Evaluation of ground-water flow and land-surface subsidence caused by hypothetical withdrawals in the northern part of the Gulf Coast Aquifer system, Texas

Abstract

During 2003–04 the U.S. Geological Survey, in cooperation with the Texas Water Development Board (TWDB) and the Harris-Galveston Coastal Subsidence District (HGCSD), used the previously developed Northern Gulf Coast Ground-Water Availability Modeling (NGC GAM) model to evaluate the effects of hypothetical projected withdrawals on ground-water flow in the northern part of the Gulf Coast aquifer system and land-surface subsidence in the NGC GAM model area of Texas. The Gulf Coast aquifer system comprises, from the surface, the Chicot and Evangeline aquifers, the Burkeville confining unit, the Jasper aquifer, and the Catahoula confining unit. Two withdrawal scenarios were simulated. The first scenario comprises historical withdrawals from the aquifer system for 1891–2000 and hypothetical projected withdrawals for 2001–50 compiled by the TWDB (TWDB scenario). The projected withdrawals compiled by the TWDB are based on ground-water demands estimated by regional water planning groups. The second scenario is a “merge” of the TWDB scenario with an alternate set of projected withdrawals from the Chicot and Evangeline aquifers in the Houston metropolitan area for 1995–2030 provided by the HGCSD (HGCSD scenario). Under the TWDB scenario withdrawals from the entire system are projected to be about the same in 2050 as in 2000. The simulated potentiometric surfaces of the Chicot aquifer for 2010, 2020, 2030, 2040, and 2050 show relatively little change in configuration from the simulated 2000 potentiometric surface (maximum water-level depths in southern Harris County 150–200 feet below NGVD 29). The simulated decadal potentiometric surfaces of the Evangeline aquifer show the most change between 2000 and 2010. The area of water levels 250– 400 feet below NGVD 29 in western Harris County in 2000 shifts southeastward to southern Harris County, and water levels recover to 200–250 feet below NGVD 29 by 2010. Water levels in southern Harris County recover to 150–200 feet below NGVD 29 by 2020 and remain in that range through 2050. A relatively small cone of depression in southern Montgomery County that did not appear in the 2000 surface develops and enlarges during the projected period, with a maximum depth of 250–300 feet below NGVD 29 in 2030, 2040, and 2050. The simulated decadal potentiometric surfaces of the Jasper aquifer each have a major cone of depression centered in southern Montgomery County that was minimally developed in 2000 but reaches depths of 550–650 feet below NGVD 29 in the 2020, 2030, 2040, and 2050 surfaces. Under the TWDB scenario the percentage of withdrawals supplied by net recharge increases from 75 percent in 2000 to 87 percent in 2050, and the percentage of withdrawals supplied by storage decreases from 25 percent in 2000 to 13 percent in 2050. Under the HGCSD scenario, withdrawals from the Chicot and Evangeline aquifers increase about 74 percent during 1995–2030; Jasper aquifer withdrawals are unchanged from those of the TWDB scenario. For the 2010, 2020, and 2030 potentiometric surfaces of the Chicot and Evangeline aquifers, the substantially greater withdrawals of the HGCSD scenario relative to those of the TWDB scenario result in progressively deeper cones of depression than those in the potentiometric surfaces associated with the TWDB scenario—for the Chicot aquifer in southern Harris County, 400–450 feet below NGVD 29 in 2030; for the Evangeline aquifer in southern Montgomery County, 700–750 feet below NGVD 29 in 2030. Although Jasper aquifer withdrawals are the same for both scenarios, the major cone of depression centered in southern Montgomery County in the 2030 potentiometric surface is 50 feet deeper at its center (600–700 feet below NGVD 29) than the cone in the 2030 surface under the TWDB scenario. Under the HGCSD scenario, the percentage of withdrawals supplied by net recharge decreases from 72 percent in 1995 to 57 percent in 2030, and the percentage of withdrawals supplied by storage increases from 28 percent in 2000 to 43 percent in 2030. About 85 percent of the increase supplied by storage is from the compaction of clay. Land-surface subsidence in the major area of subsidence centered in Harris and Galveston Counties during 2000–50 that results from simulating the TWDB withdrawal scenario expands slightly to the west and increases in places. The maximum change occurs in the Conroe area where subsidence increases from about 4 to about 13 feet during the projected period. Land-surface subsidence in the major area of subsidence during 1995–2030 that results from simulating the HGCSD withdrawal scenario increases substantially. For example, in east-central Harris County maximum subsidence increases from about 10–11 feet in 1995 to 22 feet in 2030. The hypothetical projected withdrawal scenarios are estimates of future withdrawals and might not represent actual future withdrawals. The simplifying assumptions that the downdip limit of freshwater flow in each hydrogeologic unit is a stable, sharp interface across which no flow occurs and that the base of the system is a no-flow boundary become less realistic and thus increase the uncertainty in results as drawdowns increase. The presence of uncertainty dictates that the results of the predictive simulations described in this report be used with caution in any decision-making process.