The Hydrogeology and Hydrochemistry of the Southern High Plains
James Bene’ - GIS in Water Resources, Fall 1999
Introduction
Because of mankind’s dependence on the continued availability of usable water, the long-term effect of land development on the hydrology of semi-arid areas is of great interest.Over the last two centuries, large areas of the Texas Panhandle have been developed as farmland and rangeland.In order for these industries to be economically feasible in a relatively arid region, farmers and ranchers must irrigate heavily.Unlike eastern, wetter areas of Texas that primarily use surface water for irrigation, most of the water used for irrigating the Southern High Plains is obtained by extracting groundwater.
The principal aquifer for this area is the Tertiary Ogallala Formation, an areally extensive sand and gravel aquifer that supplies potable water for most of the Texas Panhandle, western Kansas and a large section of central Nebraska.The Panhandle resident's dependence upon this source for both drinkable and irrigation water is causing change in the hydrologic cycle of the area.The high rate of aquifer withdrawal coupled with the large amounts of time needed for surface water to recharge (due to the low rate of precipitation and the high rate of evaporation—see figures 1 and 2) the Ogallala has caused large declines in the elevation of the water table in some areas.These declines are likely to continue as development continues northern Texas and the High Plains.
Not only is the hydrologic balance altered by large-scale
irrigation, but the hydrochemistry of the surface water and groundwater is also
changed.High rates of evaporation of irrigated water can cause an unwanted
concentration of soluble mineral components in irrigated soils, which can then
slowly migrate to the water table, changing the chemical composition of water
in shallower wells.Similarly, the long-term application of fertilizers and
pesticides to farm and rangeland can also adversely affect the quality of the
water supply.
The Ogallala aquifer of the Texas Panhandle covers approximately 20,000 square miles; therefore, the study area was restricted to a smaller, more manageable region that adequately incorporates the processes and conditions common throughout the area.Because the city of Lubbock, Texas is undergoing growth in an area where heavy use of groundwater for irrigation has been a way of life for many decades, the region surrounding it is ideal.The well information datafiles needed were only available in a per-county format, therefore, the nine counties surrounding the city of Lubbock, Texas(Lamb, Hale, Floyd, Hockley, Lubbock, Crosby, Terry, Lynn and Garza Counties) were chosen as a representative study area (fig. 3).
Geomorphology of the Ogallala Aquifer
The Ogallala is a large aquifer, extending from South Dakota to the Midland-Odessa region of North-Central Texas (Figure 4).It is composed primarily of sand, silt and gravel shed from the Rocky Mountains during the Tertiary Period (65-2 m.y.).The geometry of these deposits is that of a complex of braided stream and channel deposits that can impose highly heterogeneous and anisotropic local flow conditions on the groundwater (Seni 1980).
The thickness varies widely; the Ogallala can be over 1,000
feet thick in parts of Nebraska and Kansas, but the thickness in the Texas
Panhandle is generally from 500 feet to absent (Seni 1980).This variability is
due to the depositional environment at the time the sediments were laid down.As
material was shed from the Rockies during the Tertiary Period, the debris first
filled existing channels and valleys in the underlying Cretaceous (120 – 65 my)
and Triassic (235-195 m.y.) sediments.These paleo-channels were thought to have
formed because of widespread erosion during the Jurassic Period (195-141
m.y.).Eventually, the eastward movement of sediments to the Lubbock Area was
halted with the establishment of the Pecos River Basin in the late Pleistocene.
(Gustavson and Finley, 1985)
The Ogallala underlies approximately 85% of the nine counties in the study area, with only the southeast portion of Garza County and the northeast corner of Floyd County as exceptions.The differentiation between Ogallala and non-Ogallala is relatively obvious because the Ogallala forms an escarpment that can show hundreds of feet of vertical relief from the areas not covered.The edge of this escarpment forms a boundary between a playa lake dominated, primarily non-linear drainage system on the highlands, to a more familiar, linear drainage network at lower elevations.In effect, the southeastern flowing waters of the Ogallala emerge as spring water at the eastern edge of the outcrop.
Data Retrieval
In order to characterize the flow pattern and chemistry of the groundwater extracted in the Lubbock area, county well-information files were downloaded via the internet from the Texas Water Development Board (TWDB).First, nine well-information files were downloaded for the nine counties of the study area.Each of these files contains the position of registered wells contained within a county in geographic coordinates (deg, min, sec).These files also contain information such as TWDB well identification number, aquifer of completion, land surface elevation of the well head, well owner, etc.
Next, county well water-level information files were
downloaded.These files contain multi-year data on the water level in the
registered wells in each county.Unfortunately, these files reference all data
to the TWDB well identification number and not to a specific
location.Therefore, the entries in these files must be collated and linked to
the well location information located in the well-information files.Because the
data in the county water-level files is sporadic in their temporal and spatial
data distribution (not all wells were sampled every year, new wells were
drilled, old wells abandoned, some null data, etc.) it was found that
pre-processing of the data files was preferable to attempting to manipulate the
information in ArcView exclusively.In order to achieve this, a computer program
in the Turbo-Pascal language was developed.This program converts the well
location to decimal degrees, extracts all relevant data from the water-level
data sets, copies the information in an annual record format to a binary
file.This binary file was later queried, and the results written to a
tab-delimited text file for import into ArcView.
A similar procedure was used to collate the chemistry data
records:
County water quality data-files are also available for
internet download from the TWDB.As with the water-level files, the chemical
data are referenced to specific well-identification numbers listed in the
county well-information files.As with the water-level files, the information
contained in the quality files is in a multi-year format not easily manipulated
in ArcView.In addition, much of the data in the quality files must be “weeded
out” prior to regional water-quality evaluation because of the sporadic nature
of sampling well water quality.Not only were wells inconsistently sampled over
time, but a significant amount of reported data is unusable because of errors
in lab or recording procedures.(TWDB – http://www.twdb.state.tx.us/).A
second program was developed in Turbo-Pascal to extract valid data and record
it in binary-record format to facilitate speedy searches.
Once the water-level and quality data have been extracted
and collated into binary records, these records were queried by year and the
results written to individual tab-delimited text files suitable for import
into, and display by ArcView.
Water-Table Changes
The residents of the Lubbock area face a unique problem: In order for the region to continue to grow and develop, it will continue to be necessary to exact ground water from the Ogallala for irrigation, industrial and civil use.Unfortunately, although the Ogallala is very large, it is not infinite in its capacity.Figure (6) shows the distribution of registered wells in the Lubbock area.
The TWDB states that, for any given county, there are probably many unregistered wells in addition to the ones listed in their datasets.Bearing this in mind, Figure (6) shows a surprising number of wells in the region (approx. 1200).Figure (7) shows an interpolated surface based on the 1996 point water table elevation data.A distinct hydraulic gradient produces flow from northwest to southeast.However, it should be noted that the 300 m drop in water-table elevation over approximately 100 miles produces a relatively flat gradient of 0.002.This indicates that flow is steady, but slow through the matrix.
Figure (8) shows the water table elevation for 1967.A gradient of similar magnitude is apparent, indicating that no radical shift in the hydrology has occurred in the past thirty years.
However, when the interpolated water table surface of 1967 is subtracted from the 1996 surface, a distinct pattern of change is seen.
As Figure (9) illustrates, the northern portion of the study area shows a drop in water-table elevation of up to 40 m, while the southwest portion shows either no significant change or an increase in elevation.The reader is cautioned not to impart too much importance to small areas of significant change (such as the southeastern edge showing over 60 m of drop) due to the unverifiable nature of single-point data.Instead, it is preferable to make regional generalizations based on data from several locations.
The large drop in water-table elevation seen in the northern portion of the study area was expected; depletion of the Ogallala’s waters has been documented throughout the High Plains.If water is extracted from aquifer more quickly than it is recharged, the water table will drop as the aquifer is essentially mined of its groundwater.
The region of rise in water-table elevation was
unexpected.This probably indicates that much less water is being pumped in that
area than in previous years or that that region is receiving relatively rapid
recharge from areas outside the study region.Because discharge rates from wells
in Texas are not recorded (with the exception of some public wells), it is
impossible to be certain that a region has changed the volume of water
extraction over a period of time.There is evidence, however, that the section
of aquifer showing water-table rise may not be very well connected
hydraulically with the Northern region.Seni (1988) suggested that a separate
lobe of Rocky Mountain sediments is located in the southeastern region of the
study area.It is also known that an erosional remnant of Cretaceous limestone
underlies the Ogallala in the Southwest area, but is absent in the Northeast
(Reeves 1996).
Water Quality Evaluation
In order to characterize the impact of long-term extraction of groundwater on the quality of water in the Ogallala, a variety of methods were used.First, in order to classify the overall distribution of general water types, well data for all years form 1969 to 1996 were queried and extracted from the pre-processed binary record file.The pre-processing program also calculated and reported a unique integer value for each water calculated using the following procedure:
1) Because all ion concentrations were reported in mg/l, the preprocessor converted each gravimetric concentration to molar concentration.This is accomplished by dividing each ion’s weight concentration by its molar mass and recording the value.
2) The molar concentrations for all cations in each individual water sample were summed, and the same process was repeated for each sample’s anions.
3) Each cation’s molar concentration was divided by the sum of cation concentration to find that constituent’s percent of total.Each anion’s percent concentration was calculated in a similar procedure.
If an anion’s percent concentration was greater than 50%, then the water was placed into one of three categories:
An integer value of 1-4 was applied if S04 concentration was greater than 50%.
An integer value of 5-8 was applied if Ca concentration was greater than 50%.
An integer value of 9-12 was applied if HCO3 concentration was greater than 50%.
If no anion constituted more than 50% of the total, then an integer range of 13-16 was assigned.
To further classify a water sample, the integer ranges assigned through anion percentage were subdivided using a similar scheme of cation percentage:
The following table illustrates these subdivisions:
|
Dominant Anion/Cation |
|
|
|
|
|
Ca |
1 |
5 |
9 |
13 |
|
Na |
2 |
6 |
10 |
14 |
|
Mg |
3 |
7 |
11 |
15 |
|
Mixed |
4 |
8 |
12 |
16 |
Once these unique classifications were assigned and recorded, the values were plotted in figures (10,11).
In both cases, interpolated surfaces were calculated in ArcView from the point data imported from the pre-processing program.However, due to the relatively random cation distribution, the cation plot (fig. 11) is shown as plotted points to eliminate the obvious errors incurred during the surface interpolation process.
As stated above, the relatively random distribution of cations shows that no distinct correlations or patterns should be interpreted from this, and suggests that the groundwater flow pattern has been relatively stable over time.
The distribution of anions shown in Figure 10 suggests a somewhat different hypothesis.The segregation of anion classes into distinct areas shows that there may be regions of the aquifer that are hydraulically and/or mineralogically distinct from others.It should also be noted that, although the dominant anion distribution is much more segregated than that of the cations, there is still a relatively large region where waters with mixed anion concentrations are prevalent.
In addition to broadly characterizing well waters according to the dominant anions and cations, regional distributions of individual ionic constituents were plotted.Interpolated surfaces calculated form point data for a variety of ions in 1996 and 1969/70 are shown in figures (12a-12t) below.The records from 1969 and 1970 were combined to afford better aerial coverage of the study area.
As figures (12a-12t) illustrate, it can be difficult to discern differences between the two sets of plots.To illuminate the differences over time, each interpolated surface from 1967/70 was subtracted from the 1990 surface and plotted in the following figures.
Again, it should be noted that individual points of excessive difference could be an artifact of the inconsistent nature of each county’s sampling procedure.
Definite long-term trends are obvious in several plots.One of the most distinct change from 1970 to 1996 occurred in the concentration of potassium, which almost uniformly increased across the entire area of study (fig. 13d).In light of the general lowering of pH of groundwater from 1969/70 to 1996 shown in Figure 13j, a possible explanation for the potassium increase is the widespread agricultural use of potassium sulfate (K2SO4) as a pH neutralizer for basic soils.Sulfate ion concentration also shows a definite increase in the southwestern portion of the area, but appears to have been stable in the northeastern section of the study region.This apparent inconsistency is probably due to the difference in the scale of increase shown in these plots and the 2:1 potassium to sulfate dissolution ratio of the K2SO4 molecule.
Figure 13g shows a distinct increase in nitrate concentration.When the nitrate concentration difference plot is compared to the nitrate concentration plots (figs. 12o, 12p), we see that a significant portion the nitrate in the Lubbock area groundwater was introduced within the last thirty years.This finding is consistent with the heavy agricultural use of fertilizers such as ammonium nitrate.Because the federal maximum allowable concentration standard is 44mg/l for nitrate, this relatively rapid increase is cause for concern.
Total dissolved solid concentration also shows an unwanted increase in the southwest (fig. 13i).Waters with a TDS above 1000 mg/l are considered unusable for civil drinking water, and if it continues, this trend could severely restrict future development of areas near Lubbock.
Ion ratios provide evidence of the mineralogy of the rocks that make up an aquifer.A one-to-one ratio of sodium to chloride ion concentrations in sampled waters suggests that the water has interacted with salt along its flowpath.Similarly, ion ratios equal to one for calcium/magnesium and calcium/sulfate can indicate that the water has come into contact with dolomite and gypsum respectively.The following figures depicting these ratios (for 1996) suggest that there is a diverse assemblage of minerologies present in the Ogallala.
Conclusion
The Ogallala Aquifer of the Southern High Plains is a fragile natural resource that can and is being altered by extraction of its waters for commercial and civil use.Because this region experiences little precipitation and large amounts of evaporation, the Ogallala in the Lubbock area receives much less recharge than in northern states.The result of this is that the Ogallala is essentially being mined of its groundwater in some areas.The depletion of this essential resource will stunt the commercial productivity of the region, which will most likely severely restrict the area’s population growth.
The continued alteration in the quality of groundwater is also of concern.Heavy use of fertilizers and soil neutralizers on the sandy Panhandle soil, while probably necessary for high agricultural productivity, is undoubtedly changing the soil and water chemistry.Some of these effects are harmless, however, too great an increase in concentrations of compounds such as nitrate could render the groundwater unusable for years.The concentration of otherwise innocuous dissolved solids can also make waters useless for use by man.Repeated cycles of aquifer pumpage and irrigation followed by evaporation causes soluble materials to collect in soils.These compounds are then picked up by the sparse surface infiltration and redeposited in the groundwater reservoir, causing total dissolved solids concentrations to rise.This is a somewhat unavoidable consequence of irrigation in arid areas, but is detrimental to groundwater and soil quality, none the less.
Until recently, it has not been possible to study the long-term effects of heavy, regional pumping on a large aquifer like the Ogallala.Once thought to be an infinite resource, it is now apparent that man has the capability to affect not only the quantity, but also the quality of the water available in the future.
References
Gustavson, T.C., Finley, R.J.,1985, Late Cenozoic geomorphic evolution of the Texas
Panhandle and northeastern New
Mexico: Bureau of Economic Geology Report of Investigations no. 148, 42p.
Seni, S.J., 1980, Sand-body geometry and depositional systems, Ogallala Formation,
Texas:University of Texas Bureau of
Economic Geology Report of Investigation 105, 36p.