Integrating High-resolution Satellite Remote Sensing and Climate Data to Quantify Irrigation Groundwater Withdrawals and Consumptive use in the Western U.S.

Abstract Submission: In the Western U.S., the combination of ongoing and projected droughts, rising irrigation water demands, and population growth is expected to intensify groundwater consumption. Despite the pressing need to address these challenges, most irrigation systems in this region are not equipped with flowmeters required to monitor groundwater withdrawals, which is crucial to implementing sustainable water management practices. However, metering is not a trivial solution, as meters can often be faulty or inadequately calibrated, resulting in discrepancies in the recorded readings. Therefore, developing reliable and efficient solutions for monitoring groundwater withdrawals is paramount in addressing the urgent water management concerns in the Western U.S.

The existing methods for estimating withdrawals either entail significant costs and time (e.g., process-based models) or are not suited to support local-scale water management. Building on our prior research, here, we rely on Landsat actual evapotranspiration (ET) from OpenET, irrigation data (field boundaries, water source type), and climate datasets (gridMET, CONUS404, Daymet) to estimate annual groundwater withdrawals, consumptive use, and irrigation efficiencies in Nevada, Oregon, and Arizona. We use statistical (linear regression and bootstrapping) and machine learning (Random Forests, XGBoost, LightGBM) approaches and compare our groundwater withdrawal estimates with in-situ meter data at multiple spatial scales— field (30 m-100 m), local (2 km), and individual groundwater basin scales. Our models can explain 50%-80% variance in withdrawal depths and 90% variance in withdrawal volumes across these spatial scales. The estimated irrigation efficiencies (80%-90%) also align with known irrigation system efficiencies.

While these groundwater withdrawal estimates can be further improved, we consider our approach to be more accurate than simply relying on common water right duties, potential crop ET-based estimates, or assumed values. Ultimately, we aim to empower water resource communities by improving water budget information and facilitating the implementation of groundwater management plans throughout this region

Learning Objectives/Expected Outcome (Optional) :