Groundwater overdraft in the western United States has prompted water managers to develop groundwater management plans that include mandatory reporting of groundwater pumping (GP). However, most irrigation systems in this region are not equipped with irrigation water flow meters to record GP and performing quality control of the available metered GP data is difficult due to the scarcity of reliable secondary GP estimates. We hypothesize that Landsat-based actual evapotranspiration (ET) estimates from OpenET can be used to predict GP and aid in quality control of the metered GP data. The objectives of this study are to: 1) pair OpenET estimates of consumptive use (Net ET, i.e., actual ET less effective precipitation) and metered annual GP data from Diamond Valley, Nevada, and Harney Basin, Oregon; 2) evaluate linear regression and machine learning models to establish the GP vs Net ET relationship; and 3) compare GP estimates at the field- and basin-scales. Results from using a bootstrapping technique showed that the mean absolute errors and root mean square errors for field-scale GP depth are ∼11 % and ∼14 % across Diamond Valley and Harney Basin based on the OpenET ensemble mean, which showed the highest skill among all the OpenET ET models. Moreover, the regression models explained 50 %-70 % variance in GP depth and ∼90 % variance in GP volumes. Our GP volume estimates are also within 7 % and 17 % of the total reported and measured volumes in Diamond Valley and Harney Basin, respectively, and the estimated average irrigation efficiency of 87 % aligns with known center-pivot system efficiencies. Additionally, the OpenET ensemble proves to be useful for identifying discrepancies in metered GP data, which are subsequently flagged as outliers. Results from this study illustrate usefulness of satellite-based ET estimates for estimating GP and metered GP data quality control and have the potential to help estimate historical GP.

Groundwater plays a crucial role in sustaining global food security but is being over-exploited in many basins of the world. Despite its importance and finite availability, local-scale monitoring of groundwater withdrawals required for sustainable water management practices is not carried out in most countries, including the United States. In this study, we combine publicly available datasets into a machine learning framework for estimating groundwater withdrawals over the state of Arizona. Here we include evapotranspiration, precipitation, crop coefficients, land use, well density, and watershed stress metrics for our predictions. We employ random forests to predict groundwater withdrawals from 2002-2020 at a 2 km spatial resolution using in-situ groundwater withdrawal data available for Arizona Active Management Areas (AMA) and Irrigation Non-Expansion Areas (INA) from 2002-2009 for training and 2010-2020 for validating the model respectively. The results show high training (R2≈ 0.86) and good testing (R2≈ 0.69) scores with normalized mean absolute error (NMAE) ≈ 0.64 and normalized root mean square error (NRMSE) ≈ 2.36 for the AMA/INA region. Using this method, we spatially extrapolate the existing groundwater withdrawal estimates to the entire state and observe the co-occurrence of both groundwater withdrawals and land subsidence in South-Central and Southern Arizona. Our model predicts groundwater withdrawals in regions where production wells are present on agricultural lands and subsidence is observed from Interferometric Synthetic Aperture Radar (InSAR), but withdrawals are not monitored. By performing a comparative analysis over these regions using the predicted groundwater withdrawals and InSAR-based land subsidence estimates, we observe a varying degree of subsidence for similar volumes of withdrawals in different basins. The performance of our model on validation datasets and its favorable comparison with independent water use proxies such as InSAR demonstrate the effectiveness and extensibility of our combined remote sensing and machine learning-based approach.