MSWEP4Caravan: a rainfall–runoff dataset for large-sample hydrology

The MSWEP4Caravan is a large-sample rainfall–runoff dataset that extends the Caravan by providing basin-scale precipitation forcing derived from the Multi-Source Weighted-Ensemble Precipitation (MSWEP) product. In addition, basin-scale 2m air temperature derived from the Multi-Source Weather (MSWX) product is provided as a companion forcing, and a set of climate attributes derived from these forcing inputs is included. The dataset compiles 23,668 basins from the Caravan, its extensions, and related CAMELS datasets. The gridded MSWEP precipitation and MSWX 2m air temperature are conservatively aggregated over basin polygons using area-weighted averaging and converted to daily basin-scale time series. The resulting time series span from 1979-01-02 to 2024-09-30. The MSWEP4Caravan supports large-sample rainfall–runoff modelling by providing Caravan-compatible MSWEP precipitation forcing. When used together with the Caravan, it supports comparisons of hydrological modelling results obtained with the MSWEP precipitation and with the existing Caravan precipitation. This record contains the released dataset files, documentation, and metadata for version 1.0.