application of seismic refraction methods in groundwater...

$: Application of seismic refraction methods in groundwater ...pioneer.netserv.chula.ac.th/~tthanop/reading paper/Refraction 1.pdf · Seismic refraction methods have been extensively$
GEOPHYSICS, VOL. 51. NO. 2 (FEBRUARY 1986); P. 236-249, 12 FIGS., 2 TABLES.

Application of seismic refraction methods in groundwater modeling studies in New England

F. P. Haeni*

ABSTRACT

Seismic refraction methods have been extensively used in petroleum, mineral, and engineering investigations, and to some extent for hydrologic applications, during the past 30 years. Recent advances in equipment, sound sources, and computer interpretation techniques make seismic refraction methods highly effective and economical for obtaining data for groundwater modeling studies. Aquifers that can be defined by one or more seismic velocity surfaces, such as alluvial or glacial deposits in consolidated rock valleys, limestone, or sandstone underlain by metamorphic or igneous rock, or saturated unconsolidated deposits overlain by unsaturated unconsolidated deposits, are ideally suited for seismic refraction methods. Seismic refraction allows economical collection of data for one or more model input parameters and provides the basis for efficient collection of subsurface data by test drilling or aquifer tests.

Seismic refraction studies were conducted over unconfined glacial aquifers in New England to determine: (1) depth of the underlying bedrock; (2) depth of the water table; (3) saturated thickness of the aquifer in areas not acc&hlc to heavy drilling equipment; (4) areas where thick, unsaturated sediments overlie thickly saturated parts of the aquifer; and (5) locations of test holes and type of drilling equipment needed. These data were used in groundwater models and provided parameter values that required few adjustments during calibration.

INTRODUCTION

Seismic refraction methods have been extensively used in petroleum, mineral, and engineering investigations and to some extent for hydrologic applications, during the past 30 years. Recent advances in equipment, sound sources, and computer interpretation techniques make seismic refraction methods highly effective and economical for obtaining data for groundwater modeling studies. Seismic refraction should be

considered in the planning stages of hydrologic studies and used as a tool by the hydrologist to solve problems associated with pump tests, simulation models, test holes, geologic maps, borehole geophysical techniques, etc.

Classically, surface geophysical techniques have been used early in the exploration process. Hydrologists, following this example, should use these techniques early in their study, prior to using more expensive data-collection techniques such as drilling or aquifer testing. Such use of seismic refraction techniques will minimize expensive data-collection activities and will result in higher quality, more efficient hydrologic modeling studies.

To utilize the technique effectively, hydrologists must un- derstand its principles, limitations, advantages, equipment, field procedures, interpretation procedures, and potential uses.

SEISMIC REFRACTION: THEORY AND LIMITATIONS

Many textbooks and numerous journal articles present the details of seismic refraction theory (Dobrin, 1976; Mooney, 1984; Slotnick, 1959; Musgrave, 1967; Telford, 1976; Parasnis, 1979; Grant and West, 1965). The following dis- cussion only reviews the very basic principles and limitations of the seismic refraction method.

The foundation of seismic refraction theory is Snell’s law, which governs the refraction of a sound or light ray across the boundary between layers of different physical properties. As sound waves travel from a medium of low seismic velocity into a medium of higher seismic velocity, some are refracted toward the lower velocity medium, and some are reflected back into the first medium. As the angle of incidence of the sound ray approaches the critical angle (an angle where the refracted ray grazes the surface of the contact between the two media), most of the compressional energy is transmitted along the surface of the second layer, at the velocity of sound in the second layer. As this energy propagates along the surface, it generates new sound waves in the upper medium (Huygens’ principle: every point on an advancing wavefront can be re- garded as the source of a new sound wave) that in turn propa- gate back to the surface at the critical angle and at the seismic velocity of layer one. For seismic refraction to work, therefore,

Manuscript received by the Editor May 10, 1984; revised manuscript received June 27, 1985 *US. Geological Survey, 450 Main Street, Rm. 525, Hartford, CT 06103. This paper was prepared by an agency of the US. government.

236

Groundwater Modeling Studies

the velocity of sound in each deeper layer must be greater than in the layer above it. When this condition is met, the refracted wave arrives at the Earth’s surface where it can be detected by a geophone which generates an electrical signal and sends the signal to a seismograph.

From a series of geophones placed on the ground, the seismic arrival time versus the shot-to-detector distances (Figure 1) can be plotted to give a time-distance curve. Examples of time-distance curves for a wide variety of geologic sections are

Ax (ft)

“l=mO’

where

and

v1 is the velocity of sound in layer one,

Ax is the change in distance,

237

shown in Zohdy et al. (1974) and Mooney (1984). Figure 1 shows that at distances less than the crossover

AJJ is the change in time

distance, the sound has traveled directly from the sound At the crossover distance and beyond, the sound wave that source to the detectors. Because this compressional wave has has traveled through layer one, then along the interface of the traveled a known distance in a known time the velocity of high-speed layer (layer two), then back up to the surface layer one can easily be calculated. In the time-distance curve, through layer one, arrives before the wave that has been in the u, is equal to inverse slope of the plotted line, or slow layer all the time All first compressional-wave arrivals at

600

400

?jj 300 E

g g 200

100

0

I-

4

L

INTXXCXPT time

TRAVELTIME CURVE

CXOEEOVBP Dl8TANCX

0 20 40 60 80 100 120 140 160 160 200

SHOTPOINT

DISTANCE (m) TIME-DISTANCE CURVE

c GEOPHONX

V1 :BATUUATXD ALLUVIUY

VELOCITY 1.6 km11

GEOLOGIC CROSS-SECTION

FIG. 1. Diagrammatic geologic cross-section and resulting time-distance curve.

238 Haeni

geophones with distances larger than the crossover distance will therefore be refracted waves from the high-speed layer. When these points are plotted on the time-distance curve, the inverse slope of this segment equals the apparent velocity of layer two, which is the true velocity of layer two if it is parallel to the land surface. To determine the true velocity of layer two when it is dipping, the shotpoint will have to be placed at the opposite end of the geophone line, a reversed time-distance curve will have to be plotted, and the true velocity will have to be calculated.

The intercept time and the crossover distance (Figure 1) are dependent upon the velocity of the two materials and the thickness of the first layer and can be used to determine the thickness of the first layer. Analytical and graphical solutions for calculating the depth to individual refractors under various conditions (i.e., sloping boundaries, multiple layers, etc.) are presented in many texts (e.g., Dobrin, 1976, and Mooney, 1984).

Limitations of the seismic refraction method

Two special cases that ultimately affect the success or failure of seismic refraction investigations are: (1) blind zone problems. i.e., thin intermediate seismic velocity refractors; and (2) slow seismic velocity layers underlying high seismic velocity layers.

Thin intermediate seismic velocity layer problem

One of the most serious limitations of the seismic refraction method is the possible existence of a geologic layer that cannot be detected (Soske, 1954; Sander, 1978). This problem occurs when an intermediate layer in a geologic section meets the criterion that the velocity of sound increases with depth but has an insufficient velocity contrast or thickness in order to return first arrival energy. This problem is especially critical in water-resources investigations, because the intermediate

layer may be the zone of interest in the study, i.e., saturated unconsolidated material located between unsaturated unconsolidated material and bedrock. This layer cannot be defined by any alternative field array of geophones or shot- points. If the presence of this layer is suspected, however, cal- culations can be made to determine the minimum and maximum possible thickness of this unit (Mooney, 1984; Redpath, 1973; Sander, 1978). Table 1 shows the maximum thickness of this undetected layer under varying geologic conditions.

Hidden-layer or velocity reversal problem

In some hydrologic settings, the velocity of sound in the Earth’s layers does not increase with depth and low seismic velocity units underlie high seismic velocity units. An example of this is an unconsolidated sand and gravel aquifer underlying compact glacial tills. In this case, the low-velocity unit will not be detected by the refraction technique and the calculated depth of the deeper unit will be in error. Snell’s law explains the problem-the sound wave will be refracted toward the low-velocity medium-and in this case, the refracted wave would not be detected at the surface until it encountered a layer with a velocity higher than any layer previously encountered.

HYDROLOGIC USES OF SEISMIC REFRACTION TECHNIQUES IN NEW ENGLAND

The hydrologic setting of the major glacial aquifers in New England consists of unconsolidated glacial deposits underlain by consolidated bedrock. The velocities of sound in and the general stratigraphic relationship of the various hydrologic units in New England are shown in Table 2.

It is evident from Table 2 that several important hydrologic boundaries of unconsolidated glacial aquifers can be defined by large velocity changes and that the velocity of the sediments usually increases with depth. In this setting then, seis-

Table 1. Maximum thickness of an undetectable hidden layer in various hydrologic settings.

Maximum thickness of undetected

Hvdrolouic setting and Gelocity of sound in the geologic units

Thickness aquifer Range in of material in denth of

layer 1 (m)

layer 2 (m)

layer 3 (m)

Dry sand, v, = .45 km/s 3.0 2.4 3.7-5.5 Saturated sand aquifer, 6.1 4.9 7.3-l 1 .o

t’* = 1.52 km/s 15.2 12.5 18.6-27.7 Bedrock, u3 = 4.57 km/s 30.5 25.0 37.5-55.5

Till, t’i = 2.13 km/s 3.0 1.0 3.44.0 Sedimentary rock aquifer, 6.1 2.1 6.7-7.9

c2 = 3.96 km/s 15.2 5.2 16.8-20.4 Crystalline rock, 30.5 10.1 33.540.5

a3 = 4.57 km/s 61.0 20.4 66.8-8 1.4

Saturated sand and gravel, u. = 1.52 km/s 3.0 2.2 3.74.9

Limestone aquifer, 6.1 3.7 7.339.8 c2 = 3.05 km/s 15.2 8.8 18.6-24.0

Crystalline rock, 30.5 17.7 37.548.2 us = 4.57 km/s 61.0 35.1 74.7-96.0

Groundwater Modeling Studies 239

Table 2. Velocity of sound and general stratigraphic relationship of hydrologic units in New England.

Velocity of sound (km/s)

Unconsolidated stratified drift or alluvial deposits

Unsaturated Saturated

Saturated till (lodgement) Saturated Triassic sedimentary rocks Saturated crystalline rocks

.3-.6 1.2-1.8 2.1-2.4 3.44.3 4.6-6.1

mic refraction techniques can be a valuable exploration tool for the hydrologist. The technique can be used to: (1) determine the depth to the water table in unconsolidated materials (provided the saturated section is thicker than the unsaturated section); (2) determine the depth to and thickness of the lodgement till (provided it has a significant thickness and velocity contrast); (3) determine the depth to bedrock and general bedrock type; (4) plan an economical drilling program; and (5) obtain data in areas not easily accessible to drilling rigs.

EQUIPMENT

A block diagram of a typical seismic refraction system is shown in Figure 2. A wide variety of seismographs are available from different manufacturers, ranging from simple, inex- pensive single-channel units to very sophisticated, expensive multichannel units used in the petroleum industry. Almost all of the modern units record the data digitally and are compat- ible with digital computers. The type of equipment best suited for water-resources studies is somewhere in the middle of this field and is typically a 12- or 24-channel signal enhancement seismograph. These units are capable of utilizing a nonexplosive sound source because they can add the signals from several successive nonexplosive impacts to increase the amplitude of the refracted signal more than that of the random noise.

Geophones convert the physical movement of the ground to an electrical signal. In seismic refraction work, low-frequency (8-10 Hz) vertical motion geophones are used. They are at- tached to the geophone spread cable at measured intervals, and are planted for optimum physical contact between the geophones and the ground.

Geophone cables come in a variety of lengths with a prede- termined distance between each geophone connection. For water-resources studies, cables that have approximately 8, 15, or 30 m spacing between geophones are normally used.

Many types of seismic sources are available for refraction seismographs (Mooney, 1980; Beggs and Garriot, 1979).

Despite the obvious disadvantages of storage and safety, explosives are ideally suited as sound sources for refraction work. No other sources can provide the proper amount of energy under all field conditions. An alternative to the exclu- sive use of explosives, however, is either a mechanical or electrical source for the majority of the field work and selective use of explosives in areas where greater energy is needed.

For hydrologic investigations expiosives will generaily be needed under the following conditions.

(I) Deep refraction studies, requiring very long geophone lines (depth to deepest refractor 60 m or greater.)

(2) Thick unsaturated sections, especially in loose materials (unsaturated material thicker than 10 to 12 m.)

Advances in the explosive industry have virtually eliminated dynamite as a seismic source. Two-component explosives, consisting of a flammable liquid and a dry powder that are relatively safe to handle because they do not form an explosive until mixed, have replaced dynamite (Figure 3). Seismic blast- ing caps are still needed, however, to detonate the mixed explosive.

A hammer and striker plate are commonly used as a sound source for very shallow investigations. Best results are obtained when the striker plate is placed on firm ground and the resulting signal is stacked 5 to 15 times on the seismograph. A sledgehammer as seismic source is shown in Figure 4.

------7 I I ] Optional I , Digital 1

Sound Source Initiator

*

I Geophone Extension Cable _

I GROIJNL) SURFACE

Direct wave

Layer 1

- -

Refracted wave

Layer 2

FIG. 2. Block diagram ofa seismic refraction system

240 Haeni

Weight drops, shotguns, and land sparker systems all have an intermediate energy level, having two to five times the energy of a sledgehammer, but significantly less energy than explosives.

Many different types of field vehicles can be used for seismic refraction work. Figure 5 shows a pickup truck set up for seismic field work. Since explosives will most likely be used during a study, the vehicle is equipped with a small drill rig to bury the explosive charge.

INTERPRETATION

During the field operation, seismograph records and tapes are obtained that record the time the sound source is initiated and the time of the ground movement as the waves arrive at each geophone. In seismic refraction work, only the first arrival of energy at the geophone (the compressional sound wave) is utilized. Upon completion of data acquisition in the field, the interpretation phase is begun.

The first step in interpretation is to determine the time from initiation of the sound source to the first arrival of energy, for each geophone. When there is sufficient energy in the sound source, and ambient noise is minimal, the first breaks are sharp and this procedure is straightforward. When ambient noise exists (such as operation of heavy equipment or highway

traffic) and/or mechanical sound sources are used, picking first-arrival times can become difficult.

The record in Figure 6 was obtained in Maine during a modeling study of the Little Androscoggin River valley (Mor- rissey, 1983) and shows easily discernible first-arrival times and the apparent velocity of two important hydrologic units. The first six geophones form a line (A-B) that has an inverse slope of approximately 1.5 km/s. This velocity is characteristic of the speed of sound in saturated, unconsolidated sediments in New England. Geophones 7 through 12 form a line (B-C) that has an inverse slope value of 5.2 km/s, a seismic velocity characteristic of crystalline bedrock. A thin layer of glacial till is known to be present in the area, but it was not detected on the time-distance plot since it is a thin, intermediate velocity layer. The thickness of this unit varies from 0 to 3 m; its velocity of sound is approximately 2.3 km/s.

Since seismic refraction techniques have been used for a long time many interpretation schemes (Dobrin, 1976, 31% 331) and countless modifications to these schemes have been developed and published in the literature (Musgrave, 1967, 5655594). Formulas, nomographs, and most recently, computer programs, are available for almost every type of field problem. Each of these techniques has its strong points, and when properly selected and intelligently applied they usually give satisfactory results.

FIG. 3. Mixing two component explosives in the field.

Groundwater Yodeling Studies 241

One problem inherent in all geophysical studies is the ambi- near each shotpoint. The delay-time interpretation technique, guity of any particular geologic model to a set of field data. on the other hand, provides a depth to the refracting layer for The source of the problem is that geophysical techniques mea- each geophone that receives a refracted wave from that layer. sure physical properties of the Earth remotely (i.e., from the For example, a standard approach for mapping the water surface), and several different combinations of Earth materials table is to use one shot on each end of a 1Zgeophone spread. can give the same signal at the surface. The solution is appli- A standard formula interpretation would provide one depth to cation of the interpreter’s geologic or hydrologic knowledge of water near each shotpoint. The computer modeling program the study area. Successful interpretation of seismic refraction provides up to 12 depths to the water table for each shot records depends upon the hydrologist’s being the interpreter along the same spread. Consequently, much more information or at least being directly involved in the interpretation pro- on the subsurface can be obtained from the same amount of cess. Failure to involve the hydrologist inevitably leads to field work. In addition, the earth model obtained from the poor results that are usually blamed on the seismic refraction delay-time process is refined by iterative ray tracing that com- method. Success of a seismic refraction study is much more pares the sound wave traveltimes measured in the field to dependent upon the interpreter than on the specific interpreta- calculated traveltimes through the model, and adjusts the tion scheme used. earth model to minimize any differences.

The U.S. Geological Survey utilizes a seismic refraction inverse modeling computer program (Scott et al., 1972; Scott, 1977) based on the delay-time method described in Pakiser and Black (1957) and a ray-tracing modeling technique. A sign&ant advantage of this interpretation technique over the standard formula interpretations is that it uses the data collected at each geophone for interpretation. In a simple two- layer case, a standard interpretation utilizing text formulas (Dobrin, 1976) would produce a depth to the second layer

The field data, consisting of elevations and locations of shotholes and geophones, and the seismic arrival times at each geophone are input into a computer via a field portable terminal or other means, for execution of the seismic interpretation program. The program provides the interpreter with the following information:

(1) a listing of the input data; (2) a time-distance curve of the data;

FIG. 4. Using sledgehammer and striker plate for a sound source in a seismic refraction survey.

242 Haeni

(3) calculated velocities of each layer; (4) a table of depths to each refractor under the shot-

points and geophones; and (5) a plot of the computed geologic cross-section.

In addition, common field problems such as elevation correc- tions, offset shots, and profile lines that are not straight can be handled easily by the computer program.

A seismic refraction profile from a groundwater modeling study underway in Monroe, CT is shown in Figure 7. This is the final output of the seismic refraction portion of the study, and the information obtained can now be used in the groundwater modeling study.

USE IN GROUNDWATER MODELING STUDIES

A groundwater modeling study conducted by the U.S. Geo- logical Survey in Newtown, CT (Haeni, 1978) utilized seismic refraction techniques. Its purpose was to predict the effects of groundwater withdrawals on a stratified-drift stream aquifer system. Similar modeling studies in New England that also used the seismic refraction method were described in Mor- rissey (1983) and Mazzaferro (1983).

The seismic refraction survey in the Newton study was initiated after all existing data for the aquifer were inventoried and before any test drilling was started. Figure 8 shows the location of the seismic refraction profiles in the study area. Lines A-A’, D-D’, G-G’, H-H’, and I-I’ were along the axis of the

valley and were used to determine the saturated thickness of the aquifer.

Line G-G (Figure 9) shows the asymmetry of the bedrock channel beneath the stratified drift. Using these data, the hydrologist can choose the optimum placement of any planned drill holes in this study area.

Line I-I’ (Figure 9) was located in a swampy area that had limited vehicle access; the data collected by the refraction survey were the only economical means to obtain aquifer data. Based on existing data, this site was expected to show thin aquifer material (and its position at the head of the valley). Instead, the results showed that 21 m of saturated material was present. Subsequent drilling nearby proved this area had coarse-grained material and was an excellent site for future well-field investigations. The local water company subsequent- ly conducted a detailed site investigation, and two production wells were eventually located near this site.

Lines B-B’, C-C, E-E’, and F-F’ (Figure 8) were to the east side of the main valley and were located where no existing geohydrologic information was available. The purpose of these lines was to determine if any appreciable saturated aquifer thickness existed on the sides of the main valley under a thick section of unsaturated material. The results of lines E-E’ and F-F’ are shown in Figure 10.

After the results of the seismic surveys were interpreted, the data were used to design an exploratory drilling program in the valley. Test holes were located where additional data were needed and in areas that had thick, saturated sections.

FIG. 5. Pickup truck set up for seismic refraction survey. The hole is for two-component explosive.

SO.6

Groundwaler Modeling Studies 243

c Firrt arrival of comprerrional reismic wavea

0, 26 60 76 100 126 1 i0 176 200 226 260 276

time (msj

FIG. 6. 12-channel seismograph record from Little Androscoggin River valley, Maine.

I

Ik 626

Unsaturated strathed drift

Vetoclty of sound = 0 3km:s

-160

Water table

Saturated stratlfled drift Velocity of sound = 1 5 km!s

DISTANCE (m)

VERTICAL EXAGGERATION X0.93

FIG. 7. Seismic refraction profile, Monroe, CT.

244 Haeni

78’16’

41’22’20

SCALE ‘16’

41’22’ 30’

Q

b

1 2 Miles I

I I 1 2 km

EXPLANATION

LJ Stratified drift

_Q 1.5_ Land-surface contour - Shows altitude of land surface. Contour interval 30.5 m Datum is mean sea level

‘4’ Seismic Refraction Profile

FIG. 8. Location of seismic refraction profiles, Pootatuck River valley, Newtown, CT.


The existing data, the results of the seismic survey, and the results of the drilling program were combined to construct a bedrock contour map (Figure 11) and a water-table map of the aquifer. These maps were then used to generate input data for a groundwater flow model of the aquifer.

The finite-difference groundwater flow model described in Trescott et al. (1976) was used to simulate the response of the stratified-drift aquifer to imposed natural and manmade stresses. This model can simulate groundwater flow in two dimensions in a water-table (unconfined) aquifer with irregular boundaries and inhomogeneous composition. The sources of water may include aquifer storage, recharge from precipi- tation, inflow across the aquifer boundaries, and recharge through streambeds. Water is discharged through wells, evapotranspiration, leakage to surface-water bodies, and groundwater outflow.

To utilize finite-difference approximations for solution of the groundwater flow equation, the modeled area must be subdivided into blocks by a rectangular grid. The grid network for the Newtown stratified-drift aquifer model consisted of 45 rows and 102 columns. It defines 4 59q blocks, the center of which are termed “nodes.” The resulting network is referred to as a block-centered, finite-difference grid with variable grid spacing (Figure 12).

The groundwater flow equation is approximated at each node of the grid. Consequently, the properties of the aquifer

and other hydrologic parameters must be defined at each node in the modeled area. Appropriate aquifer values of hydraulic conductivity, bedrock altitude, and water-table altitude must be determined. Estimates or measurements of the other hydrologic parameters of recharge, boundary fluxes, and groundwater evapotranspiration must also be made. Entering all of these values in the model produces a water-table map of the aquifer.

The water levels simulated by the flow model can then be compared with records of actual water levels from wells and the seismic refraction data. The adequacy of the comparison then becomes one of the measures of the model’s ability to predict the results of natural or man-made stresses on the stream-aquifer system. Here the estimated natural recharge to the aquifer under varying natural conditions was used as a stress on the system.

The initial process of adjusting the model’s input parameters to duplicate a set of field conditions is called calibrating the model. If the modeler varies all of the parameters in the model during the calibration process, there is no assurance that the result is not just one of many possible combinations. The problem of lack of a unique set of model results can be minimized by limiting the number of model parameters that need adjustment during model calibration.

It is here that seismic refraction studies help the hydrologist. One parameter, the altitude of the bedrock surface, can be

JtRATIFIED DRIFT

DISTANCE (m) VERTICAL EXAGGERATION X2

go- STRATIFIED DRIFT -90

-80


FIG. 9. Seismic refraction profile, Newtown, CT.

246 Haeni

assigned values that need not be changed during the calibration process. In addition, by using this technique, problem areas within the modeled area that develop during the modeling process can be checked in the field for water level and saturated thickness data. Other advantages of using the seismic refraction technique in groundwater modeling studies are as follows.

(1) It is an economical method to obtain data for groundwater mode! input.

(2) The data can be in the form of continuous profiles instead of individual point data.

(3) The data are relatively accurate. Depths to individual refractors are normally within 10 percent of the actual depths (Zohdy et al., 1974).

E /Toddy HII Road

CONCLUSIONS

Seismic refraction techniques, long used in exploration for minerals and petroleum, can be effectively used in groundwater modeling studies. Groundwater aquifers in New En- gland have large seismic velocity contrasts at major hydrologic boundaries; consequently, this technique can readily define the geometry of the aquifer. Groundwater modeling studies in Connecticut and Maine have used this technique and show that_ it_ is_ a valuable and economical method of obtaining input data for modeling, designing drilling programs, and improving modeling results. Seismic refraction techniques are therefore a valuable tool for the hydrologist conducting groundwater modeling studies in New England and other geologically similar areas.

._.~ ~~ DISTANCE (m)

VERTICAL EXAGGERATION X2

F F I

/ Toddy H~ll Road

Waler Table

SfRAtlFICID DRIFT -


FIG. 10. Seismic refraction profile, Newtown, CT.

EXPLANATION

FIG. 11. Contour map of the bedrock surface, Pootatuck River valley, Newtown, CT.

bmm 1:24.000. lRI3. #mmrim4 1072 htslwl1:24.000.1969. ~*m6visg4 1372

0 H -.

1 Mile L__ ‘_ -1

0 .5 1 km UUUUU]

i:i* IS

EXPLANATION

FIG. 12. Map showing model boundaries and grid network for the Newtown, CT stratified-drift aquifer model.


REFERENCES

Beggs, G., and Garriot, J. C., 1979, Shotgun surface source: 49th Ann. Internat. Mtg. and Expos., Sot. Explor. Geophys., New Orleans.

Dobrin, M. B., 1976, Introduction to geophysical prospecting: McGraw-Hill Book Co.

Grant, F. S., and West, G. F., 1965, Interpretation theory in applied geophysics: McGraw Hill Book Co.

Haeni, F. P., 1978, Computer modeling of ground-water availability in the Pootatuck River valley, Newtown. Connecticut, with a section on quality of water by Elinor H. Handman: U.S. Geol. Surv. Water Resources Investigations 78-77.

Mazzaferro, D. L., 1984, Groundwater availability and water quality in Southbury and Woodbury, Connecticut, US. Geol. Surv. Water Resources Investigations Open-file Rep. 83-4221.

Mooney, H. M., 1981, Handbook of engineering geophysics: Bison Instruments, Inc.

Morrissey, D. J., 1983, Hydrology of the Little Androscoggin River valley aquifer, Oxford County, Maine, U.S. Geol. Surv. Water Re- sources Investigations Open-file Report 83-4018.

Musgrave, A. W., ed., 1967, Seismic refraction prospecting: Sot. of Explor. Geophys.

Pakiser, L. C., and Black, R. A., 1957, Exploring for ancient channels with the refraction seismograph, Geophysics, 22, 3247.

Parasnis, D. S., 1979, Principles of applied geophysics, third edition: Chapman and Hall.

Redpath, B. B., 1973, Seismic refraction exploration for engineering site investigations NTIS AD-768710.

Sander, J. E., 1978, The blind zone in seismic ground-water exploration: Ground Water, 165,394397.

Scott, J. H., 1977, SIPT-A seismic refraction inverse modeling program for timeshare terminal computer systems: U.S. Geol. Surv. Open-file Report 77-365.

Scott, J. H., Tibbetts, B. L., and Bordick, R. G., 1972, Computer analysis of seismic refraction data: Bureau of Mines Rep. of Investi- gations, RI 7595, U.S. Department of the Interior, Bureau of Mines.

Slotnick, M. M., 1959, Lessons in seismic computing: Sot. of Explor. Geophys.

Soske, J. L., 1954, The blind zone problem in engineering geophysics: Geophysics, 24, 3599365.

Telford, W. M., Geldart, L. P., Sheriff, R. E., and Keys, D. A., 1976, Applied Geophysics: Cambridge Umv. Press.

Trescott. P. C.. Pinder. G. F.. and Larson. S. P.. 1976. Finite- difference model for aquifer simulation in’two dimensions with results of numerical experiments: U.S. Geol. Surv. Techniques of Water-Resources Investigations, book 7, chapter Cl.

Zohdy, A. A. R., Eaton, G. P., and Mabey, D. R., 1974, Application of surface geophysics to groundwater investigations: U.S. Geol. Surv. Techniques of Water-Resources Investigations, book 2, chapter Dl.

application of seismic refraction methods in groundwater...

Documents