This app animates a global catalogue of drought events. Every coloured patch on the map is one drought cluster — a region contiguous in both space and time where a hydroclimatic variable stayed below its local drought threshold. Press play and watch four and a half decades of droughts emerge, migrate and decay, day by day, on a global 0.5° grid.
Drought is defined here as a dry anomaly relative to local hydroclimatic conditions, not an absolute lack of water — which makes events comparable across climates. A grid cell is in drought when the variable falls below its locally varying 10th-percentile threshold (z-scores for precipitation, PET and runoff; relative deviations for soil moisture). Consecutive deficit days are grouped into events, and a three-dimensional DBSCAN clustering (latitude–longitude–time) merges neighbouring drought pixels into spatiotemporal clusters. Clusters smaller than 50 pixels (~125,000 km²) or shorter than 30 days (10 days for PET) are discarded.
The catalogue contains 4,315 precipitation clusters (1,123 of them ocean-only), 1,724 PET, 813 runoff and 715 soil-moisture clusters. Precipitation deficits are also tracked over the oceans, so events migrating from sea to land are not split at coastlines.
The dataset is openly available under CC-BY 4.0 on Zenodo and described in the data descriptor published in Scientific Data:
Šťovíček, V., Hanel, M., Kumar, R., Moravec, V., Markonis, Y., Cammalleri, C., Řehoř, J., Trnka, M. & Rakovec, O. A global dataset of spatiotemporal drought events from reanalysis and hydrological model data for 1980–2024. Scientific Data 13, 1319 (2026). doi:10.1038/s41597-026-07750-x
Built on the ERA5 reanalysis (Copernicus Climate Change Service / ECMWF) and the mHM mesoscale Hydrologic Model. Basemap by CARTO © OpenStreetMap contributors; rendering by MapLibre GL.
Dataset authors: Vít Šťovíček, Martin Hanel, Rohini Kumar, Vojtěch Moravec, Yannis Markonis, Carmelo Cammalleri, Jan Řehoř, Miroslav Trnka and Oldřich Rakovec. Viewer developed at CzechGlobe — Global Change Research Institute of the Czech Academy of Sciences, with the help of Claude Code.