Calculated baseflow recession characteristics for streamflow gauging locations for the western and eastern United States, 1900 to 2018

This metadata record describes observed and predicted baseflow recession characteristics for 300 streamflow gauges in the western United States and 282 streamflow gauges in the eastern United States. Specifically, this record describes (1) the streamflow gauge locations (west or east) in the United States (Location), (2) the U.S. Geological Survey streamflow gauge identification numbers (USGS_Site_Identifier), (3) observed regions of similar aquifer hydraulic properties (7 regions coded by color: blue, green, red, purple, grey, pink, and orange) by k-means clustering method (Observed_Class(k-means)), (4) predicted regions of similar aquifer hydraulic properties by random forest classification models (Predicted_Class(k-means)), (5) calculated long-term baseflow recession constant at streamflow gauges (Observed_a-long[ft^(-3/2)s^(-1/2)]), (6) predicted long-term baseflow recession constant by novel empirical and physical approach (Predicted_a-long(Novel)[ft^(-3/2)s^(-1/2)]), (7) predicted long-term baseflow recession constant by random forest regression (Predicted_a-long(Random_Forest_Regression)[ft^(-3/2)s^(-1/2)]), (8) calculated short-term baseflow recession constant at streamflow gauges (Observed_a-short[sft^(-6)]), (9) predicted short-term baseflow recession constant by novel empirical and physical approach (Predicted_a-short(Novel)[sft^(-6)]), (10) predicted short-term baseflow recession constant by random forest regression (Predicted_a-short(Random_Forest_Regression)[sft^(-6)]). For more details for (3) to (10), please see Eng, K., Wolock, D. M., and Wieczorek, M., 2023, Predicting baseflow recession characteristics at ungauged locations using a physical and machine learning approach. The values entered for (5) to (10) are in scientific notation, and they are character strings that will require the user to convert numeric values using methods for their software or use case. The data are in a tab-delimited text format.