CORRELATION OF STRUCTURAL LINEAMENTS AND FRACTURE TRACES TO WATER-WELL YIELDS IN THE EDWARDS AQUIFER, CENTRAL TEXAS APPROVED: For Carmen, whose patience, love, and encouragement made this thesis possible CORRELATION OF STRUCTURAL LINEAMENTS AND FRACTURE TRACES TO WATER-WELL YIELDS IN THE EDWARDS AQUIFER, CENTRAL TEXAS by Kenneth Bower Alexander, A. A., B. S. THESIS Presented to the Faculty of the Graduate School of The University of Texas at Austin in Partial Fulfillment of the Requirements for the Degree of MASTER OF ARTS THE UNIVERSITY OF TEXAS AT A USTIN December, 1990 ACKNOWLEDGMENTS Throughout this study, many people generously donated their time, skills, and ideas in assisting my research. Without their help this thesis would not have been completed. Raymond Slade, of the U.S. Geological Survey, provided the initial impetus for my research. In addition to serving on my thesis committee, he furnished valuable technical support and introduced me to a 'higher' level of technical-report writing. Dr. Philip Bennett supervised my thesis and studies at The University. Over the past two years, Phil has provided continuous encouragement for my classwork, teaching, fieldwork, and research. His advice, covering a wide range of subjects, is greatly appreciated. I am proud to be the first student to complete a degree under his supervision. Dr. Jack Sharp also served as a member of my thesis committee. I would particularly like to thank him for his contributions to my understanding of hydrogeology and appreciation of Texas culture. This thesis was financially supported by the Barton Springs/ Edwards Aquifer Conservation District. Many thanks go to Bill Couch, the general manager, and the rest of the competent staff at the District. Tom Heathman, in particular, furnished critical information on the practical aspects of karst hydrogeology in the Edwards aquifer. Financial support was also generously provided by the Gulf Coast iv Association of Geological Societies, the UT Geology Foundation, and Dr. Jay Lehr of the National Water Well Association. Two computer experts were invaluable to my research. Rick Edson, of the UT Bureau of Economic Geology, furnished excellent assistance during the initial phase of my research especially with the lineament digitization and analysis process. Randy Ulery, a GIS computer specialist with the U. S. Geological Survey, combined the digitized lineaments with well and boundary locations to create the base map. His professionalism and generosity are greatly appreciated. I am indebted to the field assistants who helped me collect my data during the summer of 1990: Charlie Hewitt, Sevin Bilir, Annette Peloquin, Carolyn Runyon, Andrew Quarles, and Carmen Alexander. Other individuals who provided support include: • Dr. Charlie Kreitler and Dr. Rainer Senger, UT BEG • Dr. Chock Woodruff, independent consulting geologist • Laura De La Garza, formerly of the City of Austin • Jane Maler, Edwards Aquifer Research Center • Scott Thieben, UT Department of Geological Sciences • Byron Benoit, Associated Drilling Matthew X. Wickham and Gay Goodwin served as student editors for my thesis. Many thanks and beers go to them and fellow hydro students Carla Fuller, Charlie Hewitt, Gilbert Gabaldon, Malcolm Ferris, Sevin Bilir, Dana White, Fred Holzmer, Barbara Mahler, and Tom McKenna to name a few. The thesis was submitted to the Committee in November, 1990. v ABSTRACT CORRELATION OF STRUCTURAL LINEAMENTS AND FRACTURE TRACES TO WATER-WELL YIELDS IN THE EDWARDS AQUIFER, CENTRAL TEXAS by Kenneth Bower Alexander, A. A., B. S. Supervising Professor: Dr. Philip C. Bennett Lineaments are "straight lines visible from afar on the surface of the earth". In the Austin, Texas area, lineaments reflect the structural grain of the Balcones-Ouachita fault zone and may indicate subsurface geologic phenomena such as faults, fractures, and joints. These structural features often represent discrete zones of high permeability, and thus, areas of enhanced flow of groundwater capable of transmitting greater quantities of water than surrounding, non-fractured, rock. For this study more than 900 lineaments and fracture traces, identified in aerial photographs during a previous study, were detected in the Barton Springs section of the Edwards Aquifer. The endpoints of vi each linear feature were digitized and tagged with a unique identification label. Rose plots, Cartesian histograms, and a series of statistical operations were utilized to illustrate regional trends in the orientation of lineaments. As an indicator of well productivity, specific capacities of 27 wells in the area were obtained. Sixty-one water samples were collected and analyzed to test for possible chemical evidence of lineament-well interactions. The orientations of lineaments and fracture traces in the study area clearly display a bimodal distribution with a primary trend of N 40 E and a secondary peak of N 50 W. A general correlation exists between increased well productivity and decreased distances to the nearest lineament, particularly within 200 feet of lineaments. Also, 10 of the 13 largest specific-capacity values are from wells located southeast of southwest-northeast trending lineaments. Nonparametric statistical methods show that direction from lineaments is a significant factor in predicting water-well yields. Lineaments provide a tool for predicting possible sites of envi­ronmental sensitivity with respect to groundwater resources. Examples include the siting of groundwater monitoring wells for point sources of pollution, predicting the likely underground flow paths of a pollution plume or locating dam sites for recharge enhancement. Awareness of the location, orientation, and density of structural lineaments will allow the water-resource manager to identify discrete groundwater flow paths, and, thus, predict contaminant plume migration. vii TABLE OF CONTENTS I. Introduction..................................................................................................... 1 A) Previous Lineament Studies in Central Texas .......................... 3 B) Research Objectives ......................................................................... 4 II. Lineaments and Fracture Traces ..... .......................................................... .5 A) Background Information ...............................................................5 Terminology ................................................................................ 5 Structural Controls ..................................................................... 6 Historical Research ..................................................................... 9 B) Lineament Analysis ........................................................................ 12 Remote Sensing .......................................................................... 12 Physiographic Expressions ........................................................ 13 Alternative Methods of Analysis ............................................ 15 C) Relation of Lineaments to Groundwater Flow.........................17 Caves .............................................................................................. 19 III. Barton Springs Segment of the Edwards Aquifer.............................21 A) Location ............................................................................................. 21 B) Geology .... , ......................................................................................... 24 Stratigraphy..................................................................................24 Structural Geology ...................................................................... 26 a) Faults ......... , ................................................................. 27 b) Joints ............................................................................28 C) Hydrology .......................................................................................... 29 Recharge ........................................................................................ 32 Anisotropic Groundwater Flow .............................................. 33 D) Water Chemistry ............................................................................. 36 IV. Methods of Data Analysis .......................................................................... 38 A) Lineament Analysis ........................................................................ 38 Determination of Lineaments ................................................. 38 Statistical Analysis ..................................................................... .41 viii B) Well Analysis ................................................................................... 44 Productivity of Water Wells .................................................... 44 Collection of Well Data ............................................................ .45 Statistical Analysis ...................................................................... 49 C) Water Chemistry ............................................................................. 52 Field-Data Collection .................................................................. 52 Laboratory Analysis ................................................................... .54 V. Results and Discussion...............................................................................60 A) Orientations of Lineaments .......................................................... 60 B) Correlation between Well Proximity and Locations of Lineaments ................................................................................ 66 C) Correlation between Well Productivity and Directions to Lineaments ...........................................................................72 D) Correlation between Water Chemistry and Locations of Lineaments ................................................................................ 80 VI. Conclusions and Implications..................................................................89 VII. Appendices..................................................................................................93 A) Previous Pump Tests in the Edwards Aquifer .......................... 93 B) Results of Chemical Analyses from LCRA Laboratory ........... 98 VIII. Bibliography ............................................................................................... 100 Vita ix LIST OFTABLES Table 1: Summary of selected statistics for azimuths and lengths of digitized lineaments and fracture traces in the study area ...... 41 Table 2: Specific-capacity data tabulated for wells in the study area ..... .48 Table 3: Chemical analysis from water samples in the study area ....... .56 Table 4: Summary of selected statistical analysis of lineament azimuths............................................................................................. 67 Table 5: Specific-capacity values ranked in ascending order ..................73 Table 6: Summary of Mann-Whitney U test results ................................ 78 x LIST OF FIGURES Figure 1: Conceptual diagram of groundwater along fracture zones in carbonate rock ................................................................ 8 Figure 2: The Edwards aquifer in Texas ..................................................... 22 Figure 3: The study area................................................................................. 23 Figure 4: Generalized hydrogeologic column........................................... 25 Figure 5: Generalized groundwater flow in the study area ................... 31 Figure 6: Location of lineaments in the study area ............................... .40 Figure 7: Box diagrams of distribution for values of specific capacities and distances from wells to nearest lineament .... 50 Figure 8: Rose diagrams of lineament and fracture trace orientations in the study area ..................................................... 61 Figure 9: Cartesian histograms of lineament and fracture trace orientations in the study area ..................................................... 62 Figure 10: Graph showing relation between values of lineament azimuth and relative lineament length in the study area....................................................................................................63 Figure 11: Graphs showing values for mean length, sector length, and frequency of lineaments in the study area ....................... 65 Figure 12: Graphs showing well productivity and the lineament type in the study area ....................................................................70 xi Figure 13: Graph showing values of specific capacities and distances from wells to nearest lineament northwest of each SE well. ....................................................................................7 4 Figure 14: Graph showing values of specific capacities and distances from wells to nearest lineament not located northwest of each non-SE well.. ................................................. 75 Figure 15: Graph showing values of specific capacities and distances to the nearest southwest-northeast trending lineament located to the northwest or southeast of a well.. ..................... 76 Figure 16: Graph showing values of specific capacities and distances to the nearest southwest-northeast trending lineament located to the northwest of the well .......................................... 77 Figure 17: Piper diagram of samples collected in the study area ............ 81 Figure 18: Graph of carbonate chemistry for samples collected in the study area ................................................................................. 82 Figure 19: Graph of sulfate ratios versus chloride and strontium for samples collected in the study area ............................................ 84 Figure 20: Schematic cross-section across the Balcones fault zone ........ 85 Figure 21: Graph of calcium-magnesium relationship for samples collected in the study area ............................................................87 Figure 22: Graph of sodium-chloride relationship for samples collected in the study area............................................................ 88 Plate 1: Location map for wells and lineaments examined in study ......................................................map pocket xii I. INTRODUCTION This study investigates the association between structural lineaments and water well yields in the Barton-Springs segment of the Edwards aquifer. Lineaments, in simplest terms, are "straight lines visible from afar on the surface of the earth" (Woodruff & Caran, 1984). In the Austin, Texas area, lineaments reflect the structural grain of the Balcones-Ouachita fault zone and may indicate subsurface geologic phenomena such as faults, fractures, and joints. These structural features often represent discrete zones of high permeability, and thus, areas of primary groundwater movement capable of transmitting greater quantities of water than surrounding, non-fractured rock. Lineament analysis can be a useful tool for the hydrogeologist. The technique is especially valuable for predicting possible sites of environmental sensitivity with respect to groundwater resources. Examples include the siting of groundwater monitoring wells for point sources of pollution, predicting the likely underground flow paths of a pollution plume or locating dam sites for enhanced recharge. Lineament analysis may also be used to increase the probability of drilling large yield water wells. Consequently, the location, orientation, 1 and density of structural lineaments may be valuable in forecasting problems with quantity as well as quality of groundwater. In many areas in the Edwards aquifer, groundwater is available in sufficient quantities to make the process of locating wells to extract small yields of water a relatively simple task. However, other areas appear to have a limited amount of groundwater available and considerable difficulty can be experienced in obtaining an adequate water supply without drilling several wells. One serious problem with the use of groundwater as a source of water supply in the Edwards is that well yields vary tremendously. Wells with yields ranging from 2 and 500 gallons per minute (0.13 and 31.5 liters/second) can commonly be found within a short distance of one another. When larger yielding wells are required to meet the water needs of a subdivision, an industry, or a small town, the problem of obtaining sufficient well yields becomes even more difficult. The typical approach is to drill either a single well with a sufficiently high yield or drill several wells so that the combined yield will meet the estimated water needs. This approach is expensive and frequently not successful. If a procedure was available that would increase the likelihood of obtaining a higher yield in each well drilled, it would reduce the cost of developing groundwater as a source of water supply and decrease the probability of failure to acquire the quantity of water required. The expense of drilling several dry holes or wells with an inadequate yield to satisfy the owner's water needs can become quite large. The typical cost of a drilled well ranges from one to five thousand dollars or more depending on the depth of the well. Therefore, the ability to obtain the required well yield with as few drilled wells as possible is an important factor in determining the economics of groundwater as a water supply. One goal of this research project is to develop a technique that can be used to locate well sites that have a higher probability of producing large yields so that the cost of developing groundwater supplies can be minimized. Reliable methods of detecting the major groundwater flow zones in the Edwards aquifer are needed for the most economical development of its groundwater resources. A) Previous Lineament Studies in Central Texas Previous studies of lineaments in Texas investigated (1) the effects of Balcones faulting on linear features (Wermund et al., 1974, Collins & Laubach, 1990), (2) parallelism between the trends of lineaments and structural features (Dix & Jackson, 1981, Myrick et al., 1988), (3) correlation between lineaments and faults (Caran et al., 1982; Kreitler, 1976), and (4) the relationship between lineaments and geothermal potential (Woodruff & Caran, 1984). In the Barton-Springs segment of the Edwards aquifer, the relationship between lineaments and transmissivity was first investigated by De La Garza and Slade (1986a). Controlled aerial mosaics were used to map two sets of lineaments. Transmissivities of 47 wells were then estimated from specific-capacity data obtained from drilling records. After determining the distances from each of the wells to the nearest lineament, a correlation was found between increased well productivity and decreased distances to "short" lineaments (between 1000 feet and 4.5 miles (30 meters and 7.2 kilometers) in length). More recently, Woodruff et al. (1989) examined the effects on lineaments in the Edwards aquifer by (1) orientations of faults and joints, (2) drainage patterns, (3) topography, and (4) hydrology. B) Research Objectives The objective of this study is to expand a previous lineament study of the Edwards aquifer (De La Garza & Slade, 1986) by acquiring reliable data for specific-capacity and water-chemistry for as many wells as possible and accurately mapping and transferring the resulting data to a base map. In addition, this study will include statistical and chemical analyses to further investigate any conclusive evidence of lineament-well yield correlation in the Edwards aquifer. Specific objectives are to: • Examine statistical relations between values for well yields and lineaments of different types, lengths, and orientations • Investigate possible groundwater flow paths using chemical analysis • Develop a predictive model for well yields in the karstic Edwards aquifer II. LINEAMENTS AND FRACTURE TRACES A) Background Information Terminology There is a substantial amount of misleading terminology on linear features in the literature which has proliferated since the advent of spacecraft and high-altitude imagery. Part of the confusion with lineaments is a result of semantics. For example, a lineament is not necessarily a fault. Nor is any single lineament a single fracture or joint in the bedrock. Faults and joints may be expressed as lineaments, "but the term 'lineament' implies that the precise geologic nature of each individual feature is ambiguous" (Woodruff, 1989). Structural lineaments are simply surface manifestations of subsurface geologic features such as faults, fractures and joints. O'Leary et al. (1976) reviewed the origin and usage of the terms linear, lineation, and lineament. They have attempted to standardize the terminology by strictly defining each expression. For this study, their definition of lineament will be used: A lineament is defined as a mappable simple or composite linear feature of a surface, whose parts are aligned in a 5 straight or slightly curving relationship, and which differs distinctly from the patterns of adjacent features and presumably reflects a subsurface phenomenon. Furthermore, the structural lineaments reviewed in this study will be divided as recommended by Parizek (1976): linear features less than 1500 meters (4922 feet) will be classified as fracture traces; longer than 1500 meters will be classified as lineaments. Lattman (1958) noted that the term fracture trace is used as opposed to the term fracture because the surface expression of the linear feature is an indirect indication of some subsurface discontinuity. The fracture is not observed except in the case of features that are mapped on aerial photographs of an area where bare bedrock is exposed at the surface. Because the true nature of the subsurface discontinuity is usually unknown, the linear feature on the ground surface is more appropriately identified as a fracture trace. In an analogous manner, the term lineament simply identifies a long linear feature on the surface of the earth that is associated with a long subsurface discontinuity that may be continuous or discontinuous and for which the origin and characteristics are unknown. Structural Controls Although the nature of the subsurface discontinuity associated with a fracture trace or lineament cannot be observed under normal circumstances, there have been a number of situations where the characteristics of the subsurface feature could be examined. These situations have typically occurred where a fracture trace or lineament intersected a vertical cliff, deep highway cut, rock quarry, or some other type of excavation that would expose a cross-section of the discontinuity for a considerable depth. Several investigators have found examples of exposed subsurface discontinuities associated with fracture traces and lineaments and have described the nature of the features. Lattman and Matzke (1961) found a zone of joint concentration underlying a fracture trace for which a cross-section was exposed by a vertical cliff in sandstone in Wyoming. The fracture trace was expressed as an alignment of drainage features along a straight, shallow, topographic trough on the aerial photographs. The apparent linearity of lineaments and fracture traces, regardless of topography, was interpreted by Trainer and Ellison (1967) as an indication that the associated fracture zones are vertical or nearly vertical. In addition to the vertical fractures, the fracture zones usually contain a series of horizontal joints that connect adjacent vertical fractures (Figure 1). This is an important characteristic when the objective of a well-drilling program is to intersect one or more fractures with a vertical well. The characteristics of the subsurface fracture zones also appear to vary by rock type. Situations have been observed in which a subsurface­fracture zone consisted of closely spaced fractures in one rock type and more widely spaced fractures in another rock type at a different elevation within the same fracture zone (Stafford et al., 1983). The Figure 1: Conceptual diagram of groundwater along fracture zones in carbonate rock (from Lattman & Parizek, 1964). fracture zone is generally connected to the layer of residual soil at the soil-rock interface so that the soil layer can serve as a groundwater recharge zone for the fractures in the rock. In fact, there is usually a deeper layer of residual soil along lineaments and fracture traces than in adjacent areas. Apparently the fractured rock along these natural linear features weathers more rapidly to produce a layer of saprolite. The more rapid weathering of the rocks within the fracture zones is one reason that many lineaments and fracture traces occur as linear topographic depressions. In turn, these depressions provide enhanced opportunity for groundwater recharge along the natural linear features. Moore and Stewart (1983) examined a Floridian limestone aquifer and noticed increased dissolution of the underlying limestone directly below a fracture trace. Using surface geophysical techniques, they found that the linear, surface feature had been propagated upward through 20 meters (66 feet) of overburden. Trainer (1967) concluded that extensive weathering of vertical fractures near the surface was probably accompanied by subsurface weathering and possible widening of other fractures that are not expressed as fracture traces. Thus, the existence of a fracture trace at the surface typically indicates that a number of open fractures are present in the underlying bedrock. Historical Research The geologic study of lineaments has its ultimate origins in Britain during the period from 1800 to 1835 where a small number of influential geologists established that, in general, there is a system to the pattern and distribution of faults and joints and that these systems can maintain remarkably constant azimuths over significantly large areas (Hodgson, 1974). The systematic arrangement of fractures and faults, which was well documented by 1835, allowed William Hopkins to develop an advanced mechanical theory to account for the phenomena. In 1841, he published a map of the Wealdon Dome that shows directly the orthogonal relations of the major linear features of the region as predicted by his theory. The map appears to be a first attempt to show lineaments directly and in relation to other structures. William H. Hobbs (1904) has been generally credited with the introduction of modern techniques of study of natural linear features. He recognized the existence and significance of linear geomorphic features that were the surface expression of zones of weakness or structural displacement in the crust of the earth. He defined lineaments as "significant lines in the earth's face" and added that: These significant lines of landscape which reveal the hidden architecture of the rock basement are described as lineaments. They are character lines of the earth's physiognomy. John L. Rich (1928) was apparently the first geologist to report on the observation of linear features from the air. He recognized linear vegetational, tonal, and topographical alignments while flying over the limestone areas of Oklahoma. He suggested the use of aerial photographs to further study these linear features which he related to bedrock jointing. Little work was done using Rich's idea until the major oil companies became interested in the use of fracture trace studies for their exploration programs in the 1950's. Lattman and Olive (1955), Blanchet (1957), Mollard (1957), and Lattman (1958) all described interpretation of linear features in various geologic terranes. In 1964, Lattman and Parizek established the important relationship between the occurrence of groundwater with lineaments and fracture traces for carbonate aquifers. Based on analyses of 13 wells in the State College, Pennsylvania area, they determined that lineaments and fracture traces are underlain by zones of localized weathering, increased permeability, and porosity. In fact, they found specific capacities of wells situated along fracture traces to be 10 to 100 times greater than those of wells sited off the lineament trends. Three bore-hole caliper logs obtained from wells drilled on fracture traces confirmed that numerous cavernous openings were penetrated by these wells, which reached 350 feet (107 meters) in depth. As expected, caliper logs from wells drilled in interfracture areas showed few cavernous openings. However, according to Taylor (1980), their investigation had two basic deficiencies: (1) There was an insufficient number of wells intentionally located on and off fracture traces; (2) Little attempt was made to separate the influence of fracture traces from other hydrogeologic variables that affect well yields. From an analysis of data for 80 wells in the Nittany and Penns Valleys of central Pennsylvania, Siddiqui and Parizek (1971) found that fracture-trace wells were far more productive than nonfracture-trace wells and that the probability of obtaining a certain productivity is greater in fracture-trace wells than in nonfracture-trace wells. They point out that large productivity reflect large porosity and permeability around a well bore and argue that fracture traces must reflect underlying zones of increased porosity and permeability in the bedrock. A basic problem with this study is that many of these hydrogeologic factors are not easily separated which results in the authors having to make certain assumptions that are not acceptable to all investigators (Taylor, 1980). The effect of lineaments on water-well productivity was also investigated by LaRiccia and Rauch (1977). They examined 65 domestic wells in the Grove and Frederick limestones of Frederick Valley, Maryland. Specific-capacity values were found to be significantly larger for pumped wells within 100 feet (30 meters) of a lineament compared to more distant wells. Since then, many other studies have confirmed that remotely-sensed linear features are reliable indicators of areas of increased groundwater flow (Wermund et al., 1978; Caran et al., 1982; Hunter & Gutierrez, 1985; Wright, 1985). B) Lineament Analysis Remote Sensing The interpretation of lineaments is conducted from a remote vantage point. This remote view allows linear features to be perceived from a variety of clues. These features include dark or light lines in the soil, alignments of vegetation, topographic sags, aligned gaps in ridges, straight stream reaches, and other similar characteristics. Often these linear features are expressed on photographs and on the ground by a combination of features. For example, a straight stream segment may extend into soil tonal alignments in an adjacent field, which then passes into a line of slightly larger trees in a nearby wooded area, ending in an elongated sinkhole. While there are many types of aerial imagery available, there are only a few types that work well for fracture-trace mapping as applied to water-well location. Lattman's (1958) paper describes in detail the technique of mapping fracture traces. Briefly, fracture traces and lineaments are best mapped by stereoscopic examination of individual aerial photographs. Direction of flight-line overlap of at least 50 percent is necessary in order to obtain complete coverage of each photograph, with 20 or 30 percent of side lap if more than one flight line is to be analyzed. Because some of the features are only a quarter to a third of a mile in length, they are best mapped on aerial photographs at a scale of 1 to 20,000 (1" =1667'). Photographs are available for most of the United States at this scale. The examination is accomplished using a 2.5 power lens stereoscope and moving the instrument systematically over the photographs, mapping all linear features visible at each position. In addition, stereoscopic study allows the recognition of man-made linear features, bedrock schistosity, outcropping edges of dipping beds, and other features that should not be mapped as fracture traces or lineaments. Physiographic Expressions Probably the most obvious expression of fracturing visible on aerial photographs is short, straight, segments of streams and rivers. A prominent joint set or fault can markedly affect the direction of stream flow, and unusually straight segments in a drainage pattern invariably indicate some type of structural control. As previously noted, fracture traces are also expressed by the alignment of vegetation presumably due to a slightly higher concentration of moisture. A much more subtle expression is the occurrence of a line of slightly taller trees in a well-forested area (Hough, 1960). The same increase in availability of water allows these trees to grow a little taller than their neighbors, and although such a difference may not be particularly obvious, it is often distinguishable. Another type of fracture trace is the soil tonal change and alignment, particularly in an area that has been extensively planted with crops where the occurrence of vegetation and its patterns has been determined by man. Careful examination of the areas may reveal very subtle changes in the tone of the soil or of the crops themselves. Larger water content of the soil is expressed on the photographs by a relatively dark zone which allows otherwise insignificant differences in moisture to be seen. Still another manner in which fracture traces are indicated on aerial photographs is by broad but very shallow linear depressions in the topography. Mollard (1957) refers to these depressions as microrelief features and describes them as being "SO to 500 feet across, six inches to several feet deep and many hundreds of feet long." The theory of origin of these features attributes them to gradual leaching of the soil by downward percolation of water along a joint or zone of fracture. This photogeological expression of fracture traces and lineaments varies widely among different bedrock types and with different overburden thicknesses. In soluble limestone such as the Edwards Limestone, their expression can be very obvious due to enhancement through solution along the fractures, whereas in most other rock types their expression is more subtle. Alternative Methods of Analysis Several investigators have attempted to overcome the contrary, subjective nature of identifying lineaments by utilizing objective interpretive procedures. Trainer (1967) was the first to study lineaments statistically when he proposed an "objective method of investigating the areal abundance of fracture traces (lineaments) seen on aerial photographs." He argued in favor of a "uniform duration of search, in time per unit area," adding that the "rate of discovery of the traces decreases logarithmically with time." Trainer acknowledged concern over the reproducibility of his results, noting that "problems of subjectivity are inherent in the interpretation of aerial photographs." He also observed that it "is impossible to find all the fracture traces on a given image in a practical period of time." A different approach to the problem was proposed by Podwysocki et al. (1975). These investigators sought to minimize or eliminate "the effect of operator variability and subjectivity in lineament mapping" by use of several machine processing methods. They compared independent interpretations of an MSS band 5 Landsat image by four observers and related these results to those of another group. After analyzing the results, a "large amount of variability" in both the number and length of linear features was found. Podwysocki and his colleagues then attempted to use two machine-aided mapping techniques to simulate directional filters: (1) an edge-enhancement algorithm and (2) a "television (analog) scanning of an image transparency which superimposes the original image with one offset in the direction of the scan line." Although these methods created similar products, they were found to introduce processing artifacts that were mistaken for lineaments. Moreover, both methods still relied on an interpreter to detect and analyze linear features within the image so that even if the image had been faithfully enhanced, the presumed subjectivity of the interpretation would not be eliminated. The same conclusion is also applicable to most other automatic processing systems for mapping lineaments if they require decisions by interpreters after image enhancement is completed (Maffi & Marchesini, 1964; Robinson & Carroll, 1977; and McGuire & Gallagher, 1979). Other investigators used elaborate methods for evaluating the reproducibility of lineament interpretations. Burns et al. (1976) defined coefficients of reproducibility among populations of lineaments; they stipulated that the lineaments must have unit width based on pixel size. Burns and Brown (1978) refined this procedure by measuring reproducibility of digitized lineaments on a pixel--by--pixel basis. Huntington and Raiche (1978) described the degree of correlation or similarity among lineament interpretations stated in terms of the lineaments' "primary characteristics": (1) location, (2) orientation, (3) length, and (4) curvature. A drawback common to all of these procedures is extensive mathematical manipulation of the lineament data. Furthermore, the tests served only to check the relative agreement among multiple interpretations of a single image. Still other researchers have had mixed success in their efforts to develop an accurate, practical means of perceiving and analyzing lineaments. These have included: use of photos of side-illuminated raised plastic relief maps to enhance linear topography (Wise, 1969); use of transmitted rather than reflected light to view an image (Lattman, 1958); enhancement of satellite images by rotational photographic exposure of unexposed negative film through overlaid positive and negative transparencies (Lawton & Palmer, 1978); and the use of quantitative and predictive geological spatial analysis techniques coupled with digital elevation models to discern topographic lineaments (Eliason & Thiessen, 1986). Yet, none of these methods have been proven to accurately identify true structural lineaments on the surface of the earth without the aid of a human interpreter. Consequently, photo-interpretation of aerial photographs, despite its inherent subjectivity, remains the most effective procedure for identifying lineaments and fracture traces in conjunction with groundwater-resource investigations. C) Relation of Lineaments to Groundwater Flow The most important aspect of fracture traces and lineaments from a hydrogeological viewpoint is the discontinuity in the underlying bedrock associated with the features. Many rock types are almost impermeable when the rock exists as a continuous mass. Therefore, it may be difficult or even impossible to obtain an adequate supply of groundwater in an area underlain by a continuous bedrock mass. However, wells drilled in the same bedrock that intersect natural discontinuities have the potential to provide an adequate water supply for many purposes. The fractures in the rock associated with the fracture traces and lineaments provide the primary locations for the storage and transmission of groundwater. This is particularly true in relatively impervious igneous, metamorphic and limestone rocks that do not contain a significant amount of internal pore space. Therefore, the ability to obtain an adequate yield in a water well drilled in impervious rock generally depends on intersecting one or more fractures that provide the conduit necessary storage and transmission of groundwater. Because fracture traces and lineaments are commonly straight in plan view and unaffected by local topographic relief, these features are considered to be surface manifestations of vertical or near-vertical zones of fracture concentration. Such zones of fracturing are capable of transmitting larger quantities of water than the adjacent less-fractured bedrock. The location, orientation, and extent of lineaments can therefore be used to remotely locate zones of high permeability in the aquifer, and areas of high recharge potential in the unsaturated zone of the aquifer. The increase of groundwater circulation along faults, fractures, and joints leads to greater dissolution of the limestone which increases the size of the subsurface features allowing an even greater amount of water to penetrate the formation. The probability of wells intersecting one or more fractures is increased if the wells are drilled at the intersection of fracture traces or lineaments. This enhanced potential is probably related to two factors. First, a concentration of subsurface rock fractures is believed to occur at the intersection of two linear features. Second, the prospect of the wells intersecting an essentially vertical subsurface fracture is increased when the wells are located at the intersection of two linear features. However, because the subsurface fractures associated with linear features are generally close to vertical, a well drilled along a single linear feature may fail to intersect a fracture and, thus, the well will probably have a small yield that causes it to be unsatisfactory for the drilled purpose. Caves Since caves are typically formed by dissolution of limestone by groundwater, another approach to understanding and verifying groundwater development and its relation to lineaments would be to study cave locations with respect to zones of fracturing for a given area. Wilson (1977) verified that mappable structural features have been the primary influence in the localization of dissolved rock in the Cumber­land Plateau of northeast Alabama. Because zones of fracturing are known to be associated with increased permeability and solubility, the concentration of springs and caves is believed to be greater along faults and joints and their intersections. He compared the location of caves to lineaments visible on aerial photographs and satellite imagery of the area and found that 115 of 149 caves in the study area lie along lineaments. He found that simple caves, those that can be represented in a two-dimensional plan view with less than 100 feet (30 meter) vertical dimensions, tend to form along lineaments. Complex caves, caves with multi-level passages, large chambers, and great depth generally form at intersections of lineaments. Other studies relating cave development to zones of fracturing were conducted by Gregg (1974), Palmer (1975), Ogden & Reger (1977), Wermund et al. (1978), and Barlow & Ogden (1982). III. BARTON SPRINGS SEGMENT OF THE EDWARDS AQUIFER A) Location The Barton-Springs segment of the Edwards aquifer is an ideal location for an investigation of the feasibility of lineament and fracture­trace analysis. This study focuses on the part of the Edwards (Balcones fault zone) aquifer that lies within northern Hays and southern Travis counties in central Texas, for which Barton Springs is the major discharge point. Physiographically, the Balcones fault zone divides the Edwards Plateau in the west and the Blackland Prairie to the east. The Edwards aquifer is comprised of massive, highly-fractured limestone that extends over a distance of about 250 miles (400 kilometers) along a narrow, arc-shaped band that crosses southwestern and central Texas. Outcrops of formations which form the Edwards aquifer occur in parts of ten counties from Kinney, near the Rio Grande River, through Uvalde, Medina, Bexar, Comal, Guadalupe, Hays, Travis, Williamson, and Bell counties to the northeast (Figure 2). 21 Figure 2: The Edwards aquifer in Texas (from Senger & Kreitler, 1984) The areal extent of the Barton-Springs segment of the Edwards aquifer is considered to be bounded on the north by Town Lake on the Colorado River, on the west by its contact with the Glen Rose Formation of the Trinity Group and to the east by the dividing line between fresh and saline water (the ''badwater zone"). The southern boundary is a hydrologic divide near the City of Kyle that separates the Barton-Springs segment of the Edwards aquifer from the San Antonio segment. The study area covers about 155 square miles (401 kilometers2), with most of the northern third of the area generally developed and EXPLANATION mEdwards outcrop ~City of Austin A Barton Springs N t 0 0 Figure 3: The study area (from Senger & Kreitler, 1984) urbanized as part of the City of Austin and several outlying communi­ties. Figure 3 identifies the boundaries of the Barton-Springs segment of the Edwards aquifer as delineated for purposes of this study. The Edwards aquifer is one of the most critical water resources in the state of Texas due to its large usage, large yields, and good water quality. Much of the aquifer lies along the Interstate 35 growth corridor where easily accessible water is especially conducive to expansion at the rural-urban fringe. The Barton Springs/Edwards Aquifer Conservation District (BS/EACD) was created in 1987 to protect the Edwards aquifer which feeds Barton Springs, and is the sole source of drinking water for thousands of people residing within a 255-square mile area that includes parts of Travis, Hays, Bastrop, and Caldwell counties. In 1989, the approximate annual permitted pumpage for water suppliers was over one billion gallons (BS/EACD, 1990). The aquifer provides good-quality water which generally requires only chlorination as treatment prior to delivery. Withdrawals from the aquifer also provide water for industrial, commercial, and agricultural users. These demands for water are projected to increase as the regional population continues to grow and expand. B) Geology Stratigraphy In the study area, the Edwards aquifer is comprised of the Georgetown Limestone and the underlying Edwards Limestone, both of Cretaceous age (Figure 4). The Georgetown Limestone ranges in thickness from 40 to 60 feet (12 to 18 meters) in the subsurface, and consists of thin interbeds of fossiliferous and marly limestones. The Edwards Limestone ranges in thickness from about 300 to 400 feet (91 to Hydrogeologic Age Formation Unit Buda Upper Limestone Confining Bed Del Rio Cla en :::J 0 Georgetown Q) (.) Limestone ~ Q) Edwards - ..... () Limestone Walnut Lower Formation Confining Bed Glen Rose AquiferLimestone Figure 4: Generalized hydrogeologic column (from Slagle et al., 1986) 122 meters) where not weathered, and is composed of thick to thin­bedded rudist limestone, dolomite, nodular chert, and solution collapse breccias that create cavernous secondary porosity (Slagle et al., 1986). The Del Rio Clay forms an upper confining layer of the Edwards aquifer, and is composed of a calcareous, fossiliferous clay that is 60 to 75 feet (18 to 23 meters) thick in the subsurface. The Buda Limestone, stratigraphically above the Del Rio Clay, consists of an upper hard, resistant, shell-fragment limestone and a lower marly, nodular, less­resistant limestone (Slagle et al., 1986). Neither formation is known to yield water in the study area (Slade et al., 1985). The Walnut Formation, which underlies the base of the aquifer, is as much as 60 feet (18 meters) thick, and is composed of a fossiliferous limestone and layers of marl and nodular limestone. The Walnut yields little or no water in the study area and is believed to confine water within the Edwards aquifer. The Glen Rose Formation, stratigraphically below the Walnut, ranges in thickness from 500 to 900 feet (150 to 275 meters) and consists of alternating beds of limestone, dolomite, and marl. The dolomitic members of the Glen Rose are minor aquifers that locally supply small amounts of water containing relatively large sulfate concentrations (Senger & Kreitler, 1984). Structural Geology The large productivity of the Edwards aquifer is a result of early Cretaceous and late Cenozoic karstification, which has been enhanced along fractures and Miocene-age faults (Sharp, 1990). The Cretaceous strata of central Texas dip to the southeast perpendicular to the trend of the Balcones fault zone. The beds on the Edwards Plateau are near horizontal with dips of 10 feet per mile. East of the Balcones fault, the dip becomes more pronounced--approximately 100 feet per mile (McReynolds, 1958). According to Muehlberger & Kurie (1956), the regional dip is thought to be the consequence of three geologic processes: (1) initial dip towards the Gulf of Mexico; (2) subsidence of the basin of the Gulf of Mexico; and (3) uplift of the Edwards Plateau. a) Faults The study area lies along the Balcones fault zone southwest of Austin, Texas. The Balcones fault zone is a series of northeast trending, dip-slip, normal faults which displaces gently, eastward-dipping Cretaceous rocks in this area (Senger & Kreitler, 1984). The Mt. Bonnell Fault is the largest fault in the region and forms the western boundary of the study area. Most of the tectonic events responsible for this fault displacement probably occurred during the Miocene epoch. Tectonism is no longer active along this trend. The formation of the Balcones fault zone is a result of a deeply buried relict structure. The Ouachita orogen lies beneath the area and forms a hinge between the stable continental interior and the subsiding Gulf of Mexico Basin (Clark, 1982). Adjustments across this hinge were probably responsible for the dip-slip dislocations of the Balcones fault system and the majority of joints along the fault zone that have propagated upward from the underlying stress. The overall northeast­southwest trend of the Ouachita orogen is the major determining factor in the orientation of the main bounding faults of the Balcones fault zone system. The Mt. Bonnell Fault is the westernmost major fracture in the Balcones fault zone in Travis and Hays counties. The fault forms a topographic and structural divide between the Edwards Plateau to the northwest and the Gulf Coastal Plain to the southeast. Striking N 40° ­44° E, the fault can be traced from Hays County, through the community of Oak Hill, and then to a point north of Tom Miller Dam in central Travis County. Regional extensional stress existing in the rocks of the Gulf Coast geosyncline has caused faulting along a zone of underlying crustal weakness. Dip-slip movement on the Mt. Bonnell Fault has dropped the upper Edwards into contact with the Glen Rose Limestone. At least 160 feet (49 meters) of throw is transferred from the Mt. Bonnell Fault to the southeast by en echelon left faults, which breaks the downthrown block into a series of grabens and horsts (Balke, 1958). The Balcones fault zone and Luling-Mexia graben conforms to this type of stress system, having a minimum stress direction of N 50° W and near horizontal, maximum stress vertical, and an intermediate stress direction of N 40° E and near horizontal (Dunaway, 1962). Consequently, fractures in the Balcones fault zone have an average strike of N 40° E. b) Joints Based on 3233 measurements of joint orientations in central Travis County, Dunaway (1962) found 30 percent lie within an azimuth of 30° -60°. A secondary peak of 21 percent of the joints lie within an azimuth 120° -150°, a trend roughly perpendicular to the prevailing structural grain in the study area. Specifically, joints in the Edwards Limestone of the downthrown block strike in two prominent directions: N 48° E and N 42° W (Dunaway, 1962). The N 48° E striking joints are tension joints related to the Mt. Bonnell Fault and transverse jointing of the Colorado syncline. The N 42° W striking tension set is probably the result of flexing parallel to the direction throw is transferred from the Mt. Bonnell Fault to the southeast by en echelon faulting. A minor N 10° W striking tension set is probably related to the N 10° W striking joints located west of the fault (Balke, 1958). As observed by Muehlberger & Kurie (1956) and Dunaway (1962), an overwhelming number of joints had vertical or near-vertical dips. C) Hydrology A typical cavernous or karstic aquifer is exemplified by extremely large permeability but low overall porosity. Irregular dissolution of the limestones comprising the Edwards aquifer has created secondary porosity which greatly affects the hydraulic properties of the aquifer. As a result, the aquifer is extremely anisotropic. Significant porosity along particular bedding planes was created through dissolution by meteoric water during an interval of subaerial exposure at the close of the Edwards Limestone period of deposition (Abbott, 1977a). Vertical zones of greater porosity are a result of steep-angle normal faulting that began during the Miocene Epoch (Senger & Kreitler, 1984). At outcrops, these zones allow surface water to readily enter and move through the unsaturated zone to the water table. Extensive faulting, both at the outcrop and throughout the formation, is an important feature of the Edwards. It creates variations in the physical characteristics and dimensions of the aquifer and provides conveyance pathways for surface-water infiltration and groundwater movement, both of which enhance solution cavity development. Consequently, well yields vary tremendously over short distances. A narrow portion of the Edwards aquifer extending along most of the eastern boundary is overlain by the Del Rio Clay, a relatively impermeable formation that functions as a confining layer for groundwater within the underlying aquifer (21 percent of study area). Wells in the study area having the largest yields produce from this confined section, where the wells penetrate the total thickness of the Edwards. In the area west of this confining layer, particularly where the formations of the aquifer crop out, the groundwater in the study area is considered to be generally under free-surface, water table conditions (79 percent of study area). Groundwater movement within the Barton-Springs segment of the Edwards aquifer is from the southwestern and western portions of the aquifer eastward and northeastward, toward Barton Springs on the lower reach of Barton Creek (Figure 5). Historically, hydraulic gradients of the potentiometric surface have ranged from less than 20 to 200 feet per mile. It is estimated that, under "normal" conditions, the bulk of water recharged at Onion Creek would move through the aquifer for EXPLANATION filillillmJ Edwards outcrop -500 Potentiometric surface (ft above m.s.I) ._ General flow direction A Borton Springs • Well location Faults N I I ~ .I I I I 0 2 3 4 I 0 I 2 3 4 5 6 7 8km Figure 5: Generalized groundwater flow in the study area (from Slade et al., 1985) about 3 to 5 years before being discharged through Barton Springs. Barton Springs, located in Zilker Park near the center of Austin, has an average discharge rate of about 50 cubic feet per second (1.4 m3I second) and is currently the fourth largest spring in Texas (BS/EACD, 1990). The minimum discharge was measured in 1956 at a flow rate of 10 cfs (0.3 m3/second). According to Slade et al. (1986), the maximum discharge is 166 cfs (4.7 m3I second). The springs serve as one of the sources of municipal water for the City of Austin's Green Water Treatment Plant on Town Lake. On the average, about 90 percent of the total discharge from the the Barton-Springs segment of the Edwards aquifer occurs through Barton Springs and other associated springs in the immediate vicinity (36,200 acre-feet per year), with the remainder being pumped from wells throughout the aquifer for water supply purposes (BS/EACD, 1990). Recharge The Barton-Springs segment of the Edwards aquifer is recharged primarily by infiltration of surface runoff during storm events into fractures and other openings in the outcrop area of the Edwards and Georgetown Limestones, principally along water courses and streambeds. Creek water entering the recharge zone from the west infiltrates through faults and fractures along the creeks within the recharge zone. Studies by the U.S. Geological Survey (Slade, 1984) show that approximately 85 percent of the recharge into the Edwards in the Austin area occurs in the main channels of Barton, Williamson, Slaughter, Bear, Little Bear, and Onion creeks. Groundwater flow within the Barton-Springs segment of the Edwards aquifer can be summarized as follows: 1) Recharge occurs through fractures along major streams 2) Once in the aquifer, groundwater is channeled northeastward due to the predominance of northeast trending faults and fractures 3) Groundwater discharges at Barton Springs along vertical faults Anisotropic Groundwater Flow Groundwater flow in karst aquifers is very different from flow in granular aquifers. In general, the slow, dispersive, laminar flow assumed by Darcy's Law are seldom in evidence in karst terranes. Most groundwater flow in karstic aquifers like the Edwards is likely to be very rapid, convergent, and turbulent within discrete conduits (Smart & Hobbs, 1986). In fractured rock, groundwater movement is controlled by the distribution, interconnection and orientation of joints, faults, and bedding planes. The zone of saturation consists of these water bearing discontinuities separated by masses of solid rock having relatively much lower permeability (Meiser & Earl, 1982). These fractured aquifers are therefore strongly heterogeneous and anisotropic. Some aspects of fracture flow are: • Irregular, elongate cones of depression produced by wells pumping within fracture zones cause anomalous drawdowns in observation wells. In limestone aquifers, recorded drawdowns of several meters in observation wells 1000 -2000 feet (305 -610 meters) away along fracture zones are not uncommon, while very little drawdown has been noted in nearby observation wells in unfractured bedrock (Meiser & Earl, 1982). • Wells pumping in fracture zones can interfere dramatically with each other when located in the same fracture system; these effects must be addressed when evaluating the sustained yield potential of wells related by obvious fracture traces. • The total recharge area is virtually impossible to define accurately for wells in major fracture zones. Pollutants can travel tremendous distances from sources that are not readily apparent. • Groundwater flow rates along well-developed fracture zones are commonly orders of magnitude greater than flow in poorly­jointed, dense bedrock. This factor must be considered when locating monitoring wells and interpreting data for waste disposal, mining, or other activities affecting groundwater chemistry. • Fracture zone conduits perpendicular to strike of bedding can produce a step-like pattern of groundwater movement in a general downgradient direction. Therefore, strict interpretation of flow normal to "water table" contours tends to oversimplify the complex path of actual groundwater flow. • Groundwater flow in a bedrock fracture zone may vary considerably in the vertical dimension with respect to lithologic changes. For example, a sequence of shale with interbedded sandstones frequently shows larger well yields in the cleanly fractures sandstones than in the denser shales where joint openings tend to be tight (Meiser & Earl, 1982). Karst aquifers are very vulnerable to the effects of chemical spills because of the unique geologic features associated with karst terranes (Field, 1989). Sinkhole development and a lack of recognition of karst hydrological principles can allow chemical contaminants to rapidly infiltrate into the subsurface environment. Vadose storage and flow to phreatic conduits tend to concentrate contaminants for discharge at relatively few points. It is this ability to store and transmit large quantities of highly concentrated chemical contaminants to select discharge points that makes karst aquifers extremely sensitive to the effects of chemical spills. This unusual form of chemical transport and storage leads to very serious threats to human health and the environment. Thus, the ability to identify discrete groundwater flow paths by using lineament analysis will allow the water resource manager to predict groundwater plume migration and subsequent mitigation applications. D) Water Chemistry The quality of water from the Edwards aquifer varies throughout the BS/EACD area. In general, the chemical composition in the aquifer grades downdip from a calcium magnesium/bicarbonate water in the recharge area to a sodium sulfate water and finally to a sodium chloride water deep within the basin (Senger & Kreitler, 1984). This increase in mineralization of the groundwater downdip may be due to intensive faulting which creates numerous barriers to groundwater movement in an easterly direction. In addition to mineralization, the Slade, et al. (1986) reports that poorer quality water in the Glen Rose Limestone may be leaking into the Edwards aquifer, increasing its sulfate and strontium concentrations. Leakage is thought to be associated with large fault displacements, which bring the Edwards Limestone into contact with the Glen Rose Limestone updip. In general, the largest displacements occur along the Mt. Bonnell Fault and in the eastern part of the study area in Hays County and southeastern Travis County. The badwater line, which forms the eastern boundary of the study area, represents a relatively stable hydrochemical boundary separating the two distinct zones of the Edwards aquifer. In the saline "badwater zone," conditions are reducing, as evidenced by the odor of hydrogen sulfide, and the water contains 1000 mg/L to more than 10,000 mg/L of dissolved solids (Clement, 1989). The occurrence of NaCl-type water is related to abundant faults, which create pathways for deep basinal brines and restrict recharge of fresh groundwater from the west. Prezbindowski (1981) explained the water chemistry as being controlled by two processes: (1) mixing of fresh water from the Edwards aquifer moving downdip into the basin with deep saline waters moving up and out of the basin; and (2) dissolution of the Edwards Limestone by undersaturated groundwater moving downdip. A detailed discussion of the badwater zone geochemistry is provided by Clement (1989). IV. METHODS OF DATA ANALYSIS A) Lineament Analysis Determination of Lineaments Lineaments have been mapped in the study area by C. M. Woodruff, Jr., Fred Snyder, and Albert E. Ogden. These lineament maps were used by De La Garza and Slade (1986a) in a previous lineament study and are available from the City of Austin's Department of Environmental Protection. The lineaments were interpreted from mosaics of 1:20,000 black-and-white aerial photographs taken in 1937 by Tobin Aerial Research, Inc. of San Antonio. Each of the mosaics were viewed individually for two 20-minute sessions by each interpreter for a total of 120 minutes of viewing time for each 7.5-minute quadrangle. In order for a linear feature to be mapped, a minimum length of a 1.25 cm (300 meters on the ground) was established. Those lineaments confirmed as natural features were transferred to U. S. Geological Survey topographic maps. During the transfer process from the aerial photographs, errors in lineament orientation are likely to be introduced. Dix and Jackson (1981) 38 assessed the magnitude of these errors in pilot studies on two 7.5­minute quadrangles. Lineaments were transferred onto 1:24,000 topographic sheets and their orientations checked against those on the original photographs. Differences between lineament azimuths on the mosaics and topographic maps lie between 0° and 10°, with a mean of 5°. Because lineaments in this study were grouped into 10° sectors for subsequent analysis, these errors are unlikely to affect the results of the analysis. After the lineament maps were obtained, the endpoints of each linear feature were digitized and a unique identification label was assigned to each of the 938 lineaments and fracture traces identified in the study area (Figure 6). Each identification label also indicated whether the lineament had been identified by two of the three interpreters or by all three observers. One hundred and thirty-six lineaments were identified as "two-man" and 48 were identified as "three-man" lineaments. A FORTRAN program was written to determine the length and azimuth of each lineament, based on the locations of the end points. The statistical tabulation of the lineaments, computed using a program listed in Press et al., (1986), is shown in Table 1. The lineament length data was evaluated for the statistical parameters of deviation, variance, skewness, and kurtosis. The method of interpretation of these parameters is not to attach a significance to each statistical value, but is used in comparison with other samples. For example, the sample having the largest variance or standard deviation, consequently, has the Figure 6: Location of lineaments in the study area (detailed lineament map is included in map pocket) largest distribution among the values of the observations, provided all the measurements are made in the same units. Statistical Analysis The lengths and orientations of lineaments in the study area were processed in order to identify preferred orientations (peaks) and to test whether these peaks are statistically significant and, therefore, geologically meaningful. Lineament and fracture trace azimuths were divided into 18 10° sectors. Then, the total lineament length in each AZIMUTH ST A TISTICS Maximum: 179.9° Minimum: 0.7° Arithmetic Mean: 39.9° LENGTH ST A TISTICS Maximum: 5332.9 m Minimum: 310.7 m Arithmetic Mean: 1096.0 m Average Deviation: 463.3 m Standard Deviation: 639.0 m Variance: 408317.5 m2 Skewness: 1.9 Kurtosis: 5.2 Table 1: Summary of selected statistics for azimuths and lengths of digitized sector (Ls) was calculated and summed for the overall total lineament length (Lt). The relative length (Lr) was determined by: (1) Next, the length-weighted frequency (F) of the azimuth data was calculated. As described by Dix & Jackson (1981), this parameter expresses the total lineament length in a 10°-wide sector of the graph, weighted in proportion to the number of lineaments (n) in the area in question: Ls x n F=~ (2) Length-weighted frequency is used to combine lineament length and number of lineaments into a single parameter (Baumgardner, 1987). The advantage in using this measure is that values from different areas can be compared while allowing differences in number of lineaments in each area. For this study, a peak is defined as any 10°-wide sector with a magnitude larger than the average for that graph. The "peakedness" of a graph is affected by the number of lineaments in the sample. Dix and Jackson (1981) devised a measure of peakedness called the index of preferred orientation (IPO): 18 L I Lr -0.05 I x100 i=l IPO =--------(3) 1.8 They observed that values of IPO for computer-generated, geologically meaningless, random "lineaments" decreased as sample size increased. The decrease in value was rapid from 50 to 200 lineaments but slowed as the number of lineaments increased above 200. As a result, Dix and Jackson (1981) proposed that data sets should contain at least 200 lineaments to "minimize the effects of randomly oriented lineaments on geologically significant trends." To determine which greater-than-average peaks were significant, a chi-square test was used to measure the difference between each peak and the mean F value (Siegel, 1956). The x2 technique tests whether the observed frequencies are sufficiently close to the expected ones to be likely to have occurred under the null hypothesis. Dix and Jackson (1981) concluded that the 99-percent confidence level (p =0.01 level) should be used to define geologically meaningful peaks because none of their samples with more than 100 computer-generated "lineaments" had significant peaks at that level. Because a circular-normal distribution cannot be assumed, the x2 one-sample nonparametric test was applied. This test requires the use of lineament frequency rather than magnitude. To accommodate this requirement, the length­weighted frequencies were used as follows: k (F-f )2 x2 = I--==--(4) i=l F where F = arithmetic mean of length-weighted frequency In general, the larger xi, the more likely it is that the observed frequencies did not come from the population on which the null hypothesis is based. By dividing the X2 value for each peak by the degrees of freedom (u =k -1, where k equals the number of 10° sectors forming the peak), the Bemshtein accuracy criterion, H, was determined (Vistelius, 1966): xi H=­(5) u This parameter serves as a check on the actual existence of a preferred orientation. If H ~ 2, then it can be presumed that the initial hypothesis is not consistent with the observations. With H ~ 2, the initial hypothesis is not contradicted by the observed distribution. B) WellAnalysis Productivity of Water Wells The yield of a well may be expressed in terms of its specific capacity which is defined as the yield in gpm/ft of drawdown for a stated pumping period and rate. The specific capacity of a well is affected by the hydraulic properties of the aquifer (particularly the coefficients of transmissivity and storage), by the radius of the well, the pumping period, and the depth of saturated rock penetrated by the open section of the well bore (Siddiqui, 1969). The transmissivity of the aquifer has the greatest influence on the specific-capacity value of a well; as this increases so does specific capacity. However, the specific capacity of a well cannot be an exact criterion of the coefficient of transmissivity because specific capacity is often affected by storativity, partial penetration, well loss, well diameter, and hydrologic boundaries (Walton, 1988). In most cases these factors adversely affect specific­capacity values and the actual coefficient of transmissivity is greater than the value computed from specific-capacity data. Nevertheless, specific capacity is a valuable measure of the productivity of a well. Collection of Well Data In order to examine the relation between productivity of water wells and proximity to lineaments, the specific capacities of 27 wells in the study area were obtained. This data was acquired from several sources. Four of the specific-capacity values are the results of specific hydrogeologic investigations completed on wells in the study area. Due to the scarcity of pump-test data in the Edwards, these four studies provide critical information about the hydraulic properties in the aquifer and are summarized in detail in Appendix A. During the field investigations, six other specific-capacity values were determined. A 300-foot (91 meter) long electric probe (e-line) was utilized to determine the water level in the well. After measuring the static water level, the pump was energized and the subsequent drop in the water level was measured with the e-line. When the water level reached equilibrium (typically 15 -45 minutes in the Edwards), the discharge was measured using a flow meter. The specific-capacity value was calculated as the drawdown in feet divided by the discharge in gallons per minute. Unfortunately, several problems were encountered during the collection of field data. Many wells do not have an adequate access port at the wellhead for thee-line probe. In addition, due to the depth to water in the wells (typically 175 -250 feet) and the poor condition of the wells in general, obstructions in the wellbore were frequently encountered in several wells. Three wells had pumping water levels deeper than 300 feet which prevented the drawdown measurement. Specific capacities for two wells were determined using a permanently installed airline. This line, typically ~ inch (0.3 cm) in diameter, ran down the length of the well to the depth of the pump. The line was pressurized by either a compressor or a hand pump and the initial pressure reading in pounds per square inch was recorded. The pressure meter measured the amount of pressure at the pump caused by the column of water in the well. After the pump was activated, the resulting drop in the water level is reflected by a corresponding drop in the airline pressure. The difference in pressure between the static water level and the pumping water level was used to calculate the drawdown by using the relationship of 2.31 feet (0.7 meters) of drawdown to one pound of pressure. Historical data from either prior pump tests or drilling logs was used for the remainder of the specific capacities. Data derived from pump tests is more reliable because discharge and drawdown measurements are made over time. Drilling logs from all located and plotted wells in the study area were reviewed at the Texas Water Development Board's Central Records office. In the Edwards aquifer, the overwhelming number of wells do not have reliable discharge and drawdown data. According to a local driller, Byron Benoit of Associated Drilling in Manchaca, Texas, most well yields on new wells drilled in the Edwards are estimated by air-jetting. Thus, in the karstic and faulted Edwards, much of the air and water is blown into solution cavities along the wellbore. In addition, using data gathered from driller's logs is not as reliable because it is not known whether or not steady-state conditions existed before water-level declines were measured. Occasionally, however, discharge and drawdown measurements are collected during drilling procedures and can be used as a indicator of the relative productivity of a well. Well locations, specific capacities, and the method of determination are tabulated in Table 2. Several researchers (Brook et al., 1984) propose normalizing specific-capacity data in order to accurately compare values between different studies. Lattman and Parizek (1964) recommend that the specific-capacity values be divided by the total depth of saturated rock penetrated by each well to obtain what LaRiccia and Rauch (1977) designate as the specific-capacity index. However, due to the extreme vertical faulting of the Edwards aquifer, the thickness and water levels are highly variable. In addition, further manipulation of the basic specific-capacity data reduces the viability of the data due to contrived Well Number Owner Latitude Longitude Date of Discharge Drawdown Method (ddmmss) (ddmmss) Test (gpm) (feet) 'rj 58-42-812 W. F. Guyton & Assoc. 301554 974847 Jun-69 20.00 1.50 s a 58-42-821 Trigg-Forister Bldg. 301540 974838 Feb-82 16.00 10.40 p ii 58-42-8M Allen Keller Co. 301533 974747 Jun-79 100.00 60.00 s N 58-42-8S Espey Huston & Assoc. 301618 974908 Apr-82 150.00 6.00 D (/) "O 58-49-9H Charles Ranch 300933 975354 Aug-87 287.00 138.00 D Cl> 58-50-223 City of Sunset Valley 301339 974835 Jun-90 125.00 49.65 A g c=;· 58-50-414 Lee V. Johnson 301047 975027 Nov-86 51.00 18.00 D 0 58-50-704 Marbridge Found. #5 300812 975120 Feb-68 1150.00 31 .00 s Ill "O 58-50-731 Shady Hollow Estates 300858 975136 May-83 210.00 10.00 p Ill 0 58-50-830 Slaughter Cr. Acres 300937 974904 Aug-71 45.00 160.00 s ~ 58 -50-835 Onion Creek CC 300845 974845 May-69 270.00 12.00 s a. 58-50-8A Native Texas Nursery 300915 974915 Jul-90 36.00 118.96 p a Ill 58-57-307 Dahlstrom Middle Sch. 300559 975256 May-90 68.00 18.34 E ii)" cr 58-57-910 Mt. City Oaks WSC 300205 975357 Jul-90 184.00 0.25 E c: p 58-58-102 Cimarron Park #2 300622 975115 Apr-84 600.00 4.00 §I 58-58-115 Estate Utilities WSC 300723 975219 Nov-79 660.00 12.00 s Cl> a. 58-58-123 Elizabeth Porter 300634 975030 Feb-85 400.00 15.00 D 0 58-58-1 A Frank Burdette 300726 975217 Jun-90 10.75 0.29 E ""' ~ 58-58-1 B Hays Hills Baptist Ch. 300702 975228 Jun-90 71.00 7.00 p ~ 58-58-1 EE Neptune-Wilkinson 300504 975200 Apr-84 225.00 127.00 D (Ji 58-58-202 Mystic Oak WSC #1 300728 974848 unk 42.00 185.00 s ::J p 58-58-2E Hunter Industries 300537 974852 Nov-89 200.00 117.00 s- Cl> 58-58-406 Texas-Lehigh Cement 300341 975120 Aug-66 1200.00 53.00 D 5!?. 58-58-412 Plum Creek WSC 300435 975003 Jun-90 470.00 52.60 E c: '< a. 58-58-413 City of Buda #3 300420 975004 Mar-87 430.00 99.33 p Ill 58-58-506 Goforth WSC 300442 974949 Sep-77 310.00 65.00 p Cl> ""' 58-58-508 Goforth WSC 300443 974950 Jul-90 227.00 90.90 A Ill Methods: E: Field Measurement (E-line) A: Field Measurement (airline) D: Data from drillers logs S: Data from Slade & De La Garza (1986) P: Data from previous pump tests """ assumptions. The inhomogenous, anisotropic nature of the Edwards aquifer prevents the utilization of typical assumptions that could be used to translate specific-capacity data into values of transmissivity or permeability (as in Meyer, 1963) except for site-specific applications. Statistical Analysis In order to test for statistical validity of the collected data, the controlling factors must be defined and isolated and their relative importance established. In carbonate or other fractured aquifers, for example, the geologic factors influencing porosity and permeability distribution, and hence the range in well yields, are frequently not known and should be defined. Appropriate statistical techniques may be used to draw conclusions about the population from the evidence provided by the sample data. First, the distribution of the data sets must be determined. The 27 specific-capacity values collected from the study area range in value over four orders of magnitude. As expected, regression analysis failed to show a Gaussian distribution. The data was also subjected to the Lilliefors Bound Test. At the 95%-confidence level, the specific-capacity data plotted outside of the bounds of normal distribution. This variability in values for specific capacity and distance is illustrated in box diagrams (Figure 7). Consequently, nonparametric or distribution free statistical tests were conducted because the data were not normally distributed. 50 Percentile 10 25 50 75 90 0 10 20 30 40 50 60 Distribution of Specific-capacity Values (gpm/ft) Percentile 10 25 50 75 90 0 200 400 600 800 1000 Distribution of Distances from Well to Nearest Lineament (feet) Figure 7: Box diagrams of distribution for values of specific capacities and distances from wells to nearest lineament Nonparametric statistical tests do not require the use of normally distributed data nor does the shape of the distributions need to be known. Thus, a small number of observations can be used and computations are fairly simple. A further advantage of nonparametric techniques is that they may be used with data that are not exact in any numerical sense, but that in effect are simply ranks. Siegel (1956) presents these techniques and provides examples from behavioral sciences. The Mann-Whitney U Test is used to test whether two independent samples have been drawn from the same population (Siegel, p. 116). For this study area, it is an ideal technique for determining if the difference between specific capacities of wells located to the southeast of a NE-SW trending lineament are random data or structurally controlled. Thus, the null hypothesis to be tested is: H0 ­direction of a well from a lineament has no effect on specific capacity. The alternative hypothesis is H1: direction of a well from a lineament has a significant effect on specific capacity. The statistic U used in this test, is calculated by the following equations: U1 = n1n 2 + 1 [n1(n1 +1)]-R1 (6) 2 1 U2 = n1n 2 + 2 [n2(n1 + 1)] -R2 (7) where R1 =sum of the ranks assigned to the group whose sample size is n1; R2 =sum of the ranks assigned to the group whose sample size is n2; and U1 + U2 = n1n2. The smaller of the two values, U1 and U2, is the U used in the test. This U value is compared to the theoretical U at a particular significance level. If the probability associated with the observed value of U for the sample sizes n1 and n2 is equal to or less than the previously set level of significance a, the null hypothesis is rejected. A detailed explanation with hydrogeological applications of this technique is given by Siddiqui & Parizek (1972). For this study's sample size, n1 and n2, the significance of an observed value of U is determined by referring to Table Kin Siegel (1956, p. 276). For a significance level of 0.05 (that is, a 95%-confidence level), the probability associated with the observed value of U is equal to 45. Therefore, if the calculated value of U is less than 45, the null hypothesis is rejected. Table 5 gives the specific-capacity values listed in ascending order of magnitude, direction of the well from the nearest lineament, their ranks, the ranks for the two groups, the sum of these ranks, and the number of observations in the two groups. C) Water Chemistry One sample was collected and analyzed from each of 61 wells that extract water from the study area. A basic chemical survey of anions, cations, and carbon species was conducted on these samples in the Department of Geological Sciences laboratories as described in the following sections. Results from the chemical analyses are tabulated in Table 3. In addition, 20 water samples were collected with funding from the Texas Water Development Board. The Lower Colorado River Authority's Environmental Laboratory tested for 34 parameters including radioactivity and total organic carbon. The results are compiled in Appendix B. Field-Data Collection The water samples were collected over a two-month period during the summer of 1990. The sampling point was located as close to the wellhead as possible and always before the water flowed through pressure or storage tanks, chlorinators, softeners, or any filtering apparatus. Groundwater was pumped until a constant pH reading was obtained to ensure flushing of the well bore. Three parameters were measured in the field for each sample collected: pH, water temperature, and conductivity. For this study, the pH and temperature were measured using an Orion Research model SA250 pH meter. The meter was calibrated daily with standard buffers of pH 4 and 7. In most cases, the water was routed through a flow cell to simulate in-situ measurements. Because of its critical effect on pH, temperature measurements were recorded simultaneously using the temperature probe of the SA250 pH meter. Conductivity values were obtained by using a Hach model 44600 Conductivity meter. Values for total-dissolved solids (TDS) were calculated from the concentrations of the major anions and cations of each of the collected samples. Ratios of IDS to conductivity ranged from 0.5 to 0.8 with a mean ratio of 0.66. The probe was rinsed with deionized water after each sample to prevent contamination of subsequent measurements. At each well, two water samples were collected in nalgene plastic bottles for lab analysis. Both sample bottles were rinsed and filled with filtered water using a 0.45 micron (µ) membrane filter. A 60 milliliter (ml) sample was collected for the carbon and anion analyses. No preservative was added to this sample. For the cation analysis, a 30 ml sample bottle was preserved with 2 drops of nitric acid (HN03). All samples bottles were transported on ice and stored in a refrigerator until analysis. Laboratory Analysis The amount of dissolved inorganic carbon (DIC) for each sample was determined with a Dohrmann DC-180 Carbon Analyzer. After calibrating the instrument with a 50 ppm carbon standard, 500 microliters (µl) of water sample are injected by syringe. The sample is then delivered to an ultraviolet reactor for oxidation. During an analysis, the non-dispersive infrared C02 detector (NDIR) produces electrical output peaks which are integrated and displayed in ppm carbon concentration units. Using the resulting value of carbon, alkalinity was calculated by multiplying the first ionization constant, (a), by the the corresponding known value of DIC. The ionization constant is calculated as follows: H+ KzJ-1 (8) al = [ Ki + 1 + H+ where Kl is equal to J0-6.4 and K2 equals 10-10.3 at 25° C. However, since the water samples were not collected at 25° C, the equilibrium constants must be adjusted to temperature using this formula: (9) In this case, temperatures are in degrees Kelvin with T2 equal to 298° K (25° C) and T1 is the temperature at the time of collection. The resulting Ki value is substituted into Equation 8 and used to determine alkalinity. Bicarbonate values were calculated by multiplying alkalinity by the formula weight of HC03. Similar to the carbon analysis, anions in the groundwater were also determined from filtered, unacidified samples. Samples were analyzed using a Waters Ion Chromatograph for the following anions: flouride, chloride, nitrite, bromide, nitrate, phosphate, and sulfate using EPA method A-1000 (proposed). This method utilizes a 150 x 4.6 mm IC­Pak A HC column and a borate/gluconate eluent. Detection was by both conductivity and ultraviolet-absorption methods. High-and low­concentration working standards were interspersed among the water samples to ensure the accuracy of the analysis. Cations for all samples were determined using a JY-70 inductively-coupled atomic emission plasma spectrometer. Details of this process can be obtained from the instrument manual. Working standards and duplicate samples were interspersed to check calibration of the instrument. Cations analyzed include: aluminium, barium, calcium, chromium, iron, potassium, magnesium, sodium, lead, strontium, and zinc. The charge-balance error (E) for each water sample was calculated using the equation: Izmc -Izma E -x 100 (10) -Izmc + Izma where z is the ionic valence, me the molality of cation species, and ma the molality of anion species. Field Data Carbon Species Anions Well Number Owner Date Temp. pH Cond. TDS DIC Alkalinity HC03 Cl N03 504 1990 (oC) (µS) (mg/I) (ppm C) (meq/I) (mg/I) (mg/I) (mg/I) (mg/I) 58-58-1 A Frank Burdette 11-Jun 23.6 7.27 552 395 81 5.941 362.41 11 .56 6.91 7.72 58-58-115 Estate Utilities WSC 11 -Jun 23.6 7.30 528 391 82.35 6.088 371.36 9.76 5. 71 5.20 58-58-508 Golorlh WSC 12-Jun 25.2 7.30 653 446 64.95 4.802 292.90 10.85 118.26 58-58-412 Plum Creek WSC 12-Jun 24.3 7.20 662 452 77.7 5.584 340.64 10.81 0.93 86.60 58-58-102 Cimarron Park WSC 13-Jun 22.4 7.41 477 347 65.49 4.964 302.79 10.90 6.37 21 .44 58-58-114 Cimarron Park WSC 13-Jun 22 7.44 518 384 74 5.642 344.14 10.78 8.00 17.42 58-58-117 Twin Oaks Ranch 13-Jun 23.3 7.59 494 351 63.21 4.936 301 .11 9.37 0.32 32.55 58-58-407 Texas-Lehigh Cement 13-Jun 24.9 7.59 661 449 73.57 5.745 350.47 10.75 0.87 77 .03 58-58-416 Comal Tackle Co. 14-Jun 23.8 7.28 549 401 77.23 5.680 346.47 13.63 5.47 18.42 58-50-855 Village ol San Leanna 26-Jun 25 7.39 615 428 67.73 5.113 311 .86 12.68 87 .27 58-50-223 City ol Sunset Valley 26-Jun 23.7 7.05 602 416 89.52 6.093 371 .66 12.06 12.13 10.45 58-49-911 Chaparral Park #2 27-Jun 24.7 6.83 800 568 86.74 5.271 321 .55 13.44 2.06 157.46 58-50-847 Creedmoor-Maha #2 27-Jun 23.9 7.11 577 387 72.85 5.077 309.71 10.51 4.89 46.14 58-49-918 Chaparral Park #4 2-Jul 24 6.90 793 486 88.52 5.604 341 .87 16.12 2.89 62.12 58-49-915 Chaparral Park #3 2-Jul 23.1 6.84 831 533 88.62 5.419 330.55 25.37 2.24 100.63 58-50-852 J. D. Malone 3-Jul 24.3 7.29 643 414 56.96 4.200 256.21 19.73 1.19 84 .37 58-58-413 City ol Buda #3 5-Jul 26 7.30 677 439 59.35 4.388 267.64 9.57 116.99 58-58-106 City of Buda #2 5-Jul 23.5 7.24 573 380 68.67 4.995 304.69 10.35 3.86 39.82 58-58-403 City of Buda #1 5-Jul 23 7.12 583 378 72.6 5.079 309.79 10.31 5.68 27.38 58-50-830 Slaughter Cr. Ac. #1 5-Jul 24.4 7.15 614 395 59.32 4.194 255 .83 13.36 1.27 82.70 58-50-829 Slaughter Cr. Ac. #2 5-Jul 24.6 7.20 664 421 57.3 4.118 251 .21 13.91 112.59 58-57-901 Hays High School 9-Jul 23.7 7.40 506 347 64 .63 4.889 298.21 8.35 3.48 16.07 58-57-804 Michaelis Ranch 9-Jul 24.9 7.22 617 420 70.96 5.131 313.00 12.01 7.60 51.42 58-50-731 Shady Hollow Estates 9-Jul 22.8 7.08 550 375 72.66 5.006 305 .37 11 .89 3.85 17.42 58-58-219 Pool & Rogers Co. 10-Jul 24.8 7.32 803 554 56.67 4.210 256.83 41 .67 154.95 58-58-202 Mystic Oaks WSC #1 11 -Jul 24.6 7.27 792 629 61 .91 4.541 277.00 218.85 58-58-216 Mystic Oaks WSC #2 11 -Jul 24.2 7.09 1047 611 60.04 4.153 253.32 229.58 58-50-724 Manchaca Fire Dept. 11-Jul 22.7 7.14 527 349 62.68 4.416 269.38 9.62 3.46 31.27 58-57-910 Mt. City Oaks WSC 12-Jul 21 .8 6.86 558 345 68.43 4.235 258 .32 9.68 4.94 16.72 58­57-9M Leo Miller 12-Jul 23.4 7.05 570 371 73.97 5.034 307 .10 11 .43 5.53 16.16 58-57-205 Don West Ranch 14-Jul 25.5 6.80 665 424 91 .6 5.461 333 .14 11 .72 4.57 32.21 58-57-204 Don West Ranch 14-Jul 25.9 6.81 718 456 96.1 5.767 351 .78 10.96 19.51 15.27 Table 3: Chemical analysis from water samples in the study area Vl °' Field Data Carbon Species Anions Well Number Owner Date Temp. pH Cond. TDS DIC Alkaflnlty HC03 Cf N03 S04 1990 (oC) (µS) (mg/I) (ppm C) (meq/I) (mg/I) (mg/I) (mg/I) (mg/I) 58-57-202 Don West Ranch 14-Jul 25.6 6.96 537 331 67.99 4.442 270.94 9.79 0.54 18.87 58-57-607 Buda/Kyle Church 16-Jul 23.9 7.15 651 391 72.39 5.118 312.20 13.20 6.20 18.57 58-42-821 Trigg Building 16-Jul 22.3 7.22 528 362 57.64 4.168 254.25 22.50 2.38 35.34 58-50-733 Suburban Austin WSC 17-Jul 24.4 7.01 549 360 69.21 4.629 282.39 11 .53 5.85 19.33 58-49-922 Copper Hills #2 17-Jul 25.6 7.04 515 345 73.57 4.986 304.15 3.40 0.73 10.80 58-50-838 Village of San Leanna 17-Jul 25.6 7.10 611 367 59.55 4.135 252.22 1.86 74.65 58-58-208 Suburban Austin WSC 17-Jul 24.4 7.15 676 427 57.9 4.094 249.71 9.93 115.51 58-50-729 VFW Post #3377 18-Jul 24.1 7.21 548 350 61 .62 4.442 270.98 11 .54 7.04 34.03 58-58-510 Crestview R.V. Center 18-Jul 25.9 7.29 720 451 55.26 4.075 248.56 31 .25 0.11 109.08 58-58-218 AAA Petroleum Dist. 18-Jul 25.9 7.01 829 523 75.56 5.054 308.30 36.19 110.96 58-50-520 Mr. Herb Mendieta 18-Jul 24.4 7.08 559 358 71 .27 4.910 299.52 11 .34 6.68 16.85 58-42-913 Park Hill Baptist Ch. 19-Jul 23.7 7.11 643 436 80.42 5.605 341.89 19.33 8.56 23.66 58-49-9B SW Territories #5 23-Jul 25. 7 7.08 695 466 76.42 5.265 321 .17 7.59 1.09 89 .25 58-50-704 Marbridge Found. #5 24-Jul 24.3 7.12 538 358 68.64 4.801 292.89 10.68 3.48 21 .04 58-50-703 Marbridge Found. #2 24-Jul 25.1 7.08 536 362 71 .05 4.895 298.60 11 .22 5.45 19.87 58-50-854 St. Albans Epis. Ch. 30-Jul 26.5 7.06 3160 1991 56.68 3.874 236.30 545.20 598.70 58-58-2E Hunter Industries 30-Juf 27.6 7.00 759 459 54.51 3.630 221 .40 24.15 139.97 58-50-7E Mr. Richard McKeane 31-Jul 23.6 7.06 619 406 75.12 5.134 313.17 19.62 4.22 22.71 58-50-859 Onion Creek Mam. 31 -Jul 25.3 7.27 628 388 58.45 4.287 261 .52 10.17 81 .35 58-50-519 Mr. Don West 31-Jul 26.5 7.07 759 475 55.36 3.799 231.73 31 .85 0.41 129.57 58-50-835 Onion Creek CC 2-Aug 26.5 7.20 642 384 59.06 4.245 258.92 15.00 3.62 74 .66 58-50-843 Oak Forest Highlands 2-Aug 25.7 7.03 643 384 57.43 3.876 236.41 11 .11 2.53 97.81 58-50-4R Mr. Tom Roudebush 2-Aug 25.3 7.09 597 391 74.41 5.147 313.95 24.70 9.01 10.47 58-50-726 Ms. Jane Pratt 6-Aug 25.9 7.28 540 358 65.6 4.825 294.30 12.38 6.48 18.71 58-50-858 Twin Creeks WSC 6-Aug 25.9 7.07 567 351 60.45 4.148 253.04 12.01 7.58 45.53 58­50-8M Mooreland Water Co. 7-Aug 25.3 6.98 574 370 69.65 4.594 280.26 17.26 4.67 34 .27 58-58-1 H Hays Hills Baptist Ch. 7-Aug 25.1 7.25 539 384 66.88 4.879 297.59 11.84 3.49 17.72 Barton Springs City of Austin 9-Aug 21 .1 6.60 690 393 69.33 3.546 216.28 45.72 8.37 39.90 Table 3 (cont.): Chemical analysis from water samples in the study area V1 'l Cations Well Number Owner Al Be Ca Cr Fe K Mg Na Pb Sr Zn Charge (mg/I) (mg/I) (mg/I) (mg/I) (mg/I) (mg/I) (mg/I) (mg/I) (mg/I) (mg/I) (mg/I) Balance(%) 58-58­1A Frank Burdelte 0.041 0.045 79.76 0.010 0.556 23.55 5.986 0.148 0.205 0.003 12.31 58­58­115 Estate Utilities WSC 0.040 0.041 70.28 0 .012 0.429 28.76 5.269 0.127 0.265 0.005 11.87 58-58­508 Goforth WSC 0.056 0.090 61 .38 0 .018 0.028 1.685 35.53 8.545 0.187 0.418 1.67 58­58­412 Plum Creek WSC 0.045 0.209 71 .14 0 .014 1.238 30.82 6.707 0.187 0.434 2.41 58­58-102 Cimarron Park WSC 0.028 0.065 61 .08 0 .008 1.133 23.86 6.220 0.124 5.438 10.52 58­58-114 Cimarron Park WSC 0.029 0.038 68.71 0.010 1.034 25.66 6.167 0.128 0.739 0.004 9.63 58­58-117 Twin Oaks Ranch 0.043 0.043 57.49 0 .013 1.571 29.11 5.656 0.132 6.816 0.114 12.15 58-58-407 Texas-Lehigh Cement 0.048 0.198 71 .19 0.014 1.240 31 .10 6.638 0.190 0.401 3.30 58­58-416 Comal Tackle Co. 0.038 0.044 87.08 0.007 1.262 20.36 7.393 0.164 0.272 0.010 13.02 58-50-855 Village of San Leanna 0.050 0.092 63.52 0 .013 1.886 29.79 9.946 0.173 0.399 2.35 58­50-223 City of Sunset Valley 0.050 0.345 74.86 0.014 1.413 30.35 9.229 0.161 0.964 0.013 13.49 58­49-911 Chaparral Park #2 0.077 0.151 100.20 0.029 5.398 51.62 8.154 0.212 8.409 0.016 14.28 58-50-847 Creedmoor-Maha #2 0.046 0.131 69.66 0.012 1.547 26.41 6.952 0.159 0.357 9.43 58-49-918 Chaparral Park #4 0.080 0.057 107.80 0 .023 2.381 42.98 7.452 0.203 2.324 0.014 23.51 58-49-915 Chaparral Park #3 0.081 0.100 106.90 0 .023 4.294 43.40 14.470 0.233 0.387 17.93 58-50-852 J . D. Malone 0.069 0.074 65.37 0.017 3.768 33.60 22.480 0.179 0.381 14.27 58-58-413 City of Buda #3 0.083 0.094 74.86 0 .020 0.229 2.202 37.35 6.569 0.218 0.413 10.24 58­58-106 City of Buda #2 0.044 0.239 71 .51 0 .011 0.045 1 .084 28.82 6.479 0.157 0.304 13.67 58­58-403 City of Buda #1 0.045 0.169 79.37 0 .011 0.042 1 .030 26.89 6.306 0.157 0.023 17.18 58-50-830 Slaughter Cr. Ac. #1 0.064 0.084 69.21 0.016 0.043 1.765 32.34 11 .070 0.172 0.387 13.05 58­50-829 Slaughter Cr. Ac. #2 0.064 0 .091 68.79 0.017 0.089 1.724 33.69 10.760 0.182 0.403 8.46 58­57-901 Hays High School 0.055 0 .059 68.98 0 .011 0.046 1.022 30.02 5.488 0.146 2.586 0.020 20.52 58-57-804 Michaelis Ranch 0.071 0.140 75.77 0.019 0.049 2.442 39.93 7.325 0.176 0.262 17.86 58-50-731 Shady Hollow Estates 0.054 0 .053 91 .40 0.009 0.043 0 .868 24 .48 6.662 0.172 0.460 0.047 22.86 58­58-219 Pool & Rogers Co. 0.073 0 .055 61.22 0.018 0.423 6.545 36.91 68.410 0.176 0.335 11 .13 58-58-202 Mystic Oaks WSC #1 0.087 0.050 79.79 0 .024 0.073 8.935 46.17 77.730 0.208 0.340 18.76 58-58-216 Mystic Oaks WSC #2 0.063 0 .038 70.56 0.024 0.135 9.550 48.70 71.900 0.181 0.346 16.97 58­50-724 Manchaca Fire Dept. 0.071 0.134 75.15 0.012 0.047 2.130 28.04 6.421 0.172 0.259 21 .53 58­57-910 Mt. City Oaks WSC 0.039 0 .049 92.21 0.002 0.051 0.774 18.07 5.699 0.168 0.198 0.001 26.17 58­57-9M Leo Miller 0.047 0 .049 84.93 0.009 0.052 0 .965 26.35 6.646 0.157 0.362 0.123 21 .59 58-57-205 Don West Ranch 0.070 0.086 90.53 0.019 0.055 2.344 38.53 6.094 0.199 0.221 23.44 58­57­204 Don West Ranch 0.059 0 .080 123.00 0.010 0.054 0 .326 28.86 6.394 0.210 1.449 0.006 27.19 Table 3 (cont.): Chemical analysis from water samples in the study area (Jl Cations Well Number Owner Al Ba Ca Cr Fe K ~ Na Pb Sr Zn Charge (mg/I) (mg/I) (mg/I) (mg/I) (mg/I) (mg/I) (mg/I) (mg/I) (mg/I) (mg/I) (mg/I) Balance(%) 58-57-202 Don West Ranch 0.052 0.120 66.23 0 .012 0.054 3.783 32.79 5.353 0.151 7.338 0.021 25.66 58-57-607 Buda/Kyle Church 0.056 0.052 99.06 0.007 0.070 1 .257 21.95 7.550 0.178 0.244 0.019 22.43 58-42-821 Trigg Building 0.049 0.055 83.97 0.007 0.053 0.972 22.74 12.390 0.154 0.441 0.121 20.46 58-50-733 Suburban Austin WSC 0.050 0.055 89.42 0 .009 0.048 0 .857 24.13 6.885 0.165 1.213 0.011 24.47 58-49-922 Copper Hills #2 0.061 0.084 75.43 0.014 0.046 1.471 33.77 2.499 0.161 1.704 0.018 26.74 58-50-838 Village of San Leanna 0.066 0.148 71 .82 0.011 0.053 1.482 29.94 7.338 0.194 0.397 17.14 58-58-208 Suburban Austin WSC 0.077 0.095 75.14 0.017 0.070 1 .909 38.20 7 .521 0.214 0.397 12.97 58-50-729 VFW Post #3377 0.050 0.151 69.14 0 .006 1.321 25.74 7.236 0.154 0.278 15.63 58-58-51 O Crestview R.V. Center 0.052 0.033 57.09 0.012 4.739 34.40 37.050 0.146 0.349 10.03 58-58-218 AAA Petroleum Dist. 0.052 0.056 71 .72 0.016 4.110 37.68 42.030 0.167 0.324 10.78 58-50-520 Mr. Herb Mendieta 0.033 0.146 77.26 0.007 0 .565 25.25 6.500 0.126 2.751 19.00 58-42-913 Park Hill Baptist Ch. 0.047 0.086 111 .60 0.004 0.767 19.99 8.175 0.171 0.1 82 0.013 18.93 58-49-9B SW Territories #5 0.073 0.116 85.64 0.018 3.740 43.42 5.631 0.181 8.085 0.108 17.77 58-50-704 Marbridge Found. #5 0.046 0.053 86.07 0.002 0 .982 20.37 6.279 0.153 0.333 0.038 19.57 58-50-703 Marbridge Found. #2 0.046 0.047 79.17 0.006 0 .783 26.10 5.960 0.149 0.308 0.064 19.38 58-50-854 St. Albans Epis. Ch. 0.288 0.052 168.04 0.020 18.320 90.16 402.000 0.428 22.516 0.016 5.83 58­58-2E Hunter Industries 0.073 0.061 70.17 0.016 0.515 2 .811 38.42 24.360 0.178 0.351 11.71 58­50­7E Mr. Richard McKeane 0.049 0.084 104.60 0.004 1.166 18.77 11 .190 0.172 1.836 0.002 21.23 58-50-859 Onion Creek Mem. 0.065 0.136 71 .99 0.011 1.544 28.79 7.578 0.199 0.821 11.37 58-50-519 Mr. Don West 0.068 0.038 55.55 0.013 6.243 34.58 51 .460 0.173 0.342 11 .90 58-50-835 Onion Creek CC 0.047 0.134 66.83 0.006 1.033 28.22 10.080 0.175 0.332 9.46 58-50-843 Oak Forest Highlands 0.050 0.096 64.49 0.009 1.057 30.33 8.013 0.175 0.324 8.31 58-50-4R Mr. Tom Roudebush 0.043 0.064 80.98 0.010 0.309 31 .89 9.419 0.143 0.296 0.005 19.93 58-50-726 Ms. Jane Prall 0.040 0.054 77.38 0.006 0.784 25.30 6.888 0.145 2.490 0.005 19.17 58-50-858 Twin Creeks WSC 0.043 0.142 70.86 0 .006 0.930 26.36 7.300 0.152 0.284 16.11 58-50-8M Mooreland Water Co. 0.030 0.081 81 .52 0.004 0.668 23.37 8.390 0.137 6. 790 0.031 17.80 58-58-1 H Hays Hills Baptist Ch. 0.161 0.055 100.30 0.006 3.218 0.534 27.61 6.631 0.175 0.259 0.005 29.38 Barton Springs City of Austin 0.044 0.077 92.64 0.005 0.998 24.51 26.320 0.156 2.394 0.001 24.34 Table 3 (cont.): Chemical analysis fro m water sa mples in th e study area U1 '° V. RESULTS AND DISCUSSION A) Orientations of Lineaments Analyses of lineaments may be readily expressed by statistical methods. The recognition of patterns in a mass of data not visible by individual measurements and the reduction of a large number of features to a small number of significant numbers is a goal of statistics. Hence, a major emphasis of lineaments and fracture traces lies in the population of features. In this way, statistical conclusions are more important than the exact location, length, or orientation of any single lineament. Lineaments and fracture-trace azimuths in this study were separated and displayed as Rose diagrams to illustrate any regional trend in the orientation of linear features (Figure 8). Gay (1976) recommends utilizing a smooth-curve Cartesian histogram to quantitatively display the results of lineament orientations. One of the principle advantages of this type of plot is the ability to select an unique azimuth to characterize each peak in the plot. This is accomplished by dividing the peak into equal areas on each side of a vertical line. Thus characterized, the grouping of linear features can then be quantitatively compared with 60 Fracture Trace Azimuth Distribution North n =760 Percentage of Total Fracture Traces Lineament North Azimuth Distribution n =178 Percentage of Total Lineaments Figure 8: Rose diagrams of lineament and fracture trace orientations in the study area Fracture Trace Cartesian Histogram (n = 760) -------91 °------­ UI Q) CJ 15 ~ I­ ~ :J -CJ ~ 10 u. «i-0 I­ 0 5 goo 270° 315° 00 West North East Azimuth Lineament Cartesian Histogram (n = 178) - UI c Q) E as Q) c :i 10 as -0 I­ - 0 0-----.---........---....----.-----.------...---........----1 270° 45° goo West East Figure 9: Cartesian histograms of lineament and fracture trace orientations in the study area oo 315° North Azimuth other sets. Furthermore, the dispersion of orientations in a set is readily visible as the width of the plotted peak. Additionally, closely spaced peaks, which may each represent distinct sets of linear features, are easily separated and distinguished on histogram plots as well as the low or poorly developed peaks. A smooth-curve Cartesian histogram of the fracture traces and lineaments in the study area is illustrated in Figure 9. Both types of diagrams clearly show the bimodal orientation of the azimuth data. A rectangular graph of relative length versus azimuth is depicted in Figure 10 and reveals the two peaks which exceed the average relative length. The largest peak ranges in azimuth from 20° J: - C>c: Q) average ..J relative Q) length 0.06 > :;::: Qi a: a-. as \/" \r'\,.--• 0.02 a 0 090270 Azimuth (degrees) Figure 10: Rectangular graph showing relation between values of lineament azimuth and relative lineament length in the study area. -70°. A secondary peak is confined to the sector 130° -140°. These results unequivocally concur with the findings of Woodruff, et al. (1989). The significance of each peak is determined by a statistical analysis (Table 4). Tabulated data for each 10° sector includes: total lineament length (L5); average lineament length (La); number of lineaments (n); relative length (Lr); length-weighted frequency (F); and index of preferred orientation (IPO). For the two major peaks, the chi square values ( x2) and the Bernshtein accuracy criterion (H) are listed. The substantial chi square value for the largest peak (72.168) confirms the overall trend in the azimuth data. Figure 11 shows that mean length of lineaments in each 10° sector tends to increase with lineament frequency and sector length. This phenomenon has been previously reported (Haman & Jurgens, 1976; Reeves, 1976) and indicates that lineament peaks are defined both by high frequencies and by lineaments of greater mean length. However, in the Edwards aquifer, the similarity between the curves for frequency and sector length suggests that the size of a peak (sector length) is more a function of the number of lineaments forming it than of the size of these lineaments. From Table 4, the index of preferred orientation is the sum of the IPO values calculated for each sector. The resulting value for IPO (25.8%) is over twice as high as in Dix and Jackson's (1981) report for equivalent numbers of random model lineaments. This suggests that the peakedness of data in this study is not random but results from 140 120 >­ CJ c 100 Q) :l C" Q) 80 "'­ u. -60c Q) E as 40 Q) c ..J 20 0 i:t--w Frequency Mean Length 1300 1200 - -"'­"' Q) Q) 1100 E - .c: -Cl c 1000 Q) ..J c as Q) 900 :E 800 140 120 >­ CJ c 100 Q) :l C" Q) 80 "'­ u. - c 60 Q) E as 40 Q) c ..J 20 0 270 0 090 Lineament Azimuth (degrees) 140000 i:t--w Frequency 120000 -en Sector Length Q) Q) 100000 -"'­ E .c: 80000 -Cl c Q) ..J 60000 "'­ 0 - CJ Q) 40000 (/) 20000 270 0 090 Lineament Azimuth (degrees) Figure 11: Values for mean length (top figure), sector length (bottom) , and frequency of lineaments in the study area directional control of lineaments, which produces IPO values as much as twice as high as those generated randomly. The statistical analysis confirms the trend illustrated in the Rose diagrams and cartesian histograms: the statistically significant majority of lineaments and fracture traces in the study area lay in a northeastern­southwestern direction. This result is expected due to the presence of the associated faults, fractures, and joints of the Balcones fault zone. The secondary peak exists approximately 90° from the primary peak corresponding to the prevailing joint directions associated with the fault zone as described by Dunaway (1962). Myrick et al. (1988), in a study on the Northern Balcones Edwards aquifer, confirmed that lineaments are related to regional structural trends but orientation becomes increasing­ly random away from faulting. Thus, lineaments and fracture traces in the study area reflect the tectonic stresses resulting from the Balcones­Ouachita structural belt and correlate well with the primary fault trend of N 40 E and the corresponding joint trend of N 45 W. B) Correlation between Well Proximity and Locations of Lineaments The distances between each of the 27 wells with specific-capacity data and the nearest fracture trace or lineament to each well was determined using the U. S. Geological Survey's ARC/Info Geographic Information System. Due to the large ranges between values for specific capacity and distance, the data was plotted on a log-log scale. A majority RANGE IN AZIMUTH (In degrees) ';j O" ii)' ""' (/) c 3 3 Ill -< 51 C/l Cll co n. Cll a. S!?. ~­ S!?. o· ~ Ill ::J Ill -< !!!. C/l 51 ::J Cll Ill 3 Cll note: all lengths are in meters .a ~ ~~i' c g. C/l 0-1 0 10­20 20-30 30-40 40-50 50-60 60-70 70-80 80-90 Sector Length (Ls) 26407 32999 67964 122328 139046 114938 77802 41338 40189 No. of Lineaments (n) 27 39 54 99 11 8 1 02 64 38 44 Mean Length (La) 978.0 846.1 1258.6 1235 .6 1178.4 1126.8 1215.7 1087.9 913.4 Relative Length (Lr) 0.026 0.032 0.066 0.119 0.135 0.112 0.076 0.04 0 0 .039 Length-Weighted Frequency (F) 0.694 1.252 3 .570 11.780 15.960 11 .404 4 .844 1.528 1. 720 Index of Preferred Orientation (IPO) 0.030 0 .023 0.011 0.063 0 .080 0.056 0.020 0.015 0.016 Chi Square (sector) 0.011 17.009 39.395 15.449 0.305 Chi Square X2= 72.168 Bernshtein criterion H: 18.042 °''l ~ O" Sector Length (Ls) (i) ~ No. of Lineaments (n) () 0 S.. Mean Length (La) c ::J Cl> s Relative Length (Lr) (f) c 3 3 Length-Weighted Q.l -< Frequency (F) g, !!?. Index of Preferred a ~ Orientation (IPO) 0 ~ Q.l ::J Chi Square (sector) Q.l -< Chi Square (/) iii" Bernshtein criterion g, ::J Cl> Q.l 3 Cl) 3. Q.l N 3· 5. ::r (/) 90-100 40209 41 980.7 0.039 1.604 0 .016 100-110 22201 26 853.9 0.022 0.561 0.034 RANGE IN AZIMUTH (in degrees) 110-120 51225 47 1089.9 0 .050 2 .342 0.006 note: 120-130 42366 37 1145.0 0.041 1.525 0.014 X2 = H: 130-140 69497 61 1139.3 0 .068 0 .012 0 .033 0 .000 4.124 0.033 140-150 43281 39 1109.8 0.042 1.642 0.013 all lengths are in meters 150-160 22842 23 993 .1 0.022 0.511 0.033 160-170 36476 36 1013.2 0.035 1.277 0 .020 170-180 36927 43 858.8 0.036 1.545 0.020 (XJ °' of the wells with large specific-capacity values are located to the southeast of a northeast-southwest trending lineament. These wells were separately plotted and statistically analyzed to confirm if the differences observed were due to chance or some type of controlling h ydrogeologic factor. Plots showing the relation between values of specific capacity and distance based on the classification of the lineament ("1-man", "2-man" or "3-man") are presented in Figure 12. The degree of correlation between these values decreases as the number of interpreters who identified the lineament increases. It seemed intuitive at the beginning of the study to assume that a greater degree of correlation would occur with lineaments that had been identified by all three interpreters because these lineaments are more obvious and, thus, more likely to be true structural features. However, only 48 of the 938 total lineaments in the study area were identified by all three interpreters. Combined with the relatively small number of specific-capacity values, the 48 "3-man" lineaments are too sparsely distributed to show any correlation with the specific-capacity data. A good correlation exists with increased specific-capacity values and decreased distances to the nearest lineament, regardless of its classification. As shown in Figure 12, this correlation is especially evident for wells located within 200 feet (61 meters) of a lineament. Any mathematical correlation is complicated by the extremely large specific­capacity value of the well located at Mountain City Oaks Water Supply I ~ j I ' ' 200 ft l - = l!I l I l --E j I c. 100 l ' Cl i Ill l!I '! 11!1 -G) 1:1 :::s l l!I ad a I a; l l!l ~ l l > 10 l:J ­ l > l l irm '(j­ ' I l!I l' l as 1!J l El I c. ' a a ' ·a as ! I I (.) I : () ! i l :;:: l l l l '(j l!I tjEI G) j I j c. I l en 0.1 ' 1 1 0 100 1000 10000 Distance to Nearest "1, 2 or 3-man" Lineament (ft) 1000 l ' Ell ' l -' jl l = ' El l --E I l c. 100 l ' Cl l a l - a G) a l :::s l!I I El a; a_ > 10 ; l:l ' > :t:: l i ' l!I (J as l l!I c. l El a l a as ' l (.) I ' I : I () :;:: () I ~ 1£1 G) l El l!I a 1:1 l!I a a aa l!I i I ' ' I c. ; en 0.1 ' I I I 1 10 100 1000 10000 Distance to Nearest "2 or 3-man" Lineament (ft) Figure 12: Well productivity and the lineament type in the study area 1000 - = --E c. 100 C> -Q) ::s as > 10 >­ '() - as c. as 0 I 0 ;;:::: '() Q) c. en 0.1 a I a 1 I 1 I ! i ! ! I a ! ' a a ! aa Iifla ' ! a_ ! -~ 1 ! a ! a a aa I a a a a1:1 1:1 ! ! ' ~ ! ! i' ! i ! m. a a ! ' ~ ! I 10 100 1000 10000 Distance to Nearest "2-man" Lineament (ft) -1000...,.....-----------.----------...,...----------~,---a-. -------. I I e = ' a 100+-~~~~--l~~~~~-t-~~~~~~~~~~--i ' c. C> I 1:1 a -Q) ! !:la ::s ! Ill Ull a; > 10 -+-~~~~-+~~~~~-t-~~~~-1-~'P--~-m-·...._~~ >­ -, al:l: '()­ as ' artl c. ! ~a as i 0 1 ~,~~~~~+-~~~~~~~~~~ .........~~~~--t I 0 ! ;;:::: '(3 ci a Q) a c. ~ en 0.1-+---.--..........-T"....,..__-.--.-.,....,..,..,..,-ri---..---,.........,..........,.,;t--....,.......,... ........ 'TT'lm 10 100 1000 10000 Distance to Nearest "3-man" Lineament (ft) Figure 12 (cont.): Well productivity and the lineament type in the study area Company. Here, the well was pumped at 184 gallons per minute with a drawdown of only 3 inches (7.6 cm), but is located 207 feet (63 meters) from the nearest lineament. An examination of the drilling log for that well, however, shows that two large "cavernous openings" were encountered during drilling. Despite the lack of correlation between specific capacity and "3­man" lineaments, it is still recommended that lineament analyses are conducted by more than one interpreter. Typically, lineaments are interpreted from aerial photographs during a limited viewing time. To a point, multiple interpreters are able to identify more lineaments in the same time restrictions. The importance of lineament analysis is in the total distribution of lineaments and fracture traces in any one area. C) Correlation between Well Productivity and Directions to Lineaments More important than distance to the nearest lineament is the direction of the well from the lineament and the orientation of that lineament. Table 5 lists the 27 wells in order of increasing specific­capacity values. As shown in the second column from the right, 10 of the 13 most productive wells are located to the southeast of a linear feature. None of the 13 least productive wells were located within 1000 feet (305 meters) to the southeast of a lineament. These "southeast" wells were separated from the rest of the data set and graphed separately. As shown in Figure 13, wells located southeast of a lineament show a Spec. Cap. Value Direction Order of Well Number Well Name (gpm/ft) (from lineament) Rank 58-58-202 Mystic Oaks WSC #1 0.23 SW 1 58-50-830 Slaughter Creek Acres 0.28 s 2 58-50-8A Native Texas Nursery 0.30 SW 3 58-42-821 Trigg-Forister Bldg. 1.54 NW 4 58-42-8M Allen Keller Co. 1.67 NE 5 58-58-2E Hunter Industries 1.71 NE 6 58-58-lEE Neptune-Wilkinson Co. 1.77 N 7 58-49-9H Charles Ranch 2.08 SW 8 58-58-508 Goforth WSC #4 2.50 NE 9 58-50-223 City of Sunset Valley 2.52 SW 10 58-50-414 Lee V. Johnson 2.83 NW 11 58-57-307 Dahlstrom Middle Sch. 3.71 NW 12 58-58-413 City of Buda #3 4.33 NE 13 58-58-506 Goforth WSC #2 4.77 NE 14 58-58-412 Plum Creek WSC #2 8.94 SE 15 58-58-lH Hays Hills Baptist Ch. 10.14 NW 16 58-42-812 W. F. Guyton & Assoc. 13.33 SE 17 58-50-731 Shady Hollow Estates 21.00 SE 18 58-50-835 Onion Creek CC 22.50 NW 19 58-58-406 Texas-Lehigh Cement Co. 22.64 E 20 58-42-8S Espey Huston & Assoc. 25.00 SE 21 58-58-123 Elizabeth Porter 26.67 SE 22 58-58-lA Frank Burdette 37.07 SE 23 58-50-704 Marbridge Found. #5 37.10 SE 24 58-58-115 Estate Utilities WSC 55.00 SE 25 58-58-102 Cimarron Park #2 150.00 SE 26 58-57-910 Mt. City Oaks WSC 736.00 SE 27 Table 5: Specific-capacity values ranked in ascending order greater correlation with specific capacity than wells that do not lie to the southeast of a lineament (Figure 14). Because of the predominance of southwest-northeast trending structural lineaments in the study area, there is an inherent difference Wells Located Southeast of a Lineament 10 100 1000 10000 Distance from Nearest Lineament (feet) Figure 13: Values of specific capacities and distances from wells to nearest lineament northwest of each SE well between these types of lineaments as opposed to linear features that do not trend in a SW /NE direction. In order to test this correlation, all non-SW/NE trending lineaments were removed from the data set. Distances from the 27 wells to the nearest SW/NE trending lineament were determined and plotted in Figure 15. In addition, the distance to the nearest SW/NE trending lineament that lies to the northwest of the 27 wells was measured and examined in Figure 16. Restricting the type of lineaments to only those that trend in a SW/NE direction results in plots that appear to have a better predictive Wells Not Located Southeast of a Lineament 10 100 1000 10000 Distance from Nearest Lineament (feet) Figure 14: Values of specific capacities and distances from wells to nearest lineament nQ1 located northwest of each non-SE well pattern than the trends illustrated in Figures 13 and 14 when all lineaments were utilized. This is probably due to the predominance of faults and fractures that lie in a southwest/northeast orientation. Because the majority of structural groundwater conduits also run in this direction, a tolerable correlation appears to exist between increased specific-capacity values and decreased distance to the well from a lineament in a southeasterly direction. Relationship with SW/NE Lineaments Only - = --E c. 100 C> - Q) :::::s as > 10 >­ .:!::::! CJ as c. as (.) I CJ :;::: '() Q) c. (/) 0.1 10 100 1000 10000 Distance to Nearest Lineament to the NW or SE (ft) Figure 15: Values of specific capacities and distances to the nearest southwest­northeast trending lineament located to the northwest or southeast of a well The Mann-Whitney U Test was applied to the specific-capacity data to statistically determine if the difference between specific-capacity values of wells located to the southeast of a SW /NE trending lineament are random data or structurally controlled. The null hypothesis to be tested is: H0 -direction of a well from a lineament has no effect on specific capacity. The alternative hypothesis is H 1: direction of a well from a lineament has a significant effect on specific capacity. As shown in Table 6, the statistic U is calculated for both southeast wells and non-southeast wells. Since the lower computed U value, 7, is less than the Relationship with SW/NE Lineaments Only - -.. = E a. 100 C') -Cl) ::J a; > 10 >­ '(3­ as a. as (.) I () :;: '(3 Cl) a. en 0.1 1 10 100 1000 10000 Distance to Nearest Lineament to the NW (ft) Figure 16: Values of specific capacities and distances to the nearest southwest-northeast trending lineament located to the northwest of the well expected value of U, 45 (at the 95%-confidence level), the null hypothe­sis is rejected. As a result, there is a significant difference in well productivity with respect to wells located southeast of SW /NE trending lineaments. The dominance of large specific-capacity values located southeast of a SW /NE trending lineament is further reinforced upon examining the exceptions in the specific-capacity data. As Table 5 shows, three of the 13 wells with the largest specific-capacity values are not located southeast of a lineament. However, each of these three wells, Texas­ Ranks for Two Groups SE Wells 15 17 18 21 22 23 24 25 26 27 n1 =10 Non-SE Wells 1 2 3 4 5 6 7 8 9 10 11 12 13 14 16 19 20 U2 =163 Table 6: Summary of Mann-Whitney U test results Lehigh Cement Co., Onion Creek Country Club, and Hays Hills Baptist Church, have extenuating circumstances which are evident on the base map. The Texas-Lehigh Cement Co. well, while technically just east of a north-south trending lineament, is also equidistant away from the endpoint of a northeast-southwest trending lineament. Likewise, the well at the Onion Creek Country Club is northwest of the nearest lineament but is also only 300 feet (91 meters) to the southeast of a northeast-southwest trending lineament. The Hays Hills Baptist Church location is also located between a series of linear features all of which are oriented in a northeast-southwest direction. Although the well is located northwest of the nearest lineament, it is also located to the southeast of another lineament that, in fact, may have structural control on the well productivity. Of the 27 wells with specific-capacity values, the 13 largest values are all located to the southeast (within 700 feet or 213 meters) of a northeast-southwest trending lineament or fracture trace. Specific­capacity values for these wells range from 8.94 gpm/ft to 736.0 gpm/ft. Conversely, none of the 14 remaining wells are located within 1000 feet (305 meters) of a southwest-northeast trending lineament or fracture trace. These specific-capacity values range from 0.23 to 4.77 gpm/ft. Clearly, direction of a well from a lineament is more significant than distance to the nearest lineament. This is to be expected in the Barton-Springs section of the Edwards aquifer because of the attitude of the structural features in the formation. Most faults and fractures are near-vertical and any dip direction is to the southeast. Consequently, the large-yielding wells located to the southeast of northeast-southwest trending lineaments or fracture traces are most likely intersecting highly permeable fracture zones of steeply dipping faults and fractures. Based on Figure 16, a limited range for the specific-capacity values of potential water wells in the study area can be estimated by measuring the distance from the well to the nearest SW/NE trending lineament located to the northwest of the well. For example, a well located 1000 feet (305 meters) to the southeast of a lineament would be expected to have a specific­capacity value of approximately 1 -10 gpm/ft of drawdown. In this manner, lineament analysis can be used as a tool for predicting reasonable ranges of specific-capacity values for prospective wells. D) Correlation between Water Chemistry and Locations of Lineaments Table 3 lists the results of the chemical analysis of 61 water samples collected in the Barton-Springs section of the Edwards aquifer. Chemical parameters were chosen to provide an overall chemical survey of the study area. Specific parameters and ratios were used to evaluate possible chemical relationships between locations of lineaments and large yield water wells. A Piper diagram in Figure 17 displays results similar to those found in other water chemistry surveys of the area (Senger & Kreitler, 1984; Clements, 1989): calcium-bicarbon­ate waters of the Edwards grading to sodium-sulfate water in the vicinity of the ''bad-water" line. Well #58-50-854, located in the ''bad­water" zone, is isolated in the sodium-chloride section of the Piper plot. Bicarbonate values are relatively constant with a mean value of 294 mg/L. At Barton Springs, the discharge point for this section of the aquifer, the bicarbonate value, 216 mg/L, was the smallest value 80 60 40 20 20 40 60 80 ~c.a Cl --+ Cations % meq/I Anions Figure 17: Piper diagram of 61 samples collected from the study area 1000 -...J ---C> E Q) - as c: 0 .c ..... as (,) cc 100 M Ill lblll t: Ill Ill Ill l·I 1·P@ 3~1:1 liiJ t: . rF 3 Ill ·~~ ~~ Ill If:! I·l 6.4 6.6 6.8 7.0 7.2 7.4 7.6 pH Figure 18: Plot of carbonate chemistry for samples collected in the study area measured. According to Abbott (1977b), groundwater in the artesian zone of the Edwards aquifer is at least seasonably under-saturated. This undersaturation is partly due to large volumes of groundwater flowing in pipe-like voids where little of the water is actually in contact with the host rock, and partly due to the mixing effect that occurs with the addition of large volumes of undersaturated recharge. Consequently, caverns are probably being enlarged during seasonal conditions with zones of largest permeability receiving the greatest amounts of dissolution. Figure 18 illustrates the relationship between bicarbonate and pH of the collected water samples. The concentrations of three anions in the collected samples were determined: chloride, nitrate, and sulfate. No flouride, bromide, nitrite, or phosphate was identified in the samples because the concentrations of these parameters were below the method detection level. Chloride concentrations ranged from 3.4 mg/L to 545.2 mg/L. Excluding the well located at St. Alban's Church (#58-50-854), the mean chloride concentration is 15 mg/L throughout the majority of the study area. The St. Alban's Church well has an extremely large chloride concentration which is to be expected with the brine-like water in the "bad-water" zone. Relatively large chloride values appeared along the western, northern, and eastern boundaries. Nitrate concentrations averaged 4.7 mg/L with a maximum value of 19.5 mg/L. This value, however, occurs on the Don West Ranch (#58-57-204) and is likely due to the cattle ranching in the area. Sulfate and strontium concentrations from the 13 wells with the largest specific-capacity values were also examined. The mean values for sulfate and strontium in the Edwards aquifer from this study are 60.1 mg/Land 2.2 mg/L, respectively. If the fractures near these wells penetrated the underlying Glen Rose Limestone, larger concentrations of sulfate and strontium may have occurred. Sulfate values exist for 9 of the 13 large-flow wells and the mean concentration value is 29.8 mg/L. Similarly, the mean strontium values for the large flow wells are about one-half of the average of all the wells in the study area. In addition, values for both sulfate and strontium vary widely in the large-flow --..J C" Cl) E a - (.) 1:1 -g a "l:f' 0 di! CJ) a a a 1:1 a .1 .01 . 1 1 10 Sulfate (meq/L) 1000 - ~ 100 C" Cl) E - "­ CJ) - "l:f' 100 CJ) 1 .01 .1 1 10 Sulfate (meq/L) Figure 19: Sulfate ratios versus chloride (top) and strontium (bottom) for samples collected in the study area wells. Sulfate ratios as compared to chloride and strontium are plotted in Figure 19. Consequently, no direct correlation can be made with sulfate and strontium concentrations and lineaments. This may be due to a paucity of data, a complex distribution of constituents or a lack of upward West East Figure 20: Schematic cross section across the Balcones fault zone (from Senger & Kreitler, 1984) groundwater movement through vertical fractures from the Glen Rose. The last point is confirmed by Senger & Kreitler (1984): "Leakage from the Glen Rose Limestone is probably not upward through the Walnut Formation into the Edwards Limestone but instead is lateral across fault surfaces" (Figure 20). Similarly, Slade et al. (1986) presents an alternative conclusion: .... vertical displacements along faults which exceed the thickness of the Walnut Formation would cause the upper Trinity and Edwards aquifers to be in direct contact along these faults. Water movement could then occur directly between these two aquifers. In both cases, lineaments, which indicate vertical faults and fractures, would not be locations of large sulfate and strontium concentrations. The four major cations, calcium, potassium, magnesium, and sodium all reflect substantial variability among the 61 water samples. Neither calcium or magnesium concentrations display a predictive distributive pattern except for the tremendously large values for both constituents at the St. Alban's Church well in the ''badwater" zone. The mean calcium and magnesium concentrations are 78.5 mg/L and 30.62 mg/L respectively. Figure 21 presents a plot of calcium versus magnesium. Sodium concentrations appear to plainly delineate the "bad­water" line. While the mean sodium value is 13.4 mg/L, all of the sampled wells that lie within 5000 feet (1524 meters) of the ''bad-water" line (and Interstate 35) have sodium values greater than 25 mg/L including a concentration of 402 mg/Lat the St. Alban's Church site. A sodium-chloride plot is illustrated in Figure 22. The mean potassium concentration is slightly larger than 2 mg/L in the study area. No significant associations are evident between locations of lineaments and trace metal concentrations. Except for an anomalous value at Hays Hills Baptist Church (#58-58-lH), the mean concentration of aluminum is 0.06 mg/L. Both the aluminum concentration (0.16 mg/L) and the iron value (3.2 mg/L) are large at this well probably due 4 -..J c-3 Q) E - E ::::s u 2 ca (..) Ill Ill a Ill 13 13 I·~ Ill Ill 13 .,. Ill IG a 13 "" "' :&'"~~Ill ! i§'i . .. t;l Ill ~ .gJ . 1!11!1 l!I Iii 0 1 2 3 Magnesium (meq/L) Figure 21: Calcium-magnesium relationship for samples collected in the study area to its recent completion. Chromium, lead, and zinc concentrations are all relatively small with no anomalously large values measured among the 61 samples. In this study, no correlation can be made between locations of lineaments and water chemistry. Perhaps a larger database could possibly reveal some type of relation. However, chemical analysis is a site-specific application. Not all linear features are necessarily predictive, thus, site-specific techniques such as water chemistry and dye tracing may not be the most appropriate tools for lineament analysis. In contrast, the combination of remote sensing of lineaments with their -..J c-2 Q) E - E ::s "O 0 en 0 L:I . Ill i a Ill ;Ill Ill a Jdflll ~ a . L:I Ill Ill Ill a Ill 0.0 0.5 1.0 1.5 Chloride (meq/L) Figure 22: Sodium-chloride relationship for samples collected in the study area statistical analyses can produce useful clues to the probable location of areas of increased permeability. VI. CONCLUSIONS AND IMPLICATIONS This study investigates the correlations between structural lineaments and water-well yields in the Barton-Springs segment of the Edwards aquifer. In the Austin, Texas area, lineaments represent the structural grain of the Balcones-Ouachita fault zone and may indicate subsurface geologic phenomena such as faults, fractures, and joints. These structural features often represent discrete zones of large permeability, and thus, areas of enhanced flow of groundwater capable of conveying greater quantities of water than surrounding, non­fractured rock. Specific conclusions are: 1. Lineaments and fracture traces in the study area represent the tectonic stresses resulting from the Balcones-Ouachita structural belt and correlate with the primary fault trend of N 40 E and the corresponding joint trend of N 45 W. 2. The total length of lineaments in each 10° azimuth sector is a function of the number of lineaments within the sector rather than the length of the individual linear features. 3. Although any given lineament is not necessarily predictive of grain, the overall trend defined by all lineaments provides a clue to the structural grain in the study area. Likewise, the total number of lineaments is more important than the number of interpreters who identify a particular lineament. 4. A good correlation exists with increased specific-capacity values and decreased distances from each well to its nearest lineament, regardless of its classification. This correlation is especially evident for wells located within 200 feet of a lineament. 5. SW/NE trending lineaments have a greater influence on well yields than lineaments that are not oriented in this direction. 6. Wells located southeast of SW /NE trending lineaments indicate a greater correlation with specific-capacity values than do other wells. 7. Limited ranges for specific-capacity values of potential water wells in the study area can be reasonably estimated by measuring the distance from the well to the nearest SW/NE trending lineament located to the northwest of the well. 8. No direct correlation between locations of wells with respect to lineaments and water chemistry can be made for this particular study. 89 It is important to recognize that using the described lineament analysis to locate water wells will not necessarily guarantee success. Any particular well may fail to intersect a sufficient number of subsurface fractures to provide the well yield required to satisfy its desired use. Also, a particular fracture or set of fractures that may be intersected in a well may not have a sufficient storage and transmission capability to produce a large well yield. However, the use of the described lineament analysis to locate water wells can maximize the probability of obtaining a large-yield well. Fracture-trace and lineament analysis can be particularly useful in determining the locations of groundwater monitoring wells. Because groundwater flow preferentially follows the most permeable pathway, monitoring wells should be based on fracture traces or lineaments. For example, if a hazardous-waste storage lagoon is located in an area of fractured bedrock, at least one of the downgradient monitoring wells, as required under the Resource Conservation and Recovery Act, should be located on a fracture trace or lineament. Lineaments provide the hydrogeologist with a tool for predicting possible sites of environmental sensitivity particularly with respect to groundwater resources. Examples include the siting of groundwater monitoring wells for point sources of pollution, predicting the likely underground flow paths of a pollution plume or potential recharge enhancement dams. Thus, the location, orientation, and density of structural lineaments, along with the described statistical analyses of lineaments, will provide the water resource manager with the ability to identify discrete groundwater flow paths, predict contaminant-plume migration and, subsequently, to apply appropriate mitigation procedures. Future research should focus on improving the understanding of the hydraulics of the Edwards aquifer. To this end, validated specific­capacity data should be acquired in the field from various well locations throughout the aquifer. Long-term pump tests should be conducted to complement specific-capacity data. Down-hole geophysical techniques could be employed to identify subsurface cavities. If the results conform to this study, a water well should be drilled to the southeast of a SW/NE trending lineament to substantiate the results presented in this paper. In addition, further study is needed to determine the change in groundwater flow paths with decreasing water levels in the Edwards aquifer. VII. APPENDICES A. Previous Pump Tests in the Edwards Aquifer Due to the paucity of aquifer test data in the Barton-Springs section of the Edwards aquifer, the results of four pump tests, conducted by private hydrogeological consulting firms, are described below to illustrate the extreme ranges of transmissivity, permeability and specific-capacity values in the study area. The karstic features of the Edwards result in turbulent flow through crevices, dissolution cavities, fractures, and channels throughout conditions of varying hydraulic gradients, air entrapment, and hydrostatic pressures. This accounts for the widespread variation in evaluation of the aquifer at a specific location. Locations of the following wells can be found on the Plate 1. Well #58-42-821 In January 1982, Underground Resource Management, Inc. of Austin, Texas was retained to review the records related to a water well located at 2502 Loop 360 South, Austin, Texas. The 460-foot (140 meter) deep well was drilled and completed in April 1981 by Central Texas 93 Drilling Company. As indicated in the driller's report, 6-5/8" diameter steel casing was set from ground surface to 350 feet (107 meter) below the ground surface. After completion, a 2 h.p. Red Jacket submersible pump was installed in the well. On February 2, 1982, a pumping test was performed to determine the specific capacity of the water-supply well. The static water level was 262.2 feet (80 meter) below the top of the casing. The pumping rate was measured at 16 gpm with a maximum drawdown of 10.4 feet (3.2 meter) after 90 minutes of pumping. The resulting specific capacity of the well was calculated to be 1.54 gpm/ft of drawdown. Well #58-50-731 In 1983, Underground Resource Management, Inc. of Austin, Texas supervised the installation of a water well for the Shady Hollow Estates Subdivision north of Manchaca, Texas. A 6.5 inch (16.5 cm) diameter test hole was drilled by Central Texas Drilling Company to a depth of 420 feet (128 meters). As noted in the driller's log, large fractures were first encountered at a depth of 231 feet (70 meters) (509' msl). From 231 feet to 330 feet (101 meters) (410' msl) the action of the drill stem and the nature of the returns from the hole suggested that solution enlargement of the secondary fractures associated with the fault had been extensive. The quantity of water blown to the surface with the returns increased substantially. After the test hole was drilled to depth, it was reamed to a diameter of 9-7 /8" to a depth of 438 feet (134 meters) (302' msl). At this depth, the density of encountered fractures had lessoned considerably and it was felt that the hole was nearing the bottom of the Edwards Formation. The hole was cased with 6-5/8" (16.8 cm) plain-end welded steel casing to a depth of 433 feet (132 meters) (307' msl). In order to evaluate the potential of the well, a 24-hour pump test was conducted using a 10 h.p. Red Jacket submersible pump. The pump was set at 315 feet (96 meters) (425' msl) on 3" (7.6 cm) drop pipe with an airline strapped to the pipe. One inch (2.54 cm) diameter tubing was run alongside the drop pipe to a depth of 315 feet (96 meters) (425' msl). A pressure gauge attached to the airline and an electric probe run through the tubing were used to determine the change in water level throughout the test. At a pumping rate of 210 gpm, the water level inside the well was drawn down approximately 10 feet (3 meters) at the end of the 24-hour period. The resulting specific capacity of the well was calculated to be 21 gpm/ft of drawdown. Well #58-58-2E A pump test was conducted on Well #58-58-2E near Buda, Texas in November 1989 by Jack H. Holt & Associates, Inc. The property is located in Hays County approximately 2 miles (3.2 km) northeast of Buda at the southeast corner of Turnersville Road and the Interstate 35 frontage road. This site is owned by Hunter Industries and is used as a construction staging yard and a temporary concrete batch plant site. A 700-foot deep (213 meters) well was drilled by Kucher Drilling of San Marcos, TX in October 1989. The 8 inch (20.3 cm) diameter well was cased to a depth of 460 feet (140 meters) and grouted with a cement slurry. A 20 h.p. submersible pump with a 3 inch (7.6 cm) discharge pipe was placed at a depth of 300 feet (91 meters) from the ground surface. The purpose of the pump test was to determine flow rates, well drawdown, and possible effects of well drawdown on the Phillips Well located approximately 500 feet (152 meters) to the northwest. The pump discharge was a constant 200 gpm as verified by meter readings at the discharge pipe. The test was run for a period of 7 hours with a maximum drawdown of 117 feet (36 meters). The resulting specific capacity of the well was calculated to be 1.7 gpm/ft of drawdown. Water level in the well completely recovered within two hours. The discharge from the well did not effect the Phillips Well. Calculated transmissivity values were invalid due to erratic drawdown values caused by karstic groundwater flow. Well #58-57-8A In June 1990, Jack H. Holt & Associates, Inc. conducted a drawdown test at Native Texas Nursery in south Austin. The recently drilled well is located approximately 1.4 miles (2.2 km) south of Slaughter Lane and approximately 0.7 miles (1.1 km) east of Manchaca Road. The water will be used for irrigation for a plant nursery of approximately 3.5 acres. The Native Texas Nursery Well was drilled by Associated Drilling Company of Manchaca, Texas and completed on 26 June 1990 to a depth of 500 feet (152 meters). The well is cased with 5 inch (12.7 cm) diameter PVC to a depth of 500 feet (152 meters)and screened from 360 to 480 feet (110 to 146 meters). A 5 h.p. pump was set to a depth of 400 feet (122 meters) with a 1.25 inch (3.2 cm) PVC discharge. The discharge pipe is connected to a 2000 gallon steel storage tank approximately 15 feet (4.6 meters) from the well head. A pumping test was conducted on 29 June 1990 to determine the productivity of the well. The static water level was 171.6 feet (52.3 meters) below the top of the casing. As verified by meter readings (in gallons), the pump discharge was a constant 36 gpm throughout the 6 hour test. With a maximum drawdown of 118.96 feet (36.3 meters), the specific capacity was calculated as 0.30 gpm/ft of drawdown. The aquifer transmissivity was determined to be 950.4 gal/dayI ft. Pool and Rogers Co. Mystic Oak• wsc #1 Cimarron Park #1 Trigg Building Suburban Austin WSC Mr. Herb Mendieta Park Hill Baptist Ch. St. Alban'• Epia. Ch. Comal Tackle, Inc. ~. :;;d Parameter Units #58 -58·219 #58-58·202 #58-58·114 #58-42·821 #58 ·50· 733 #58 -50 -520 #58-42-913 #58-50-854 #58-58-416 t"t) rJl Alkalinity, tolal mg/ L 228 262 258 224 268 269 299 228 270 ~- Alkalinity, bicarb. Alpha, gross mg/L pCi/L 228 8.4 262 18.8 258 3.2 224 1.1 268 1.4 269 2.9 299 4.9 228 270 rJl Aluminum, dissolved mg/ L 0.04 0.02 0.01 <0.01 <0.01 <0.01 <0.01 0.02 <0.01 0 Arsenic, dissolved mg/ L <0.01 <0.01 <0.005 <0.005 <0.005 <0.005 <0.005 <0.005 <0.005 Ho. Barium, dissolved mg/ L Boron, dissolved mg/L Cadmium, dissolved mg/ L Calcium, dissolved mg/ L Carbon, tolal organic mg/L 0.03 0.37 <0.0t 52.26 2 0.04 1.2 0.03 63.28 2 0.04 0.32 0.01 65.13 3 0.04 0.27 <0.01 67.93 2 0.04 0.21 <0.01 72.2 3 0.14 0.18 O.Ot 70.82 2 0.07 0.15 <0.01 95.95 3 0.05 1.38 <0.01 130.98 1.3 0.06 0.14 <0.01 82.65 2 ('):::r t"t) s Chloride Chromium, dissolved Copper, dissolved mg/ L mg/ L mg IL 44 <0.0t <0.01 51 <0.01 0.02 12 <0.01 0.02 24 <0.01 <0.0 1 13 <0.0 1 0.01 12 <0.0t <0.01 21 <0.01 <0.0t 273 <0.01 <0.0t 14 <0.01 <0.01 ~· n SlJ- Flouride Iron , dissolved mg /L mg/L 3.6 0.32 4 0.01 0.4 0.01 0.2 <0.01 0. 2 <0.01 0.3 <0.01 0.2 <0.01 3.9 0.08 0.2 <0.01 > Lead, dissolved mg IL Magnesium, dissolved mg/L <0.01 35 .58 <0.01 46.26 <0.01 27.58 <0.01 21.81 <0.0t 23.4 <0.0 1 25.65 <0.01 20.09 <0.005 99.84 <0.005 21 :::::s SlJ- Manganese, dissolved mg /L <0.01 0.02 <0.01 <0.0t <0.01 <0.01 <0.01 <0.01 <0.01 '- Total Hardness Zinc, dissolved Silica mg/ L mg IL mgI L 277 <0.01 1t.46 349 0.03 t2.73 276 0.01 10.2 259 0.12 8.89 277 0.01 10.14 282 <0.01 10 322 <0.01 10 738 <0.01 14.36 293 <0.01 11 . 19 ~ SlJ O"' 0 ~ SlJ- 0 ~ '° 00 City ol Buda Dahlstrom Village ol Cily ol Chaparral Creedmoor- Mr. J . D. llays lligh Shady Golorth O:l. Well #1 Middle Sch. San Leanna Sunset Vall. Park #2 Maha #2 Malone School Hollow Est. WSC #4 ~ Parameter Un it s #58· 58 -403 #58-57-307 #58-50-855 #58-50-223 #58-49-9 t 1 #58-50 -847 #58-50-852 #58-57-901 #58-50-731 #58-58-508 ro Alkalinity, Alkalinity, total bicarb. mgll mg/ L 276 276 262 262 224 224 288 288 292 292 240 240 222 222 250 250 277 277 228 228 'Jls:: 1--',..,. Alpha. gross pCill 4.5 2.3 7.1 1.2 7.3 7.3 6.9 2.9 1.8 5.1 'Jl Aluminum, dissolved mg/ L 0.02 0.01 <0.01 <0.01 <0.02 <0.02 <0.01 <0.01 <0.01 0.02 0 Arsenic, dissolved mgll <0.005 <0.005 <0.0 1 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 to+> Barium, dissolved mg/ l Boron, dissolved mgl l Cadmium, dissolved mg/l Calcium, dissolved mg/ L Carbon, total organic mg/ l 1.1 <0.01 73.05 0.9 0.47 0.04 69.49 0.8 0.07 0.14 <0.01 63.22 1.2 0.32 0.08 <0.01 73.08 0.8 0.12 0.11 <0.01 93.89 1.7 0.12 0.06 <0.01 67.77 1.7 0.05 0.14 <0.0 1 56.93 2 0.03 <0.01 <0.01 58.71 1.9 0.03 <0.01 <0.01 80.22 2 0.07 0.02 <0.01 62.75 2 (j:::r ro s Chloride Chromium, dissolved mg IL mg l l 10 <0.01 12 0.02 14 <0.01 14 <0.01 15 <0.01 11 <0.01 20 <0.01 10 <0.01 1 3 <0.01 12 <0.01 I-'• ("'} ~ Copper. dissolved mg/ L 0.05 0.03 <0.01 <0.01 0.01 0.01 <0.0 1 <0.01 <0.01 <0.01 1--' Flouride Iron, dissolved mg IL mgll 0.4 0.03 0.2 0.03 2.1 <0.01 0.3 <0.01 0.7 <0.01 0.8 <0.01 2.5 <0.01 0.4 <0.01 0.2 <0.01 3.2 0.29 ~ Lead, dissolved mg l l <0.005 <0.005 <0.01 <0.01 <0.01 <0.01 <0.01 <0.0 1 <0.01 <0.0 1 ~ Magnesium, dissolved mg l l Manganese, dissolvod mgll Mercury, dissolved mg l l 26.34 <0.01 <0.001 24.68 0.02 <0.001 30.58 <0.01 <0.00 1 31.75 <0.01 <0.001 51.38 <0.01 <0.01 26. 12 <0.01 <0.01 33.11 <0.01 <0.01 28 .32 <0.01 <0.01 23.67 <0.0 1 <0.01 35.75 <0.01 <0.01 ~ 1--' '