Data for Modeling Interbasin Transfers of Water in Colorado and the Northeast Region, United States.

Data used to predict flow characteristics of transfers of water between hydrologic basins at the hydrologic unit code 8 (HUC8 scale) using tree-based ensemble models—random forest models for Colorado and M5 cubist models for the Northeast Region (parts of Pennsylvania, New Jersey, and New York)—are presented and documented in this data release. Interbasin transfers (IBTs) of waters are important components of water balances of basins and can have substantial impact on national and (or) regional water availability for a variety of human uses, such as public supply and irrigation. This data release contains all input files necessary to reproduce the results of the flow prediction models described in the associated journal paper (Eng and others, 2024). For reasons described in the journal paper, there are differences in input data, model approach, and flow prediction models between the models for Colorado and the Northeast Region.

This data release presents five different types of files, all in csv format: (1) NE_flow_char.csv and CO_flow_char.csv present interbasin flow information from reported data sources for the Northeast Region and Colorado, respectively; (2) NE_class_bootstraps.csv and CO_class_bootstraps.csv present the list of eight HUC8 basins for each of 100 validation runs of the classification models for the Northeast Region and Colorado, respectively; (3) NE_CUB_bootstraps.csv and CO_RF_bootstraps.csv present the list of eight HUC8 basins for each of 100 validation runs of the flow prediction models for the Northeast Region and Colorado, respectively; (4) NE_predictors.csv and CO_predictors.csv present predictor variables for the M5 cubist and random forest models for the Northeast Region and Colorado, respectively; and (5) Predictors_DataDictionary.csv, which defines the 166 predictor variables presented in the files NE_predictors.csv and CO_predictors.csv.