Data Exploration: Characterization of the

Groundwater resources of the Leona Aquifer

Term Project

GIS in Water Resources

Fall 2002

Lloyd Hemphill

Department of Geological Sciences

The University of Texas at Austin

Motivation: Why should shallow alluvial aquifers be studied?

Shallow alluvial aquifers are an important but very fragile source of groundwater. Flat alluvial plains are the sites of both agriculture and urban development which makes these aquifers a convenient source of groundwater. Historically alluvium has been a popular source of ground water because hand-dug wells did not need to be very deep to reach water. Shallow alluvial aquifers are also very abundant worldwide. One problem with these aquifers is the fact that they are easily contaminated by numerous means. As a result, many have been abandoned for deeper and better protected aquifers. Even so, shallow aquifers should not be ignored because they are often the recharge areas for deeper aquifers. The Leona Aquifer in Central Texas is a good example of an important shallow alluvial aquifer.

Background

This project deals with the Leona Formation as it occurs in Caldwell and Hays County in central Texas. The only major city located on this formation is the City of Lockhart. The Leona stretches from Interstate Highway 35, near Kyle, TX to the southeast of Lockhart. This is about 38 kilometers or 23 miles.

The Leona formation consists of stratified gravels, sands, and clays. These are Quaternary sediments eroded from the Edwards Plateau. (Follett, 1966) Underlying the Leona Formation is the Pecan Gap Chalk, Navarro and Marlbrook Marl Group, Midway Group, and the Wilcox Group. For this project, the most important characteristic of these formations is their degree of permeability. The Wilcox Group is relatively more permeable than the other units.

Figure 1: Geologic Map of the Leona Formation as it occurs in Hays and Caldwell County, TX.

Qle: Leona Formation; Kpg: Pecan Gap Chalk; Kknm: Navarro and Marlbrook Marl;

Emi: Midway Group; Ewi: Wilcox Group.

Project Goals and Objectives

There were many goals for this project. The first goal was to learn to use GIS to gather information and begin to create a picture of the character of the Leona Aquifer. To do this the data needed to be put into a more useable format. Most of the well data was stored in multiple ASCII text or .pdf files. The second goal of the project was to assess the extent of existing data in space and time. This is helpful in determining what can be done with existing data and where new data should be collected. The next goal was to use GIS to create a map to aid in locating wells for future data collection. The last two goals of this project were to examine spatial changes in water quality in the Leona aquifer and compare water quality within the Leona Aquifer to that of surrounding aquifers.

Methods

Data Acquisition

Several different types of data were used for this project. Some of the data was taken from class exercises used by the GIS for Groundwater Resources course, taught by Dr. David Maidment of The University of Texas at Austin.

The following are the data used:

· A Digital Elevation Model (DEM) was generated from the National Elevation Dataset archive for San Marcos Basin.

· The Hydrologic Unit Code (HUC) coverage for San Marcos Basin was obtained from the US Geological Survey.

· HydroEdge.shp file prepared by Dr. Maidment from National Hydrography Dataset data.

· Coverages of major roads and highways for Hays and Caldwell counties were obtained from the TNRIS web data page.

· Well data was obtained from Texas Water Development Board (TWDB) files. This data included well locations (degrees, minutes, and seconds), water levels, and water quality.

· Geology data was obtained from the Bureau of Economic Geology Geologic Atlas Seguin Sheet.

All of this data except the DEM and the geologic map were added to a geodatabase as personal a geodatabase feature classes. The raster data could not be stored in the geodatabase.

Data Processing

Well Data

The well data from the TWDB required some processing before it could be used in ArcGIS. First, the data was downloaded as ASCII tab delimited text files. Next, the data was opened using Excel and saved as an Excel spreadsheet. The well data was stored in three main tables: Well data, Water Quality, and Water Levels. The most important information stored in the Well data table is latitude and longitude of each well, the Aquifer code, and the well depth. The Water Quality table contained analyses of the major minor chemical constituents in the water and the date in which it was collected. The Water Levels table contained the depth to the water table from the land surface datum and water level change since the last measurement. The elevation of the water table was calculated by the following equation:

Water Table elevation = (Elevation of land surface datum – depth to water).

All wells were identified by the state well number in each of these tables. Selected data, described later, was organized in an Excel spreadsheet and saved as a DBF 4 *.dbf file. This file was imported into the geodatabase and then added to an ArcMap document.

Displaying Well as Points

To plot the wells on the map, I first displayed the X, Y data (Longitude, Latitude) as an event. This event had an attribute table and showed the location of the wells, but its versatility was limited. Next, this event was exported as a shapefile and added to the project as a layer. This new layer could be analyzed using tools such as spatial analyst.

Five main layers of well data were used in this project. All of the wells used were in Caldwell County. The first layer added was the well data for all wells in Caldwell County. The second, was well data for all wells in the Leona Formation. Water quality and water level layers required more sorting in Excel. The purpose of these layers was to display spatial distributions at during a relatively short span of time. For the water quality layers I chose to include data taken during the span of six months as one “time slice” and for the water level layer four months were included in the “time slice”. This makes the assumption that there are no seasonal variations. This may not be a true assumption, especially when dealing with water levels. To achieve the best spatial distributions it is necessary to find time slices in which there are many data points. The third and fourth layers were added using 1946 water quality data for the Leona Aquifer the Wilcox aquifer. This time period was a focus for this project because of the abundance of data points and because of interesting water quality trends discussed later. The fifth layer was water level data collected between January and March 1964.

DEM

For this project, the DEM was primarily used to examine the topography and geomorphology of the study area. A surface analysis was conducted using the spatial analyst tool in ArcGIS. Contours and slopes were drawn from the DEM to show the distinct topography of the Leona formation.

Figure 2: The surface analysis function.

Geologic Map Raster

The only geology data available for the area was a paper copy of the Bureau of Economic Geology Seguin Sheet. No digital data was available. In order to use the geologic map, the map was scanned and saved as a *.jpeg file to save file space. Later, the map may be stored as a higher resolution image. This image file was opened in ArcCatalog and the coordinate system was set to Universal Transverse Mercator (UTM) Geographic Coordinate System using the North American Datum 1927. Next, this map was loaded into ArcMap. Since this map was a scanned raster image, it contained no referencing information and needed to be geo-referenced.

These are the steps used:

1. Zoomed into a location where the coordinates were known (intersection of latitude and longitude).

2. Clicked add control point button on geo-referencing toolbar.

3. Clicked on a single point twice.

4. Clicked view link table button on geo-referencing toolbar.

The link table contained X and Y coordinates for the source (the unreferenced raster) and X and Y coordinates for the map (the actual coordinates of the control point).

5. Changed the X and Y values to the actual coordinates.

6. Repeated process three times.

7. Rectify. (This permanently attached the geo-referencing information to the image.)

This was mainly done as a learning experience so only three control points were selected. To achieve the best map many more control point should be added.

Spatial Analysis

Water Quality

Nitrate and Chloride levels were examined in the Leona Formation and in the Wilcox Group. One idea was to determine if the water flowing from the Leona had any effect on the water quality in the underlying Wilcox group. In order to best view the spatial distribution of water quality, it was necessary to use the spatial analyst tool to interpolate between data points and create a raster. Water quality data from 1946 was used. Both the inverse distance weighted (IDW) and kriging methods were used.

Water Levels

Water level data from 1964 was used to create a potentiometric surface for the Leona Aquifer. The interpolation to raster was done using the kriging method. Two points on the resulting surface were used to calculate the hydraulic gradient between the northwest and southeast ends of the aquifer.

Temporal analysis

Excel was used to plot graphs to illustrate changes over time. These graphs were connected to the spatial analyses discussed above. This juxtaposition of data allows data to be examined over both space and time. This method was used to look for trends in chloride levels in the City of Lockhart and water level fluctuations in the Leona Aquifer.

Results

Topographic Analysis

The surface analysis tools illustrated the unique topography of the Leona Formation. The Leona is a classic example of inverted topography. The alluvium that became the Leona was most likely deposited in a topographically low area. Modern drainage has cut down around the unit and now it is a topographically high area.

Figure 3: DEM and contours of the Leona Formation. Light blue on the DEM represents higher elevations.

The surface of the Leona Formation forms a very flat plain. The gradient between the two points located on figure 3 is 0.12 degrees. The very gentle topography is also demonstrated by a slope analysis. (Figure 4)

Figure 4: Slope analysis of the Leona Formation. Yellow areas indicate very gentle slopes. The red dots are wells in the Leona Aquifer.

Nitrate Distribution

Figure 5: 1946 Nitrate concentrations in mg/l (as nitrate) in the Leona Aquifer.

IDW interpolation of Nitrate in the Leona and Wilcox Aquifers highlighted a few nitrate hotspots. The maximum contaminant level for nitrate in drinking water is 44.27 mg/l as nitrate. (Explanation of TWDB Groundwater Database) This means that the dark blue areas were suitable (in terms of nitrate) for drinking water in 1946. These interpolations may exaggerate some of the higher concentration areas especially those based only on one data point. It may be unrealistic to assume that high nitrate is found a kilometer or more away from the tested well.

Note the nitrate hotspot southeast of Lockhart in Figure 5. Compare it to the high nitrate in found in the same area in the underlying Wilcox Aquifer (Figure 6). This may be showing a plume of water that is flowing out of the Leona into the Wilcox.

Figure 6: 1946 Nitrate concentrations in mg/l (as nitrate) in the Wilcox Aquifer.

Chloride Distribution

IDW interpolation of chloride data points shows similar areas with high chloride concentration. Figure 7. Some of these are the same wells that had high nitrate levels. Chloride is not a health risk; therefore it does not have a maximum contaminant level. Water does begin to become unpleasant to drink when the chloride concentration is more than 250 mg/l. This threshold is called the secondary maximum contaminate level.(ABAG, 1997) As before, the blue areas have the best drinking water in terms of chloride concentration.

Figure 7: 1946 Chloride concentrations in mg/l in the Leona Aquifer.

Excel graphs of Chloride concentration plotted over time show some interesting trends. Figure 8. In the early 1940’s Chloride levels were high but have decreased over time. Note that the graphs preserve the relative proportions of chloride that are shown in the spatial data. One possible explanation for the high chloride levels is the improper disposal of oil brines in the 1940’s.

Figure 8: Chloride level changes over time in the Leona Aquifer, Lockhart, TX.

Leona Water Table

Figure 9: Water Table in the Leona Aquifer in 1964.

Krige interpolation of water levels shows that the water table in the Leona Aquifer strongly mimics the surface topography. (Figure 9) Higher hydraulic head is found to the northwest and lower head to the southeast. The slope of the water table is 0.15 degrees which is nearly the same as the slope of the land surface. Groundwater should flow down this slight gradient toward the southeast. The krige does not illustrate water level changes in the north-south direction. Groundwater likely also flows toward the north and south margins of the unit.

Water level fluctuations in the Leona are shown by the graphs in figure 10. The eastern well shows about 12 feet of change and the other shows about 11 feet of change overall. This is important to notice when you consider that the wells are 29 ft and 14 ft deep respectively.

Figure 10: Water level fluctuations in the Leona Aquifer.

Geologic Map

Figure 11. Geo-referenced geologic map of the Leona Formation. The blue dots are wells in the Leona Formation.

Problems

A few problems were encountered in this project that need more work. The first problem was with the TxDOT highway data. Even when the same projection was set, the orientation of the data for different counties did not match each other. The second item deals with the well locations. There has been no ground truthing done to determine how accurately the wells are plotted on the map. Most of the Leona Aquifer wells are in the area mapped as the Leona Formation on the geologic map so the locations are approximately correct.

Future Work

There are a couple other tasks that could be done in the future. Land use data could be added to the database and used to determine possible sources of contaminants. It would also be useful to gather precipitation data to compare with the water level changes in the Leona Aquifer. Another possibility is to digitize a polygon around the Leona Formation and calculate the area which could be used in future modeling.

References:

ABAG Environmental Help line, 1997

http://www.abag.ca.gov/govnet/environmental_help/answers/112.html

Explanation of TWDB Groundwater Database

(http://www.twdb.state.tx.us/data/waterwell/groundwaterexplanation.htm)

Follett, C.R., 1966, Groundwater resources of Caldwell County, Texas: Texas Water Development Board: Report 12.