WELLS FOR HYDROGEOLOGY: IMPROVING, EXPANDING, AND AUTOMATING GEOLOGIC INPUTS TO NEAR-SURFACE GROUNDWATER FLOW MODELS USING EXISTING WELL DATABASES

As the scarcity of surface water continues to increase and industrial pollution threatens existing groundwater reserves, understanding the controls on groundwater flow is an increasingly important goal for scientists, planners, policy makers, and other community stakeholders. This is especially true for communities that rely primarily or solely on groundwater as a water source. Among the most important controls on groundwater flow in near-surface aquifers is the distribution of different texture and lithology of the sediments within them. In many places, the most abundant and precise source of this information is contained in well logs or descriptions. In the United States, many statewide geological surveys or other agencies assemble and maintain databases of this information from water well drillers, geotechnical engineers, environmental consulting agencies, or geologists. These databases provide a wealth of information, but often must be converted from raw descriptions in human-readable format to discretized lithologies and, potentially, target materials that can be used to define property zonation in computer models, such as MODFLOW. Described here is a new process using an open-source python toolset for creating n-layer hydrogeologic models from well datasets such as these. This process can be applied anywhere appropriate datasets exist. These datasets include a) a database of wells with unique keys (e.g., API numbers), geospatial coordinates, and depth-defined geologic descriptions; b) a surface elevation raster defining the top of the hydrogeologic model, and c) a bedrock elevation raster (or another raster) defining the bottom of the model. The process described can be applied across broad scales, including county-to-sub-county scales (~1000 km²) and state-wide scales (~100,000+ km²). Also presented are lessons learned from cases in near-surface glacial-fluvial aquifers in Illinois where it has been tested. The core methods used here have been implemented elsewhere, but the tools presented represent a new low-cost and low-barrier means of rapid characterization of any near-surface aquifers of interest.