Introductory overview: Error metrics for hydrologic modelling - A review of common practices and an open source library to facilitate use and adoption

Error metrics quantify predicted flow accuracy and compare different predictions. Hydrologists commonly use and report metrics with little justification or discussion of the selected metric or metric strengths and weaknesses. Metric selection requires clear objectives, as different metrics are sensitive to different bias or error types. We review over 60 different error metrics along with various common modifications. We provide a brief metric description of these metrics and a more in-depth discussion of metrics often reported in hydrological literature. We recommend that multiple metrics be used to evaluate model accuracy. We present the open source HydroErr library, implemented in Python and MATLAB (R), which contains the error metric functions reported here to facilitate greater use of these metrics and encourage metric exploration related to relative metric strengths and weaknesses. We demonstrate the library with short case studies and provide a supplement with additional detail.