Catchment Attributes and MEteorology for Large-Sample SPATially distributed analysis (CAMELS-SPAT): streamflow observations, forcing data and geospatial data for hydrologic studies across North America

Abstract. We build on the existing Catchment Attributes and MEteorology for Large-sample Studies (CAMELS) dataset to present a new dataset aimed at hydrologic studies across North America, with a particular focus on facilitating spatially distributed studies. The dataset includes basin outlines, streamflow observations, meteorological data and geospatial data for 1426 basins in the US and Canada. To facilitate a wide variety of studies, we provide the basin outlines at a lumped and semi-distributed resolution; streamflow observations at daily and hourly time steps; variables suitable for running a wide range of models obtained and derived from different meteorological datasets at daily (one dataset) and hourly (three datasets) time steps; and geospatial data and derived attributes from 11 different datasets that broadly cover climatic conditions, vegetation properties, land use and subsurface characteristics. Forcing data are provided at their original gridded resolution, as well as averaged at the basin and sub-basin level. Geospatial data are provided as maps per basin, as well as summarized as catchment attributes at the basin and sub-basin level with various statistics. Attributes are further complemented with statistics derived from the forcing data and streamflow and focus on quantifying the variability of natural processes and catchment characteristics in space and time. Our goal with this dataset is to build upon existing large-sample datasets and provide the means for a more detailed investigation of hydrologic behaviour across large geographical scales. In particular, we hope that this dataset provides others with the data needed to implement a wide range of modelling approaches and to investigate the impact of basin heterogeneity on hydrologic behaviour and similarity. The CAMELS-SPAT (Catchment Attributes and MEteorology for Large-Sample SPATially distributed analysis) dataset is available at https://doi.org/10.20383/103.01306 (Knoben et al., 2025).