Aquaculture and Irrigation Water Use Model (AIWUM) 2.0 Input and Output Datasets for the Mississippi Alluvial Plain, 2014-2020 (ver. 1.1, April 2026)

The Aquaculture and Irrigation Water Use Model (AIWUM) 2.0 employs remote sensing data sets and Distributed Random Forests (DRF), an ensemble machine learning algorithm, to predict annual groundwater use associated with irrigation and aquaculture from 2014 to 2020 in the Mississippi Alluvial Plain (MAP) and Mississippi embayment regions at 1-kilometer and 100-meter resolution (Majumdar and others, 2023). The response variable for the model are groundwater withdrawal rates from flowmeters in the Mississippi delta region associated with the Delta Voluntary Metering Program (DVMP). Explanatory variables include a variety of climate, remote sensing, and GIS data. Annual predictions of groundwater use from the DRF are disaggregated to monthly groundwater use using crop-specific groundwater withdrawal data from existing flowmeters across the MAP. This data release provides input and output datasets associated with AIWUM 2.0. Processing steps for attributing explanatory variables to groundwater withdrawal rates at flow-meters are available in metadata for the Inputs found in this data release and at the companion software release (https://code.usgs.gov/map/wu/aiwum-2.0-hydromap_ml-mirror). This data release was originally published on January 23, 2024, and revised in April 2026. The revision corrects errors in all 1 km millimeter-depth raster files. During resampling from 100 m resolution to 1 km resolution, a scaling issue caused the values in the 1 km rasters to be 100 times larger than intended. The corrected files are included in this revision.