Global DFs V2: Design flood estimation for global river networks based on machine learning models

This dataset provides design flood estimates for near-global coverage using machine learning models. More details about the estimation method are described in the HESS paper (https://doi.org/10.5194/hess-25-5981-2021). We made some improvements in this version. a) We use river networks from HydroSHEDSdatasets as benchmarks. b) We use another open access CN dataset in model development. (https://doi.org/10.6084/m9.figshare.7756202.v1). c) We estimate design floods under multiple return periods (2-1000 years). The flood discharge Q is presented in log10(Q) value in each tiff file and the unit is m³/s. Please let me know ([e-mail]) if you have any questions about this dataset. Thanks.