Abstract:
This study aims to integrate machine learning, hydrogeochemistry, and physical-based hydrological analysis techniques to first understand, and then replicate dynamics of the Casper Aquifer at a monitored mountain front location (Belvoir Ranch) under first drought, and then normal recharge conditions from the recent past. To accomplish this objective, existing surface and subsurface data, including hydrometeorological measurements and monitored groundwater and stream water levels and their hydrogeochemistry, will be jointly interpreted to identify aquifer recharge mechanism at the mountain front, i.e., timing, pathway, rate. An improved groundwater model, informed by this understanding, will be built, calibrated, and verified using both pumping test data and the monitored groundwater and stream water levels from the study area. With this model, we will predict aquifer dynamics (i.e., water level, flow rate, residence time, recharge area) under pumping and non-pumping conditions while considering a range of future recharge variability. Throughout this research, we will employ co-production principles by regularly seeking and implementing input and feedback from the water managers of the Cheyenne Board of Public Utilities (BOPU), our main stakeholders for the area. Our study method, if successful, will contribute to the development of a sustainable aquifer development plan for BOPU that will account for recharge variability. Though this study will focus on one mountain front location as a testbed, our method integrating physical, hydrogeochemical, and statistical techniques can be extended to evaluating other aquifers within and outside the State facing recharge variability and/or water development pressures. For the Lone Tree Creek subwatershed in Belvoir Ranch, preliminary results indicate a high potential for the proposed study to identify localized recharge pathways to the aquifer from existing surface and subsurface characterization and monitoring data. Therefore, a hydrostratigraphic model created from a past WRP grant can be quickly updated to incorporate localized flowpaths identified through this research.
