Random Forest Regression Model Archive for Estimating Low-Streamflow Statistics at Ungaged Locations in New York, excluding Long Island

This USGS data release contains R models (R Core Team, 2020) for estimating 7Q10 and 30Q10 [lowest annual 7-day and 30-day average streamflow that occurs (on average) once every 10 years] statistics at ungaged locations in or adjacent to New York State excluding Long Island. Details on model development are available in Stagnitta and others (2025).

basin_characteristics.csv - 224 basin characteristics represented by variables that describe basin geometry, climatic conditions, land cover, soils and surficial geology, and other characteristics for 213 unaltered gaged locations determined from (Stagnitta and others, 2024). Basin characteristics were used as predictor variables within the models to estimate low-streamflow statistics for ungaged locations.

R scripts used to flag unaltered streamgage datasets for redundant basins, and to train, tune, test, and bias-correct random forest regression models for estimating 7Q10 and 30Q10 are included in model.zip. Users are encouraged to read the readme file in this zipped file for details on the scripts and associated files used to generate the statistics.

References:

R Core Team, 2020, R—A language and environment for statistical computing: R Foundation for Statistical Computing web page, accessed August 2024 at

Stagnitta, T.J., Graziano, A., Woda, J., Glas, R., and Gazoorian, C., 2024, Low-flow statistics for New York State, excluding Long Island, computed through March 2022: U.S. Geological Survey data release, https://doi.org/10.5066/P9NOM6FR.