The data were generated by modeling wetted width, thalweg depth, bankfull width, and bankfull depth measurements from the USEPA's National Rivers and Streams Assessments. Models were developed with StreamCat data (https://www.epa.gov/national-aquatic-resource-surveys/streamcat-dataset-0) as predictor variables in random forest models. Models were then applied to perennial NHDPlus (version 2.1) stream segments to produce 1.1 million estimated values across the conterminous US. The data, upon publication of the companion journal article, will be also distributed as part of the StreamCat dataset.

This dataset is associated with the following publication: Doyle, J., R. Hill, S. Leibowitz, and J. Ebersole. Random forest models to estimate bankfull and low flow channel widths and depths across the conterminous United States. JOURNAL OF THE AMERICAN WATER RESOURCES ASSOCIATION. American Water Resources Association, Middleburg, VA, USA, 59(5): 1099-1114, (2023).