Shallow Seismic Investigation of Hydrogeologic Problems in the Brazos River Alluvium, Texas A&M Plantation, Burleson County, Texas
M. A. McBrayer
Twenty-two shallow, reversed, seismic refraction profiles were conducted in the Brazos River floodplain to test the feasibility of using seismic methods to provide hydrogeologic information in this province. The specific objectives were to map the total and saturated thickness of the alluvial deposits and to outline gravel lenses within the alluvium.
It was found that the water table was the only interface at which the acoustical properties of the deposits above and below changed sufficiently to be mapped by seismic methods. The alluvial deposits above the water transmitted compressional waves at an average velocity just greater than the velocity of sound in air. The saturated alluvial deposits transmitted seismic waves at an average velocity of slightly greater than the velocity of sound in water.
The saturated alluvial deposits and the bedrock appear to have a continuous increase in velocity with depth rather than a significant change in acoustical characteristics.
The seismic measurements in themselves could not delineate gravel lenses within the alluvium. However, the zones of greater permeability were indicated on the resulting contour map of the water table by the areas of gentle gradients. Zones of greater permeability in the alluvial deposits are probably gravel lenses.
The object of this study was to investigate the hydrogeology of the alluvial deposits of a portion of the Texas A&M Plantation using the techniques of shallow refraction seismology. An attempt was made to use these methods to map the total thickness and the saturated thickness of the alluvial deposits and to outline zones of gravel lenses within the alluvium. Domestic water wells can be made in the alluvium almost anywhere in the floodplain, but the high discharge irrigation wells are limited in location by the erratic distribution of gravel lenses. Although geophysical techniques do not result in as much detailed or as accurate information as does test drilling, they may outline the more favorable areas for well locations without the necessity of the present procedure of extensive test drilling.
The area chosen for this investigation was located on the Texas A&M Plantation, Burleson County, Texas. The Plantation is that portion of the University farm in the Brazos River floodplain, locally known as the "Brazos Bottoms." The test site, an area approximately 3,OOO feet by 1,2OO feet, was bordered by the Brazos River on the northeast side (figure 1). The topography is fairly flat, the maximum relief being about 6 feet.
The standard seismic refraction technique was followed and the velocities and depths were calculated by standard mathematical solutions of the travel-time plots. A discussion of the general theory and methods are given in Jakosky (195O), Dobrin (196O), and Griffiths and King (1965). Briefly, the seismic method consists of the generation of an elastic pulse near the surface of the ground and recording the resulting motion of the ground at nearby points on the surface. Measurements of the time intervals between the generation of the pulse and its detection at the geophones at various distances give the velocity of propagation of the energy in the ground. Where the subsurface structure is simple, the values of elastic wave velocities and the positions of boundaries between regions of differing velocity can be calculated from the timedistance data. These calculations require a simple model of the structure; that is, the velocity zones should be homogeneous and isotropic with planar interfaces between each zone. It should be remembered that a seismic profile is an average profile of the refracting horizon, with the highs and lows being smoothed out. Hence, the difference in depth between the true surface and the seismic interfaces may become a relatively large percentage of the total depth in shallow-studies.
As the depth of seismic refraction investigations becomes shallower, the differences between the theoretical model and the geologic conditions become pronounced. Observational errors, errors due to assumptions, and errors due to rapid changes in conditions within short vertical and horizontal distances may result in a relatively large error in the depth estimate when making short refraction profiles. The disturbing factors are due to the method of generating the seismic wave, the shot hole conditions, the topography of the ground surface, inhomogeneity and horizontal and vertical changes in the stratification, and irregularities in the refracting surfaces.
Shallow seismic studies are useful because they rapidly provide a picture of the refracting interfaces to serve as a guide for any subsequent drilling. The accuracy of the depths to and nature of the refracting layers would be greatly improved if some preliminary test drilling information were available.
Shallow seismic refraction methods can be used to map the top of the saturated zone in the unconsolidated or poorly indurated sediments of the Gulf Coastal Plain with some accuracy. Although the seismic method cannot differentiate sediment types below the water table, the contour spacIng on the resulting map of the water table should outline zones of higher permeability. To test this assumption, a profile could be made across this zone before and during a period of heavy irrigation withdrawal. If these are zones of greater permeability, the water table should decline during the pumping period.
The elastic properties of the saturated alluvium and the bedrock did not contrast sufficiently for this interface to be mapped by seismic methods. The poor induration of most of the Tertiary sediments in the Coastal Plain would hinder mapping of the alluvium-bedrock interface by seismic methods. Upstream, however, where the alluvium overlies more indurated Cretaceous deposits, shallow refraction methods could probably map this interface.
The range of uncertainty in the depth determinations is probably an indication of the combined effects of inhomogeneity of the deposits and the departure from flatness of the seismic interfaces. This range is not a criterion for judging the correctness of the interpretation as the interpretation was made in the original grouping of the data before making the least-squares solutions.