Sensitivity of Surface Water/Groundwater Interaction Models to Land Surface Elevation Data in Riparian Areas

Small fluctuations in aquifer water table elevations can result in dramatic changes in evapotranspiration (ET) rates and the associated river gains/losses in riparian areas.  In numerical models such as MODFLOW, the magnitude of ET flux can be very sensitive to the simulated depth to groundwater, particularly in arid regions where ET loss is a major component of the water balance. This sensitivity suggests the need for high-resolution and accurate data sets of land surface elevation from which depth to water and ET calculations are made. Digital elevation models (DEMs) provide relatively uniform and reliable elevation data for large regions and are easily acquired, but often do not capture the subtle topography variations required to accurately characterize the impacts on groundwater levels from riparian area ET.  An analysis of the sensitivity of surface water/groundwater models to elevation data was simulated using 10-meter DEMs and 1-foot Light Detection and Ranging (LIDAR) data. The data sets were incorporated into a MODFLOW model with 125ft by 250ft cells along a 2-mile river corridor in the Southwest US. A 2-meter vegetation classification coverage was utilized for all model runs. While data point resolution is an obvious difference between the two data sets, differences are also noted with respect to the average cell elevation and representation of topographic features. Results compare the distribution and magnitude of fluxes using the two approaches, with comment on the data set limitations.