they measured the transmission of 8.2- to 12.4-GHz waves through the samples. Two different waveguide tech- niques were used in the tests, to measure a sample with finite length as well as to simulate a sample of infinite length. Later, Balanis, Shephard, Ting, and Kar- dosh (_7) reported on the anisotropic properties of coal in the radar frequency ranges, by using a two-path interferom- eter at 9 GHz. They measured the dielec- tric constant and conductivity of coal as a function of the direction of propaga- tion of the transient signal and as a function of the polarization of the elec- tromagnetic wave in the coal. Using four Eastern bituminous coals and one Eastern anthracite coal, they also studied the effects of certain other physical proper- ties upon high frequency electromagnetic conduction. These included the rank of the coal, the pyrite concentration and distribution, the mineral content, and the moisture content. Anthracites have higher conductivities by a factor of 10, and higher permittivities by a factor of 2 to 3 , than do bituminous coals. The investigators also found that high fre- quency electromagnetic anisotropy corre- lated quite well with optical anisotropy. APPLICATIONS OF GPR RADAR MEASUREMENTS OF ICE AND SNOW THICKNESS While basic work on earth dielectrics and conductivity continued, there was al- so further research ( 101 ) into the prac- tical applications of GPR. An earlier proposal of Cook had been the use of air- borne VHF pulse radar for the measurement of ground ice and snow (33) . The idea soon became reality in Antarctica, where intensive multinational explorations were in progress. Airborne VHF pulse radar in various vehicles proved an invaluable tool in profiling the snow and ice masses of the continent (52-53, 55, 106 ). The methods were quickly applied elsewhere. Luchinonov (84) reported on the use in the Soviet Union of airborne radar to delineate mountain glaciers. In Canada, airborne radar was used to measure the depths of ice and snow ( 99 , 102 ) . Camp- bell and Orange (20) reported on the use of pulse radar for the continuous pro- filing of sea and fresh water ice thick- ness in Arctic regions. GPR was also shown ( 88 ) to be useful in permafrost as it penetrated deeply enough to be meaningful. Annan (_3) reported in 1976 on the use of a wideband (150 MHz), short pulse ra- dar with a 110-MHz frequency, 50-W out- put, and a 50-kHz repetition rate for underground surveying in the Arctic. The long wavelengths of the radar limited the resolution to features near the surface. Higher frequencies were suggested for fu- ture use. Despite its inadequacies, GPR in general proved its utility as an un- derground surveying method, particularly if used in conjunction with drilling. Wide-angle reflection and refraction sounding techniques (WARR) , used in seis- mic sounding, were suggested for use in layered areas. Reliable sounding from 3 to 30 m should then to be possible with GPR in permafrost. In a related experi- ment, Bertram, Campbell, and Sandler (11) reported using short pulse radar to lo- cate large masses of underground ice. Harrison and a British group ( 65 ) used a pulse-modulated radar at 35 MHz from a U.S. Navy aircraft to penetrate glacial ice and allow the creation of a graphical reconstruction of the underlying terrain. The radar succeeded in delineating the various interfaces between ice and rock, ice and air, brine and ice, brine and air, and so on. The radar signal returns were recorded on 35-mm film for later analysis, which proved difficult because of the complicated ray paths of the radar signals and echoes. The relative- ly long wavelengths used did not give great resolution, but the system provided a valuable tool for subglacial explora- tion through maximum ice depths of 1 to 3,000 m. On the frozen Great Lakes, Cooper, Mul- ler, and Schertler (40) experimented with the use of a GPR mounted on an all- terrain vehicle for rapid profiling of the lake ice depth. Earlier experiments from a C-47 aircraft at a 2,300-m alti- tude and a 75-m/s groundspeed had given inconclusive results because of the in- herent difficulties in locating the air- craft's position accurately enough to make the later calibration measurements by surface auger meaningful. They could not determine how close to reality the observed thicknesses (10 to 92 cm) actu- ally were. Experiments with a Coast Guard heli- copter showed similar disadvantages. The airborne radar did not react to ice thicknesses of less than 10 cm, and any surface melt water over 1 mm thick blocked the radar. Hoping to overcome these problems, a short pulse, 2.86-GHz radar with 20-W peak power was mounted on an all-terrain vehicle to be used on the frozen lake surface. Extensive laboratory measurements of the dielectric constant of lake ice were made under various conditions before the actual tests were made on the lake. As the vehicle-borne radar proceeded across the lake ice, work teams bored auger holes for verification measurements. The results showed a maximum depth overesti- mate by radar of 9.8 pet, a maximum un- derestimate of 6.6 pet, and an average error of 0.1 pet. Upon analysis, it ap- peared that most of the Inaccuracy was due to error in auger depth measurements and the rest to nonuniform dielectric constants or the resolution of radar at the wavelength used. RADAR MEASUREMENTS OF SALT DOMES GPR's have also been used in salt domes where the much lower RF attenuation of salt has allowed its use over much great- er penetration ranges. Several research- ers associated with Texas A, & M, Univer- sity (69^, 116 , 122 ) did early work in this area. One group ( 71 ) utilized bore- hole radar, made small enough and rugged enough to be used in an oil well under 10,000-lb/in2 pressure and at 250° F. The 250-MHz pulse radar with a 400-Hz pulse-repetition rate and a 10-kW peak power was lowered into the borehole. At- tenuation through salt ranged from 1.2 to 2.8 dB per 100 ft, which gave penetration ranges through rock and salt of up to 8,208 ft. Internal reflecting surfaces within the salt body produced anomalous reflections that were difficult to inter- pret. Despite this and other difficul- ties of interpretation created by an in- sufficient data base, the technique was considered a viable tool for geophysical exploration. Another group ( 115 ) made extensive tests of a 440-MHz radar in the Cote Blanche Salt Dome. It was used within the mine to outline the salt body and to detect discontinuities within the salt. The roof of the salt dome was also outlined. Penetration ranges of up to 2,040 ft were achieved at low power, and 3,177-ft ranges were achieved at high power. ARCHEOLOGICAL AND GEOTECHNICAL APPLICATIONS GPR has been used in archeology as an exploration tool by a group ( 43 ) seeking a means of detecting hidden chambers within the Great Pyramids of Egypt. The "sounder" operated in the frequency range 16 to 50 MHz, with a pulse 1-1/2 cycles in length and a peak power of 0.2 MW. It was tested in a Western dolomite mine where empty chambers 100 to 130 ft from the transmitter produced good radar echoes. Lytle and Lager (86) used a 3- to 50-MHz transmitter at the Wowona tun- nel in Yosemite National Park for tests from within the tunnel to the surface, through 300 m of solid granite. They measured the in situ bulk conductivity of the rock as well as the relative dielec- tric constant as a function of frequency. Geophysical Survey System, Inc. , engi- neers (99) , have put together a small portable array of GPR's designed for near-surface location of various targets. The signal is a wideband (120 MHz) pulse with a repetition rate of 50 kHz. The antenna is a small wheeled carriage that can be towed or pulled by hand up to 2-3 mi/h. The radar receiver's output is by means of a graphic recorder, a continuous strip chart that provides profiles of the underground target. The set is easily portable even in an experimental form. It has low power requirements (12 V, 1.5 A) and uses the same antenna for trans- mission and reception. The system uses a time-domain sampling technique, which reduces the signal frequency to audio ranges where it can be conveniently re- corded on the strip chart. The unit has been successfully tested for the detec- tion of pipes buried up to 13 ft below the surface, for use in the presence of multiple targets, to detect underground wires, and to profile lake and delta deposits. Another imaginative use of GPR came from New Mexico where National Bureau of Standards (NBS) investigators, working with the National Park Service (91) , used a FM-CW radar to find the structural weaknesses of an ancient Spanish mission undergoing restoration. The technique holds much promise for archeological work, particularly since the radar re- turns give an accurate representation of the water distribution in the sample area. GPR was also suggested (10) as a basic tool for geotechnical site assessment in conjunction with traditional survey meth- ods. Radar was suggested as a cheap and effective method to fill in between ex- pensive exploratory boreholes. It also could be used in hydrological and soil studies, and analyses of hazardous mate- rial disposal sites. However, GPR ap- pears to be extremely "site specific" in its ability to penetrate the earth be- cause of the differences in soil and the presence or absence of water. The U.S. Air Force (12) has also shown interest in GPR for use in rapid evaluation of air- field pavements for damage under combat conditions. SUBHIGHWAY MEASUREMENTS There is a growing awareness of the po- tential for subsurface testing offered by GPR. It is obvious GPR is a viable new tool for geologists and engineers ( 106 ) . Highways continue to be a fruitful area for the application of the new technol- ogy. A persistent problem for highway builders is the instability of sloping ground, caused by subsurface geologic anomalies. Exploratory drilling is ex- pensive and frequently gives incomplete data. Working with the Federal Highway Administration (FHWA) and the U.S. Geo- logical Survey (USGS) , NBS developed four microwave probing systems ( 120 ) . From them, the FM-CW system was chosen for highways investigation because of its suitable resolution at ranges of inter- est. It had adequate penetration, could be built with commercially available com- ponents , and had previously been success- ful in a subsurface survey. The FM-CW system was built and subjected to 5 weeks of testing in the Pike National Forest in Colorado. Boreholes and borehole TV cam- eras were used to determine the details of underground geology at the test site and to verify the results of the radar tests. The FM-CW radar successfully lo- cated a series of fractures in granite at depths between 5.96 and 6.86 m. Moore, Echard, and Neill (98) reported the use of a short pulse radar (approxi- mately 1 ns) to detect the presence of voids under concrete slabs on an inter- state highway. Experiments were done on test blocks and on a highway. The vary- ing radar responses seemed to delineate a true picture of the underslab voids, but it was impossible to verify them. When the group attempted to drill into the voids for verification, the cooling water used in the drills rushed through the newly opened holes into the voids , carry- ing with it dirt and concrete, disturbing the voids. Also, the effect of the dif- ferences in dielectric constant of the concrete and earth was poorly understood and was not compensated for properly. The technique works , but a large body of experimental data will have to be built up for it to reach full potential. MICROWAVE MEASUREMENTS OF COAL SEAMS Delineation of coal seams (thickness, shape, and orientation) using microwave radar was a subject of several investiga- tions. A theoretical basis for the new technology was derived by Wait ( 139 ) . Under a Bureau of Mines contract, NBS (47) used a microwave (0.5 to 4.0 GHz) radar to measure coal seam thickness in two coal mines. The technique worked to measure coal thicknesses from 10 to 40 cm, but some anomalies in the coal seams confused the results. However, the re- searchers recognized that detection of these anomalies might be the most valu- able characteristic of microwave sensing. They also found that the dielectric con- stant was a function of the water content. NBS subsequently developed a FM-CW ra- dar system operating in the 1 to 2 GHz range (48-49) . It was tested in the lab- oratory and in three coal mines to mea- sure coal seam thickness up to 55 cm. The technique was suggested for possible use in controlling the operation of auto- mated mining machinery at the coal face. GPR SYSTEMS VIDEO PULSE RADAR APPLIED TO GPR Video pulse radar techniques have been applied to GPR. A video pulse radar uses broadband signaling but with the wave- forms consisting of multiple frequencies (from near direct current to megahertz or gigahertz), instead of broadcasting many cycles of one frequency, as does a tradi- tional radar. This multif requency wave will be perturbed by any change in under- ground electrical characteristics such as different soil or rock types that have differing dielectric constants, or in- terfaces such as cavities, joints, or faults. The transmitter sweeps through the frequencies, and the reflected elec- tromagnetic wave is picked up by the re- ceiver to give a voltage-versus-time out- put. The waveforms can be processed for both time and frequency domains. This is done experimentally by a digital comput- er, but in actual operating units a built-in logic circuitry would do it. Using a video pulse radar, Moffatt ( 94 ) easily detected a tunnel through 20 ft of limestone, as well as some previously un- suspected steel supports. An effort to detect faulting in a quarry produced am- biguous results, however. The complex fracturing of the rock made it difficult to relate the electronic information to physical reality. A study by Burrell and Peters ( 17 ) pointed out that using the "Low Frequency Window" (LFW) , a frequency range in which the attenuation of RF waves is relatively low (below 100 kHz), penetration ranges of several kilometers were possible. One problem was the need to decouple the pow- erful transmitted signal from the weak reflected wave, which it could mask. So that the return signal could be observed, it was necessary for the output signal to fall to zero very rapidly. This was ac- complished by including a high frequency portion of the pulse and by developing orthogonal antennas to electronically separate the output and input signals. Video pulse radar was used by Cook in a series of experiments (37) for the British National Coal Board. The British needed to detect unmapped, abandoned mineworkings in advance of the mining face, because abandoned galleries, if penetrated unexpectedly, could release floods of mud or water under high pres- sure. Cook and his associates reported much work in the laboratory and tests in mines. They reported the relative radar transparency of various minerals and an- tenna design parameters in an actual mine test. Good results were obtained with a radar operating at 100 MHz with one-way propagation ranges from 6 to 56 ft and reflections through 28 ft of coal. These results were obtained despite having to use an inadequate antenna designed for the open air. Attenuation was from 1 to 1.6 dB/ft of coal, and particularly good reflections were obtained from any target that was wet or contained water. Long wavelengths gave poor target resolution, but there was far less loss of energy by scattering if the impinging wave had a wavelength greater than either the thick- ness of the layer of coal, shale, or limestone, or the individual rock crystal size. The radar successfully detected mine openings through 30 to 60 ft of rock in narrow, cluttered, dirty, wet under- ground workings , and was unaffected by the presence of steel paneling or beams nearby. The outlook for deep electromag- netic probing in soft ground such as sand, earth, or mud was not so favorable. Even at low frequencies , 50 to 70 ft seemed to be the ultimate range. The ra- dar signals. Cook reported, were easy to interpret about half the time. Image in- terpretation was still in its infancy in underground radar, and there were many unexpected and unexplained results. There were spurious echoes on some occa- sions and no reflections on others. More work, it was suggested, needed to be done in mines under known conditions. The most important factor was the absorption of rock; less important were the choice of principal frequencies, the transmit- ter's electronic efficiency ("performance figure"), or the antenna's directivity. GPR, Cook said, offers great potential for exploring large volumes of rock at 10 low cost in advance of mining. Other tests (96) , of a road tunnel (with steel supports) in limestone, a fault in a do- lomite quarry, and a coal mine tunnel, seemed to bear out Cook's prediction. SHORT PULSE GPR Researchers at ENSCO, Inc., now XADAR Corp., developed a short pulse GPR, which was tested in coal mines in 1977 and 1978 (39, 51 , 57) . The portable set consisted of a transmitter, a receiver, and a con- troller. The transmitter emits a short, high frequency (30- to 600-MHz) pulse that is coupled to the ground through the antenna. The transmitted pulse is shaped like a single cycle of a sine wave. The pulse is radiated from the antenna into the earth in a broad pattern. Whenever the radar pulse encounters a medium with different electrical properties, a por- tion of the electromagnetic energy is re- flected back to the receiver. The change in electrical properties coincides with changes in the nature of the mediiam, such as a fault or clay vein passing from the coal seam into shale or rock. By oieasur- ing the time of passage of the radar wave and its reflection, and by knowing the velocity of the wave, the distance to the anomaly can be determined. However, in coal and rock, unlike air, the speed of propagation is not a constant and must be separately determined for each site. It is normally in the range of 0.2 to 0.5 times the speed of light. Using this type of GPR, ENSCO researchers were able to propagate (one way) a signal 100 ft in a coal seam, using signals in the 40- to 500-MHz range. Voids were detectable us- ing the GPR at 50 ft, in the frequency range of 10 to 160 MHz. An uncased bore- hole was detected over 25 ft away. BOREHOLE RADAR Radar was also applied ( 119 , 145 ) to borehole survey. Under a Bureau con- tract. Southwest Research Institute ( 117 ) developed a borehole radar. The entire package was required to be less than 4 in. in diameter and 1 m long, which re- quired a special antenna design to fit the cylindrical shape. Representative models of what the various geologic anom- alies in mines would look like on a radar return had to be developed. A short pulse, high-power radar, with a 15- to 30-m range surrounding the borehole, was developed using a sampling oscilloscope and signal averaging technique. The data were digitized and fed into a computer for clutter rejection. An experimental system was built and tested in single- and cross-borehole operations. There were problems of data interpretation and with penetration range. Distant targets were dim because of medium attenuation and normal geometric dispersion, while near targets reflected very strongly. A time-varying gain amplifier was used to overcome this. The system proved more suitable for hard rock than for sedimen- tary rock, and there were problems of spurious signals from electric cables and other components. In a later phase of work, a borehole directional radar probe was developed by Southwest Research Institute ( 116 ) and subjected to demonstration testing in 1982. The radar was a video pulse radar. Each pulse had a time duration of 12 ns and a useful frequency spectrum of 30 to 300 MHz with a 1-kW peak power output. The probe (16.7 ft in length) was de- signed to fit within a 4-in borehole. The unit comprised the borehole probe and the surface control unit connected by or- dinary armored four-conductor wire line cable. The borehole probe transmitter utilized a unique rotating directional antenna. Earlier units tested with sin- gle antennas for transmitting and receiv- ing had proven unsatisfactory. The new antenna scanned in eight angular steps (45° apart) according to the command of the surface control unit. The actual or- ientation was determined by a built-in fluxgate north-seeker circuit for verti- cal drill holes or a pendulum vertical reference for horizontal holes and was reported to the surface unit. The re- ceiver antenna was a short, omnidirec- tional monopole. A time-domain sampler developed a low frequency replica of the reflected signal from the receiver output 11 by repeated sampling and enhancement by signal averaging. This replica was typi- cally in the audiofrequency range (3 to 30 Hz or 300 to 3,000 Hz) and could be transmitted by the standard geophysical armored four-conductor wire line cable. The unit was field-tested in York Can- yon Mine near Raton, N. Mex. This mine was chosen for the tests because of ease of access and for its favorable electro- magnetic propagation characteristics, and because it had a well-documented geologic anomaly to serve as a target. The anom- aly was a 7-ft coal seam displaced 50 ft by a complex fault that had stopped min- ing development and had been encountered at points 700 ft apart. In the actual tests, the returns appeared to be omnidi- rectional, presumably because of the poor directional response of the antenna sys- tem. The fault was identified by one borehole probe as a target reflection from 22 ft away. Another borehole probe, 20 ft from the fault, failed to get a clear response from the fault. As part of their work in developing cross-borehole methods ( 23 , 25 ) for use in detection of mining hazards for the Bureau of Mines , Lawrence Liveirmore Na- tional Laboratory (LLNL) undertook exper- iments at the Gold Hill Mine test site near Boulder, Colo., in 1980. The tests involved using a high frequency electro- magnetic (HFEM) system to locate a tunnel 49.5 m below the surface between two boreholes 10 m apart. A transmitter, broadcasting at a constant frequency and amplitude at frequencies between 45 and 70 MHz, was lowered down one borehole. The receiver was lowered down the other. Two experimental scenarios were used; in the first, the transmitter and receiver were lowered at a constant rate so that they were at the same or "common depth" at all times. In the second, the trans- mitter and receiver were lowered with a constant difference between their depths, which was maintained to provide "offset views" for better definition of the tar- get. Common-depth tests were run at 45, 50, 55, 60, 65, and 70 MHz and offset- view tests at 45, 50, and 55 MHz. Data were plotted as signal strength versus depth, with the tunnel location indicated by signal minima. The tunnel was accu- rately located; it lay between test bore- holes only 10 m apart. The size of the tunnel was not so accurately determined; it was measured as having a height of 1.7 m when it actually was 2.8 m high. Lawrence Livermore Laboratory also de- veloped and tested a high frequency (mi- crowave), cross-borehole electromagnetic system (81) designed to detect water- or air-filled underground cavities. At Man- atee Springs, Fla. , tests were run on known water-filled caverns using micro- wave propagation (5 to 105 MHz) between two boreholes , one containing the trans- mitting antenna and the other the receiv- ing. In this test, the test signal fre- quency was varied from 5 to 105 MHz and the depth from 30 to 38 m to encompass the area where the cavern was expected to be located. At Medford Caves, an air- filled cavern was mapped by using a single-frequency scan (100 MHz) and vary- ing the depth from 25 to 70 m. The tests indicated that air- or water-filled cavi- ties have characteristic electronic sig- natures but are difficult to identify if the cavity is irregularly shaped or if the matrix is extensively fractured. A medium of high conductivity creates problems also, by making it difficult to propagate a signal of sufficient strength. SYNTHETIC PULSE RADAR Under Bureau contract, XADAR Corp. (formerly ENSCO) is investigating a syn- thetic pulse radar for coal mine use (51). Short pulse systems have many dis- advantages: the need to separate the transmitted and reflected signal, sam- pling techniques that use only part of the transmitted energy, and complicated antenna design. A synthetic pulse radar system is designed to overcome noise problems, to give a higher signal-to- noise ratio, and to increase penetration using lower energy. The latter is neces- sary for permissibility for use in under- ground coal mines. The synthetic pulse 12 radar transmits a single frequency at a time. The amplitude and phase of the re- ceived signal are measured, and the fre- quency is then stepped to the next value. If the proper frequencies are transmitted and received, then a pulse can be recon- structed by taking the inverse Fourier transform of the received data. The sys- tem does require more signal and data processing and more sophisticated equip- ment. In the initial test, a breadboard unit was built and tested. It was de- signed for operation within a transmis- sion frequency range of 20 to 160 MHz with a minimum spacing of 100 kHz. This should give a maximum time "window" of 10,000 ns or a reflection range of ap- proximately 2,500 ft in coal, much great- er than needed. In the breadboard tests, many problems with internal interference were encountered, and finally fiber op- tics had to be used to isolate some sys- tem elements. There were also problems with ghost signals. Nevertheless, it was shown that a prototype system having an effective penetration range of 200 ft could be developed. From the experience gained from the breadboard unit, a prototype synthetic pulse GPR was built (146). It was com- pleted in June 1982 and then tested. There were many changes from the original breadboard unit. The system used a het- erodyne receiver, which uses two mixers. This allowed the detection circuits to operate at the lower intermediate fre- quency (IF) of 1 MHz, which eliminated many problems associated with the higher frequency circuits. The signal sent to the control unit from the receiver was at the intermediate frequency. Despite equipment difficulties that prevented accomplishment of all the experimental objectives, the unit was successfully tested in Eastern and Western coal mines. It was found that the synthetic pulse GPR could penetrate greater distances (ap- proximately 2 times farther) than short pulse GPR. Reflection depths of 40 ft were achieved in the Deseret Mine, Utah, and the background noise was found to be 60 dB lower than the main signal arrival. It was estimated, based upon these re- sults, that reflection ranges of 180 to 200 ft are possible. Because of equip- ment difficulties, the antenna tuning feature planned for the unit was not built and the display unit was not fin- ished in time for the mine tests. More work on this equipment is planned. STRATA CONTROL RADAR The Bureau is also investigating a GPR for the identification of roof conditions in underground coal mines that might lead to roof falls. Most roof falls in coal mines are associated with geologic anom- alies within a few meters of the immedi- ate mine roof. Therefore, the primary interest of researchers is in very small, but hazardous rock features within a very narrow band of rock above or behind the surface of mine openings. This means that higher frequencies are required than for most GPR systems, in order to resolve the tiny anomalies, for example, cracks. Although attenuation is much greater for higher frequencies, the area of interest is of such shallow depth that the in- creased attenuation is an acceptable trade-off for better resolution. Primarily utilizing off-the-shelf com- ponents, a short pulse GPR was put to- gether in 1981, for tests and evaluation in the laboratory and mines. The trans- mitter of this radar emits a single pulse of 4-ns duration and 400-V amplitude at 100 or 250 MHz. Both transmitter and re- ceiver utilize triangular dipole anten- nas. The receiver consists of a voltage variable attenuator, a high-gain pream- plifier, and a sampling unit. It is con- trolled by a small computer that performs several functions. The computer provides an external signal to the sampling unit to set the sampling gate on a particular point of the received pulse. It converts the sampling unit output to digital form and averages the signal from a number of pulses to remove noise. It then steps the sampling gate to the next point on the pulse and repeats the process. When a full pulse has been sampled, the 13 computer stores the data on a magnetic tape. In addition, it supplies an expo- nentially variable voltage to the voltage variable attenuator to adjust the receiv- er system sensitivity to compensate for the signal attenuation through the rock. At the completion of a run, the small computer can be connected to a large com- puter and the data transferred to a disk file for later analysis. Several surface and three underground tests of the strata control GPR were made in 1982 and 1983 in mines near Tusca- loosa, Ala. Figures 3 and 4 show the computer control console and antenna in- stallations underground. Radar signa- tures were correlated with rock strati- graphies. In tests, three rock strata were identified; the middleman rock 2 to 3 ft thick, the overlying coal seam 3 to 6 ft thick, and the main roof. These were correlated with ground-truthing test holes , drilled and borescoped in the mine roof. The data are currently being computer-enhanced so that they can be displayed in a two-dimensional graphic representation that reveals stratigraphic conditions and any anomalies encountered in the radar scan. FIGURE 3. - Radar control console of GPR used during underground testing. 14 FIGURE 4, - Antenna deployment against mine roofo GPR ANTENNAS A group from the Electromagnetics Divi- sion of NBS evaluated antennas suitable for GPR for the Bureau of Mines (8^). Af- ter a literature review had narrowed the possible choices, several antenna types were tested, most of which were horn-type antennas. Eventually one was selected as being the best compromise among competing requirements, for example, small physical size needed in mine work and some supe- rior electronic characteristics of larger antennas. Transverse electromagnetic (TEM) horns appeared to be the most prom- ising, so most of the tests were run on modifications of this basic type. Even- tually three antennas were built, which were TEM horns, labeled NBS-TG, They were small (49 cm long, 50 cm high, 21.5 cm wide) , metal TEM horns with variable width and variable flare angle. The NBS group also put together two GPR systems , a pulse radar time-domain system and a FM-CW system. Only initial evaluations of these two systems were undertaken. INTERPRETATION OF GPR DATA The question then becomes not whether radar can penetrate the ground and be used to map ahead of mining, but increas- ingly one of analysis (9^, _22, 82^, 92, 97 , 105 ) of radar returns to determine the nature of the target. For example, an ambitious study (23) was made by investi- gators at Ohio State whose purpose was to develop a simple, automatic, "single look," real-time method of identify- ing subsurface targets. A commercially available radar was used. It was a video pulse radar (direct current to more than 3 GHz) with a very short pulse that al- lowed for "time isolation" or a radar "dead zone" between the transmitted and reflected signals. Separate orthogonal dipole antennas for transmission and re- ception aided signal separation. The or- thogonal antenna was Insensitive to lay- ering within the earth. The receiver used a sampling oscilloscope that reduced 15 signal noise by averaging a large num- ber of waveforms rather than by using filters. A signal processor was used for clutter and noise reduction, and another unit for target characterization and identification. A general purpose digital computer was used for characteri- zation and identification of the targets as well as for controlling the sampling oscilloscope. In the experiment, five targets of dif- fering shapes and materials (plastic landmine, brass cylinder, aluminum sphere, copper sheet, and woodboard) were buried at known positions and depths in a small yard at Ohio State University, Ra- dar reflection images of each target were obtained. The result was a processed time-domain waveform. These waveforms exhibited a transient behavior (a super- imposed signal upon the basic return waveform) in the time region late in the signal pulse, where only the natural re- sponse of the target had any effect. On analysis, the Fast Fourier Transforms (FFT) of these waveforms showed strong peaks indicative of resonance behavior. These proved unchanging with variations in antenna location or orientation. It was concluded from this that "the complex natural resonances of the target are ex- citation invariant." Thus, the individual targets produced "complex natural resonances," depending upon their shape and constituent mate- rial, which could be used as the basis for target characterization and identifi- cation. The signal level of these sever- al natural resonances varied greatly, creating a problem of dynamic range, and varying ground conditions introduced even more complications. The value of the complex natural resonances of the targets was calculated from experimental data by using Prony's method, a two-century-old system for the mathematical treatment of waveforms . These calculated waveforms were then compared to the actual wave- forms of the buried objects by the computer. The researchers at Ohio State continued to improve their process for using radar in identifying buried objects (25) . A redesigned antenna was a step forward. The new antenna resonated at the middle of the range of the band of complex reso- nances of the target. This provided a better match between the resonant fre- quency of the radar system and the fre- quency band spanned by the target reso- nances and allowed improved target identification. Interpreting the signal return re- flected back from a buried target is a difficult process. For targets in air, analytical methods modeled on physical optics are used to image the targets. The situation is more complex for buried targets, but target imaging is similarly based on the backscattered field ( 24 , 148 ) . Investigators at Ohio State work- ing under an Army contract found that , for below-the-surface target imaging, the most important part of the signal is the early return (high frequency) part of the backscattered field. The backscattered field is attenuated by the ground, and this attenuation determines the signal strength and, therefore, the depth of penetration. The backscattered field contains several complex functions of the reflected radar pulse and is very diffi- cult to interpret. The problem of the interpretation of reflected subsurface radar signals remains one that will un- doubtedly require extensive attention in the future. CONCLUSIONS It has long been known that electromag- netic waves can penetrate earth and rock. Various GPR's in portable, low-powered forms have been used successfully to lo- cate subsurface objects and voids in mil- itary operations, highway construction. and archeological work at short ranges. This same technology could be used to de- tect hidden geologic anomalies in mine roof , ribs , or faces , which might pose a safety hazard to miners. Bureau of Mines research efforts have been directed 16 toward investigating various types of GPR systems capable of detecting such hidden hazards, ahead of and during mining. The electromagnetic radar systems that may have significance for exploration and mine safety include microwave radar. borehole radar, short pulse radar, syn- thetic pulse radar, and strata control radar. More research is needed to per- fect the systems that have been initi- ated, especially in the area of data analysis. BIBLIOGRAPHY 1. Alongi, A. V. Pavement Thickness Measured, Voids Detected by Downward- Looking Radar In New York Test. Ind. Res. News, Fall 1973, pp. 36-39. 2. Anderson, W. L. , and R. K. Moore. Frequency Spectra of Transient Electro- magnetic Pulses in a Conducting Medium. IRE Trans, Antennas and Propag. , v. AP-8, No. 4, 1960, pp. 603-607. 3. Annan, A. P., and J. L. Davis. Im- pulse Radar Sounding in Permafrost. Ra- dio Sci., V. 11, No. 4, 1976, pp. 383- 394. 4. Arcone, S. A. Distortion of Model Subsurface Radar Pulses in Complex Di- electrics. Radio Sci., v. 16, No. 5, 1981, pp. 855-864. 5. Balanis, C. A., J. C. Jeffrey, and Y. K. Yoon. Electrical Properties of Eastern Bituminous Coal as a Function of Frequency, Polarization and Direction of the Electromagnetic Wave and Temperature of the Sample. Paper in Electromagnetic Guided Waves in Mine Environments. Pro- ceedings of a Workshop (Boulder, Colo. , Mar. 28-30, 1978), ed. by J. R. Wait (contract HOI 55008, Nat. Telecommun. and Inf. Administration). BuMines OFR 134- 78, 1978, pp. 87-102; NTIS PB 289 742. 6. Balanis, C. A. , W. S. Rice, and N. S. Smith. Microwave Measurements of Coal. Radio Sci., v. 11, No. 4, 1976, pp. 413-418. 7. Balanis, C. A., P. W. Shepard, F. T. C. Ting, and W. F. Kardosh. Aniso- tropic Electrical Properties of Coal. IEEE Trans. Geosci. and Remote Sensing, V. GE-18, No. 3, 1980, pp. 250-256. 8. Belsher, D. R. Detection of Lost Oil Well Casings and Unknown Water Filled Voids in Coal Mines Through Development of a Microwave Antenna System (contract H0272007, NBS). BuMines OFR 6-79, 1978, 94 pp. 9. Benson, R. C. Application of Ground Penetrating Radar to Geotechnical, Hydrological and Environmental Assess- ments. Abstracts, 48th Annual Meeting and Exposition, San Francisco, Calif., October 1978. Soc. Explor. Geophys., Littleton, Colo., 1978, p. 279. 10. Benson, R. C, and R. A, Glaccum. Radar Surveys for Geotechnical Site As- sessment. Pres. at Am. Soc. Civil Eng. Conf., Geotechnical Engineering Div. , At- lanta, Ga. , October 1979, preprint, pp. 1-16; available for consultation at Tus- caloosa Research Center, Bureau of Mines, Tuscaloosa, Ala. 11. Bertram, C. L. , K. J. Campbell, and S. S. Sandler. Locating Large Masses of Ground Ice With an Impulse Radar Sys- tem. Proc. 8th Int. Symp. on Remote Sensing, Univ. Mich., Ann Arbor, Mich., October 1972, pp. 241-260. 12. Bertram, C. L. , R. M. Morey, and S. S. Sandler. Feasibility Study for Rapid Evaluation of Airfield Pavements. U.S. Air Force Weapons Lab., Dayton, Ohio, AFWL-TR-7 1-178, June 1974, 38 pp. 13. Bhattacharyya, B. K. Propagation of Transient Electromagnetic Waves in a Conducting Medium. Geophysics, v. 20, No. 4, 1955, pp. 959-961. 17 14. Bhattacharyya, B. K. Propagation of Transient Electromagnetic Waves in a Medium of Finite Conductivity. Geophys- ics, V. 22, No. 1, 1957, pp. 75-88. 15. Biggs, A. W. Radiation Fields From a Horizontal Dipole in a Semi- Infinite Medium. IRE Trans. Antennas and Propag. , V. AP-10, No. 4, 1962, pp. 358- 362. 16. Biggs, A. W. , and H. M. Swarm. Radiation Fields of an Inclined Electric Dipole Immersed in a Semi-Infinite Con- ducting Mediinn. IEEE Trans. Antennas and Propag., V. AP-11, No. 3, 1963, pp. 306- 310. 24. Chan, L. C. , D. L. Moffatt, and L. Peters, Jr. Subsurface Radar Target Imaging Estimates. IEEE Trans. Anten- nas and Propag., v. AP-29, No. 2, 1981, pp. 413-417. 25. Chan, L. C. , L. Peters, Jr., and D. L. Moffatt. Improved Performance of a Subsurface Radar Target Identification System Through Antenna Design. IEEE Trans. Antennas and Propag., v. AP-29, No. 2, 1981, pp. 307-311. 26. Chew, C. , and J. A. Kong. Elec- tromagnetic Field of a Dipole on a Two- Layer Earth. Geophysics, v. 46, No. 3, 1981, pp. 309-315. 17. Burrell, G. A., and L. Peters, Jr. Pulse Propagation in Lossy Media Using the Low-Frequency Window for Video Pulse Radar Application. Proc. IEEE, v. 67, No. 7, 1979, pp. 981-990. 18. Caffey, T. W. H. Two-Way Communi- cation With an Earth Penetrator. Radio Scl., V. 11, No. 4, 1976, pp. 267-273. 19. Caldecott, R. Electromagnetic Pulse Sounding for Surveying Underground Water. Electroscience Lab. , Ohio State Univ., Columbus, Ohio, Rep. 401X-1, 1967, 429 pp. 20. Campbell, K. J. , and A. S. Orange. A Continuous Profile of Sea Ice and Freshwater Ice Thickness by Impulse Ra- dar. Polar Rec. , v. 26, 1974, pp. 31-41. 21. Campbell, K. J., and J. Ulrichs. Electrical Properties of Rocks and Their Significance for Lunar Observation. J. Geophys. Res., v. 74, No. 25, 1969, pp. 5867-5881. 22. Chan, L. C. A Digital Processor for Transient Subsurface Radar Target Identification. M.S. Thesis, Ohio State Univ., Columbus, Ohio, 1975, 141 pp. 23. Chan, L. C. , D. L. Moffatt, and L. Peters, Jr. A Characterization of Subsurface Radar Targets. Proc. IEEE, v. 67, No. 7, 1979, pp. 991-1000. 27. Chlamtac, F., and F. Abramovici. The Electromagnetic Fields of a Horizon- tal Dipole Over a Vertically Inhomogene- ous and Anisotropic Earth. Geophysics, v. 46, No. 6, 1981, pp. 904-915. 28. Clay, C. S. , L. L. Greischor, and T. K. Kan. Matched Filter Detection of Electromagnetic Transient Reflections. Geophysics, v. 39, No. 5, 1974, pp. 683- 691. 29. Clough, J. W. Electromagnetic La- teral Waves Observed by Earth-Sounding Radars. Geophysics, v. 41, supplement, 1976, pp. 1126-1128. 30. Collett, L. S., and T. J. Katsube. Electrical Parameters of Rocks in Devel- oping Geophysical Techniques. Geophys- ics, V. 38, No. 1, 1973, pp. 76-91. 31. Cook, J. C. An Electrical Cre- vasse Detector. Geophysics, v. 21, No. 4, 1956, pp. 1055-1070. 32. . Electrical Properties of Bituminous Coal Samples. Geophysics, v. 35, No. 6, 1970, pp. 1079-1085. 33. . Proposed Monocycle-Pulse, VHF Radar for Airborne Ice and Snow Mea- surements. Trans. AIEE, Commun. and Electron., v. 79, pt. 2, 1960, pp. 588- 594. 18 34, Cook, J, C. Radar Exploration Through Rock in Advance of Mining, Trans, Soc, Min. Eng. AIME, v, 254, June 1973, pp. 140-146, 35. Radar Transparencies of Mine and Tunnel Rocks. Geophysics, v, 40, No, 5, 1975, pp, 865-885. 36. Radar, Seeing Through Rock With Ch, in N, Am, Conf. on Rapid Ex- cavation and Tunneling, Henniker, N,H, , 1972. AIME, Chicago, 111., v. 1, 1972, pp. 89-101. 37. Status of Ground-Probing Radar and Some Recent Experience. Proc. Eng. Foundation Conf., Subsurface Explo- ration for Underground Excavation and Heavy Construction, Henniker, N.H. , Aug. 11-16, 1974. Am, Soc. Civil Eng., New York, 1974, pp. 175-194. 38. . A Study of Radar Explora- tion of Coalbeds (Teledyne Geotech) . Bu- Mines OFR 5-72, 1971, 83 pp; NTIS PB PB 207 362. Electro-Magnetic Sounder Experiment. Geophysics, v. 39, No. 1, 1974, pp. 49- 55. 44. Duckworth, K. Electromagnetic Depth Sounding Applied to Mining Prob- lems. Geophysics, v. 35, No. 6, 1970, pp. 1086-1098. 45. Eberle, A. C, and J, D, Young. The Development and Field Testing of a New Locator for Buried Plastic and Metal- lic Utility Lines. Natl. Acad. Sci. — Natl. Res. Council, Washington, D.C., Transp. Res. Rec. 631, 1972, pp. 47-52. 46. Ellerbruch, D. A. Coal Layer Thickness Measurement Using FM-CW Sys- tem in 1-2 GHz Band. IEEE Trans. Geosci. Electron., v. GE-16, No. 2, 1978, pp. 126-133. 47. Ellerbruch, D. A. , and J. W. Adams . Microwave Measurement of Coal Layer Thickness (contract SO144086, NBS). BuMines OFR 101-75, 1974, 33 pp; NTIS COM 74-11643. 39. Coon, J, B, , J, C. Fowler, and C, J. Schafers, Experimental Uses of Short-Pulse Radar in Coal Seams. Geo- physics, V. 46, No. 8, 1981, pp. 1163- 1168. 40. Cooper, D. W. , R. A. Muller, and R. J. Schertler, Remote Profiling of Lake Ice Using an S-Band Short-Pulse Ra- dar Aboard and All-Terrain Vehicle. Ra- dio Sci., V. 11, No. 4, 1976, pp. 375- 381. 41. Curtis, W. L. , and R. J. Coe. A Microwave Technique for the Measurement of the Dielectric Properties of Soils. IEEE Trans. Microwave Theory and Tech. , V. MTT-11, No. 3, 1963, pp. 211-212. 42. Dolphin, L. T. , W. B. Beatty, and J. D. Tunzi. Radar Probing of Victorio Peak, New Mexico. Geophysics, v. 43, No. 7, 1978, pp. 1441-1448. 43. Dolphin, L. T. , R. L. Bollen, and G. N, Oetzel. An Underground 48. Ellerbruch, D, A,, and D, A, Bel- sher. Electromagnetic Technique of Mea- suring Coal Layer Thickness, IEEE Trans, Geosci, Electron,, v, GE-15, No. 2, 1977, pp. 126-133. 49. . FM-CW Electromagnetic Technique of Measuring Coal Layer Thick- ness. NBS, NBSIR 76-840, May 1976, 67 pp. 50. England, A. W. , and G. R. Johnson, Microwave Brightness Spectra of Layered Media, Geophysics, v, 42, No, 3, 1977, pp. 514-521. 51. ENSCO, Inc. Mine Roof Stratigra- phy Using Electromagnetic Radar (contract J0366076). U.S. Dep. Energy, DOE FE- 9141-1, June 1978, 221 pp. 52. Evans, S. Progress Report on Ra- dio Echo Sounding, Polar Rec, v, 11, 1967, pp. 413-420. 19 53. Evans, S. Radio Techniques for the Measurement of Ice Thickness. Polar Rec, V. 3, 1963, pp. 406-410. 54. Fitz, D. C. Application of Remote Sensing in Hydrology. M.S. Thesis, Colo. State Univ., Ft. Collins, Colo., 1980, 131 pp.; NTIS PB 81-111932. 55. Forbes, L. M. (ed.). Radio Echo Exploration of the Antarctic Ice Sheet, 1967. Polar Rec, v. 14, 1968, pp. 211- 213. 56. Fountain, L. S., F. X. Herzid, and T. E. Owen. Detection of Subsurface Cav- ities by Surface Remote Sensing Tech- niques, Fed. Highway Administration, Rep. FHWA-RD-75-8, June 1975, 81 pp.; NTIS PB 253-379. 57. Fowler, J. C. Subsurface Reflec- tion Profiling Using Ground-Probing Ra- dar. Min, Eng. (Littleton, Colo.), v. 33, No. 8, 1981, pp. 1266-1270. 58. Fowler, J. C, S. D. Hale, and R. T. Houck. Coal Mine Hazard Detection Using Synthetic Pulse Radar (contract H0292025, ENSCO, Inc.). BuMines OFR 79- 81, 1981, 86 pp.; NTIS PB 81-224412. Dissipation Factors for Six Rock Types Between 20 and 100 Megahertz. BuMines RI 6913, 1967, 21 pp. 64. Hancock, J. C. , and P. A. Wintz. Signal Detection Theory. McGraw-Hill Book Co., New York, 1966, 247 pp. 65. Harrison, C. H. Reconstruction of Subglacial Relief From Radio Echo Sound- ing Records. Geophysics, v. 35, No. 6, 1970, pp. 1099-1115. 66. Heacock, J. G. (ed.). The Struc- ture and Physical Properties of the Earth's Crust. Am. Geophys. Union, Wash- ington, D.C., 1971, 284 pp. 67. Helstrom, C. W. Statistical The- ory of Signal Detection. Pergamon Press, Inc., New York, 1968, 470 pp. 68. Hipp, J. E. Soil Electromagnetic Parameters as Functions of Frequency, Soil Density, and Soil Moisture. Proc. IEEE, V. 61, No. 1, 1974, pp. 98-102. 69. Hluchanek, J. A. Radar Investiga- tion of the Hockley Salt Dome. M.S. The- sis, Texas A. & M. Univ., College Sta- tion, Tex., 1973, 149 pp. 59. Frischknecht, F. C. Fields About an Oscillating Dipole. Q. Colo. Sch. Mines, v. 62, No. 1, 1967, p. 326. 60. Ghose, R. N. Mutual Couplings Among Subsurface Antenna Array Elements. IEEE Trans. Antennas and Propag. , v. AP- 11, No. 3, 1963, pp. 257-261. 61. Glenn, W. E. , J. Ryu, S. H. Ward, W. J. Peeples, and L. L. Phillips. The Inversion of Vertical Magnetic Dipole Sounding Data. Geophysics, v. 38, No. 6, 1973, pp. 1109-1129. 62. Grant, F. S., and G. F. West. In- terpretation Theory in Applied Geophys- ics. McGraw-Hill Book Co., New York, 1965, 584 pp. 63. Griffin, R. E., and R. L. Marovelli. Dielectric Constants and 70. Hoekstra, P., and A. Delaney. The Dielectric Properties of Soils at UHF and Microwave Frequencies. Program and Ab- stracts of Papers Presented at Meeting of URSI-Int. Union Radio Sci., Group on An- tennas and Propagation, Boulder, Colo., August 1973, pp. 112-130. 71. Holser, W. T. , J. J. Brown, 0. A. Fredrickson, and R. R. Unterberger. Ra- dar Logging of a Salt Dome. Geophysics, v. 37, No. 7, 1972, pp. 889-906. 72. Ilsley, L. C, H. B. Freeman, and D. N. Zellers. Experiments in Under- ground Communications Through Earth Strata. BuMines TP 433, 1928, 60 pp. 73. Jakosky, J. J. Radio as a Method for Underground Communications in Mines. BuMines RI 2599, 1924, 4 pp. 20 74. Jakosky, J. J., and D. H. Zellers. Factors Retarding Transmission of Radio Signals Underground and Some Further Ex- periments and Conclusions. BuMines RI 2651, 1924, 10 pp. 75. Jesch, R. L. High Resolution Sensing Techniques for Slopability. Fed. Highway Administration, Rep. FHWA-RD-79- 32, January 1979, 47 pp. 84. Luchinonov, U. S. Radar Sensing of Mountain Glaciers. Sov. Phys. — Tech. Phys. (Engl. Transl.), v. 18, 1973, p. 415. 85. Lytle, R. J. Measurement of Earth Medium Electrical Characteristics; Tech- niques, Results, and Applications. IEEE Trans. Geosci. Electron., v. GE-12, No. 3, 1974, pp. 81-101. 76. Joyce, J, W. Electromagnetic Ab- sorption by Rocks With Some Experimental Observations Taken at Mammoth Cave in Kentucky. BuMines TP 497, 1931, 28 pp. 86. Lytle, R. J., and D. L. Lager. The Yosemite Experiments: HF Propagation Through Rock. Radio Sci. , v. 11, No. 4, 1976, pp. 245-252. 77. Kerr, J. L. Short Axial Length Broadband Horns. Abstracts, 22nd Ann. Symp., USAF Antenna Research and Devel- opment, Univ. 111., Oct. 11-13, 1972, pp. 14. 87. Lytle, R. J., E. F. Laine, D. L. Lager, and D. T. Davis. Cross-Borehole Electromagnetic Probing to Locate High- Contrast Anomalies. Geophysics, v. 44, No. 4, 1979, pp. 1667-1676. 78. Khalafalla, A. S. , and J. M. Viner. Rock Dielectrometry at Mega- and Giga-Hertz Frequencies. Program and Ab- stracts of Papers Presented at Meeting of URSI-Int. Union Radio Sci. , Group on An- tennas and Propagation, Boulder, Colo. , August 1973, pp. 146-163. 79. Kraichmann, M. B. Handbook of Electromagnetic Propagation in Conducting Media. U.S. Dept. of the Navy, Document NAVMAT P. 2302, 1970, 1128 pp. 88. Lytle, R. J., E. F. Laine, D. L. Lager, and J. T. Okada, The Lisbourne Experiments: HF Propagation Through Per- mafrost Rock. Lawrence Livermore Lab. , Livermore, Calif., Rep. UCRL-51474, 1973, 147 pp. 89. Lytle, R. J., E. K. Miller, and D. L. Lager. A Physical Explanation of Electromagnetic Surface Wave Formulas. Radio Sci., v. 11, No. 4, 1976, pp. 235- 243. 80. Kuo, J. T. , and Dong-heng Cho. Transient Time-Domain Electromagnetics. Geophysics, v. 45, No. 2, 1980, pp. 271- 291. 81. Laine, E. F. Detection of Water- Filled and Air-Filled Underground Cavi- ties (BuMines contract H0202007, Lawrence Livermore Lab.). Lawrence Livermore Lab., Livermore, Calif., Rep. UCRL-53127, December 1980, 15 pp. 82. Lee, K. H. , D. F. Pridmore, and H. F. Morrison. A Hybrid Three- Dimensional Electromagnetic Modeling Scheme. Geophysics, v. 46, No. 6, 1981, pp. 796-805. 91. McGehan, F. P. Radar to the Res- cue. Dimensions/NBS, v. 63, No. 9, 1979, pp. 5-9. 90. Meakins , B. J. Mechanisms of Di- electric Absorption in Solids. Ch. in Progress in Dielectrics-3. John Wiley & Sons, Inc., New York, 1961, 486 pp. 92. Mofatt, D. L. Electromagnetic Pulse Sounding for Geological Surveying With Application in Rock Mechanics and Rapid Excavation Program (BuMines/ARPA (DOD) contract H02 10042, Ohio State Univ.). Electroscience Lab., Ohio State Univ., Columbus, Ohio, Semiannual Tech, Rep. 3190-1, October 1971, 79 pp. 83. Lerner, R. M. Ground Radar Sys- tem. U.S. Pat. 3,831,173, Aug. 20, 1974. 21 93. Mofatt, D. L. Subsurface Appli- cation of Periodic Electromagnetic Video Pulse Signal, Paper in Thru-the-Earth Electromagnetic Workshop (Golden, Colo. , December 1973) , ed. by R. G. Geyer (con- tract GOl 33023, Colo. Sch. Mines). Bu- Mines OFR 16-74, 1974, pp. 112-117; NTIS PB 231 154. 94. Subsurface Video Pulse Radar. Proc. Eng. Foundation Conf., Sub- surface Exploration for Underground Exca- vation and Heavy Construction, Henniker, N.H. , Aug. 11-16, 1974. Am. Soc. Civil Eng., New York, 1974, pp. 195-212. 95. Moffatt, D. L. , and R. K. Mains. Detection and Discrimination of Radar Targets. IEEE Trans. Antennas and Prop- ag., V. AP-23, No. 3, 1975, pp. 358-367. 102. Page, D. F., G. 0. Venier, and F. R. Cross. Snow and Ice Depth Measure- ments by High Range Resolution Radar. Can. Aeronaut, and Space J., v. 19, No. 10, 1973, pp. 531-533. 103. Parkhomenko, E. I. Electrical Properties of Rocks. Plenum Press, New York, 1967, 314 pp. 104. Peters, L. , Jr. Electromagnetic Transient Underground Radar (ETUR) for Geophysical Exploration. Proc. IEEE An- tennas and Propag. Soc. Int. Symp., Am- herst, Mass., Oct. 10-15, 1976, APSIS 786, pp. 203-204. 105. Porcello, L. J. The Apollo Lunar Sounder Radar System. Proc. IEEE, v. 62, No. 6, 1974, pp. 769-783. 96. Moffatt, D. L. , and R. J. Puskar. A Subsurface Electromagnetic Pulse Radar. Geophysics, v. 41, No. 3, 1976, pp. 506- 518. 97. Moore, D. F., and E. A. Quincy. Bayes Classification of Subsurface Elec- tromagnetic Responses. Radio Sci. , v. 11, No. 4, 1976, pp. 395-403. 98. Moore, J. R. , J. D. Echard, and C. G. Neill. Radar Detection of Voids Under Concrete Highways. Proc. IEEE Int. Radar Conf., Washington, D.C., April 28- 30, 1980, pp. 130-135. 99. Morey, R. M. Application of Downward Looking Impulse Radar. Proc. 13th Annu. Can. Hydrographic Conf. , Cana- da Centre for Inland Waters, Burlington, Ontario, March 1974, pp. 83-99. 100. Continuous Subsurface Profiling by Impulse Radar. Proc. Eng. Foundation Conf., Subsurface Exploration for Underground Excavation and Heavy Con- struction, Henniker, N.H. , Aug. 11-16, 1974. Am. Soc. Civil Eng., New York, 1974, pp. 212-232. 101. Morey, R. M. , and W. S. Harring- ton, Jr. Feasibility Study of Electro- magnetic Subsurface Profiling. U.S. EPA, EPA R2-72-082, October 1972, 28 pp. 106. Robin, G. de Q. , S. Evans, and J. T. Bailey. Interpretation of Radio Echo Sounding in Polar Ice Sheets. Philos. Trans. R. Soc. London, Ser. A, v. 265, 1969, pp. 437-505. 107. Robinson, L. A., W. B. Weir, and L. Young, Location and Recognition of Discontinuities in Dielectric Media Using Synthetic RF Pulses. Proc. IEEE, v. 62, No. 1, 1974, pp. 36-43. 108. Rollwitz, W, L, , and J, D, King. Coal Thickness Gauge Using RRAS Tech- niques. (Natl. Acad. Sci. contract NAS 8-32606, Southwest Res. Inst.). Final Rep., pt. 1, SWRI Project 15-4967, June 10, 1978, 55 pp. 109. Rubin, L. A., and J. C. Fowler. Ground Probing Radar for Delineation of Rock Features. Eng. Geol. (Amsterdam), V. 12, No. 2, 1978, pp. 163-170. 110. Saint-Amant, M. , and D. W. Strangway. Dielectric Properties of Dry Geologic Materials. Geophysics, v. 35, No. 4, 1970, pp. 624-645. 111. Sinha, A. K. , and P. K. Bhatta- charya. Vertical Magnetic Dipole Buried Inside a Homogenous Earth. Radio Sci., V. 1, No. 3, 1966, pp. 379-395. 22 112. Slichter, L. B. and Theoretical Response Ing Spheres. Trans. AIME, pp. 443-459. The Observed of Conduct- V. 97, 1932, 121. Unterberger, R. R. Looking Through Rock. With Radar. Mln. Congr. J., V. 63, No. 6, 1977, pp. 38-41. 113. Smith, B. D. , and S. H. Ward. On the Computation of Polarization Ellipse Parameters. Geophysics, v. 39, No. 6, 1974, pp. 867-869. 114. Stewart, R. D. Radar Investiga- tion of the Cote Blanche Salt Dome. M.S. Thesis, Texas A. & M. Univ., College Sta- tion, Tex., 1974, 347 pp. 115. Stewart, R. D. , and R. R. Unter- berger. Seeing Through Rock Salt With Radar. Geophysics, v. 41, No. 1, 1976, pp. 123-132. 116. Suhler, S. A., and T. E. Owen. Field Demonstration of Deep-Penetrating Geophysical Techniques (contract H0212018, Southwest Res. Inst.). Draft final rep., January 1983, 59 pp.; avail- able from Denver Research Center, U.S. Bureau of Mines, Denver, Colo. 117. Suhler, S. A., T. E. Owen, J. E. Hipp, and W. R. Peters. Development of a Deep Penetrating Borehole Geophysical Technique for Predicting Hazards Ahead of Coal Mining (contract H0252033, Southwest Res. Inst.). BuMlnes OFR 77-80, 1978, 124 pp.; NTIS PB 80-208614. 118. Tarn, K. Dielectric Property Mea- surements of Rocks In the VHF-UHF Region. Ph.D. Thesis, Texas A. & M. Univ., Col- lege Station, Tex., 1974, 416 pp. 119. Tarantolo, P. J., Jr., and R. R. Unterberger. Radar Detection of Bore- holes In Advance of Mining (Rock Salt). Geophys. Prospect., v. 26, No. 2, 1978, pp. 359-382. 120. Trompeter, J. Where No Man Has Gone Before: A Report of Microwave Tech- niques In Subsurface Investigation. Con- str. Specifier, November 1980, pp. 38-47. 122. Salt. Geophys. Prospect., Radar Propagation In Rock V. 26, No. 2, 1978, pp. 312-328. 123. U.S. Bureau of Mines. Annual Re- port of the Director of the Bureau of Mines to the Secretary of the Interior, 1924, p. 16. 124. Von Hlppel, A. R. Dielectrics and Waves. MIT Press, Cambridge, Mass., 1954, 284 pp. 125. Walt, J. R. A Conducting Perme- able Sphere In the Presence of a Coll Carrying an Oscillating Current. Can. J, Phys., V. 31, No. 4, 1953, pp. 670-678. 126. . A Conducting Sphere In a Time-Varying Magnetic Field. Geophysics, V. 16, No. 4, 1951, pp. 666-672. 127. . Electromagnetic Waves in Stratified Matter. Pergamon Press, Inc., Oxford, England, 1962, 372 pp. 128. Ing Magnetic Induction by an Osclllat- Dlpole Over a Two-Layer Appl. Scl. Res., Sec. B, v. 7, Ground. 1959, pp. 73-80. 129. . Mutual Coupling of Loops Over a Homogeneous Ground. Geophysics, V. 20, No. 3, 1955, pp. 630-637. 130. . Mutual Coupling of Wire Loops Lying on a Homogeneous Ground. Geophysics, v. 19, No. 2, 1954, pp. 290- 296. 131. . Mutual Electromagnetic Coupling of Loops Over a Homogeneous Ground-An Additional Note. Geophysics, V. 21, No. 2, 1956, pp. 479-484. 23 132. Walt, J. R. Note on the Theory of Transmission of Electromagnetic Waves in a Coal Seam. Radio Sci., v. 11, No. A, 1976, pp. 263-265. 133. On Anomalous Propagation of Radio Waves in Earth Strata. Geophys- ics, V. 19, No. 2, 1954, pp. 342-343. 134. Propagation of Electro- magnetic Pulses in a Homogeneous Conduct- ing Earth. Appl, Sci. Res., Sec. B. , v. 8, No. 7, 1960, pp. 213-253. 135. The Radiation Fields of a Horizontal Dipole in a Semi-Infinite Dissipative Medium. J. Appl. Phys., v. 24, No. 3, 1953, pp. 958-959. 136. Radiation From a Small Loop Immersed in a Semi-Infinite Conduct- ing Medium. Can. J. Phys., v. 37, No. 3, 1959, 672-674. 137. Transmission and Reflec- tion of Electromagnetic Waves in the Presence of Stratified Media. J. Res. NBS, V. 61, September 1958, pp. 205-232. 138. Wait, J. R. (ed.). Electromag- netic Probing in Geophysics. The Golem Press, Boulder, Colo., 1971, 391 pp. 139. Wait, J. R. , D. Chang, R. J. Ly- tle, M. A. Ralston, K. R. Umashankar, and R. C. Wittman. Analyticas Bases for Electromagnetic Sensing of Coal Proper- ties. U.S. Dep. Energy, Rep. DOE FF 8972-1, July 1978, 34 pp. 140. Ward, S. H. Electrical, Elec- tromagnetic and Magnetotelluric Meth- ods. Geophysics, v. 45, No. 11, 1980, pp. 1659-1660. 141. Ward, S. H. Electromagnetic The- ory for Geophysical Applications. Min. Geophys., v. 2, 1967, pp. 10-197. 142. Ward, S. H. , and D. C. Fraser. Conduction of Electricity in Rocks. Ch. in Mining Geophysics, Soc. of Explor. Geophys., Tulsa, Okla. , v. 2, 1970, 36 pp. 143. Ward, S. H. , and H. F. Morrison (eds.). Special Issue on Electromagnetic Scattering. Geophysics, v. 36, No. 1, 1971, pp. 1-258. 144. Watt, A. D. , F. S. Mathews, and E. L. Maxwell. Some Electrical Charac- teristics of the Earth's Crust, Proc. IEEE, v. 51, No. 6, 1963, pp. 897-910. 145. Wu, T. T., and R. W. P. King. The Cylindrical Antenna With Nonreflect- ing Resistive Loading. IEEE Trans. An- tennas and Propag. , v. AP-13, No. 5, 1965, pp. 369-373. 146. XADAR Corp. Coal Mine Hazard De- tecting Using Synthetic Pulse Radar (con- tract H0212016). Draft final rep., De- cember 1982, 71 pp.; available from Denver Research Center, U.S. Bureau of Mines, Denver, Colo. 147. Yatsyshin, V. I., N. I. Zhuk, K. M. Salamatov, and G. E. Yakovitskaya. Methods of Measuring the True and Effec- tive Electrical Conductivities of Rocks in Coal Mines. Sov. Min. Sci. (Engl. Transl.), v. 7, No. 7, 1971, pp. 453-456. 148. Young, J. D. Radar Imaging From Ramp Response Signatures. IEEE Trans. Antennas and Propag., v. AP-24, No. 3, May 1976, pp. 276-282. INT.-BU.OF MINES, PGH., PA. 27296 %.** v" .^^.o«:* ^ ^Pv^, 4* vv -oV^^ . **^\ 3 . .<^/7/>vv>2, .Q V^ o«». "^o ' -<'^'' . ♦ s * V^ "'■^ ' ' -4 V . " O .0' .L^C.'* > ^ ►'- .-^^^ 'V V.** .'I«SM. v./ /^-V %.<.^ -•^SfA-. ^<51-...^* '>( ■*.,.* v-^ %- A:^' ^. ^^^ .<5U^„ .*\.-^;^-,-*.. .^°'.^tX /^«.Xo^ ■ <^^^ a5°X. -•'/' V^^\/' V^"''/' "V"^^-'<^*'' %"'^V - -^v^^^^ 4^ % 0'"" oOJL"''^ "*o-, .-^ •■•' -6 v/^^ ^ *% WERT BCXDKBINDINC MIODLETOWN PA JUNE 84 ^^^'V. °.: ^V %.** .'M&". ■\/ .•^\ %,** ,'^", %/ .-J •ono" x^^ ^^^vP. * *7 ".ii, « .^ -^^ lA^Ls iVvrv --- ' .,^^l^ LIBRARY OF CONGRESS II 1 002 955 8 9A li^ «t 4 <