The Mississippi River Valley alluvial aquifer is a water-bearing assemblage of gravels and sands that underlies about 32,000 square miles of Arkansas, Kentucky, Louisiana, Mississippi, Missouri, and Tennessee. Pumping of groundwater from the alluvial aquifer for agriculture started in the early 1900s in the Grand Prairie area for the irrigation of rice and soybeans. From 1965 to 2005, water use in the alluvial aquifer increased 655 percent. In 2005, 6,242 million gallons per day of water were pumped from the aquifer, primarily for irrigation and fish farming. Water-level declines in the alluvial aquifer were documented as early as 1927. Long-term water-level measurements in the alluvial aquifer show an average annual decline of 1 foot per year in some areas.

In this report, the utility of the updated 2009 MODFLOW groundwater-flow model of the alluvial aquifer in northeastern Arkansas was extended by performing groundwater-flow assessments of the alluvial aquifer at specific areas of interest using a variety of methods. One such area is along the western side of Crowleys Ridge, which includes western parts of Clay, Greene, Craighead, Poinsett, Cross, St. Francis, and Lee Counties. This area was designated as the Cache Critical Groundwater Area by the Arkansas Natural Resources Commission in 2009 for the alluvial and Sparta/Memphis aquifers, because of the rate of change in groundwater levels and groundwater levels have dropped below half the original saturated thickness of the alluvial aquifer.

Three scenarios were simulated, in part, to allow assessment of the role that pumping in Jackson and Woodruff Counties has on water levels and flow rates into and out of the cone of depression located along the western side of Crowleys Ridge. In scenario 1 (the baseline scenario), the 2005 pumping rate is applied from 2005 through 2050 without change. In scenario 2, pumping is the same as in scenario 1 except that the pumping rate in Jefferson and Woodruff Counties is specified as zero from 1998 to 2050. In scenario 3, pumping is the same as in scenario 1 except that the pumping rate in Jefferson and Woodruff Counties is specified as half the rate specified for stress periods from 1998 to 2050.

Particle tracking using MODPATH was done to assess the direction and time of travel that particles take from specific model locations. Factors that affect how far particles will travel include: (1) pumping rates in the vicinity of particles, (2) when and if model cells with pumping wells go dry, and (3) changes in pumping rates during the simulation period. Particles can travel further if cells do not go dry and pumping in a model cell can continue even if the total pumping rate from the model is specified with a lesser rate. Particles introduced in Jackson and Woodruff Counties travel somewhat further for scenario 2 than for scenario 1.

Flow vectors were generated using the built-in graphics capability in the Groundwater Modeling System. Groundwater-flow vectors depict both magnitude and direction of simulated groundwater flow. Groundwater-flow magnitude is the product of the hydraulic gradient and hydraulic conductivity. Only two values of hydraulic conductivity (230 and 730 feet per day) are specified over the model domain. Variability in groundwater-flow magnitude is caused largely by variation in hydraulic gradient. Groundwater-flow vectors and hydraulic-head maps for scenarios 1 and 2 for the beginning of 2010 and the beginning of 2050 were generated to allow for a comparison of flow rates and direction. The lengths of the vectors presented represent the horizontal magnitude of flow, and the orientation of each vector indicates the horizontal direction of flow.

Zone-budget analyses using ZONEBUDGET were performed to assess the rate of water entering and leaving specified zones within the model. Zone-budget analysis was performed on the simulated groundwater flow by dividing the model into four separate zones and calculating the individual flow components for each zone at different simulation times for the three pumping scenarios. Reduction of pumping in scenarios 2 and 3 resulted in substantially more groundwater flow into the counties east of Jackson and Woodruff Counties, and more flow to rivers within the model area.

First posted December 2, 2010