Impact of Vertical Discretization of Aquifers in Models on Simulation of Stream Aquifer Interaction

Stream-aquifer interaction will be one of the most difficult sustainability indicators to be evaluated by all the Groundwater Sustainability Agencies (GSAs) under California’s Sustainable Groundwater Management Act (SGMA) due to its complicated nature and lack of general understanding of the process and appropriate data. Integrated water resources models will serve as important tools for understanding and assessment of stream-aquifer interactions in many groundwater basins in California.

This paper investigates the minimum, optimum, and desirable number of aquifer layers needed to simulate stream-aquifer interactions with desired level of accuracy using two integrated water resources models, CA DWR’s IWFM and USGS’s MODFLOW models. Finer vertical layering will theoretically provide better estimates of stream-aquifer interaction if supported with appropriate filed data. However, in a real world, the calibration parameters might take away the gains made by finer discretization, specifically for big scale regional models. Many test cases were developed to evaluate the sensitivity of vertical discretization on model output related to stream-aquifer interaction due to pumping from a well based on the distance of the well from the stream, depth of pumping and range of various aquifer and stream bed parameters such as aquifer hydraulic conductivity, storage coefficient, and stream bed conductance.