Code and analysis-ready data for "Reservoir drought often persists beyond meteorological recovery worldwide"

作者：Alireza Daneshi, Hossein Azadi, Soraya Yaghobi, Anwar Eziz · 发表于：Zenodo (CERN European Organization for Nuclear Research) · 年份：2026 · DOI：10.5281/zenodo.21790941 · 研究领域：Computer science、Meteorology、Database、Data mining、Environmental science、Operations research、Information retrieval、Remote sensing

This archive contains analysis-ready event, feature, prediction and external-validation ledgers; unrounded source data for all main and Extended Data figures; executable analysis and verification code; environment records; final figures; numerical dictionaries; claim-boundary ledgers; and SHA-256 manifests supporting the manuscript “Reservoir drought often persists beyond meteorological recovery worldwide”. The locked analysis covers 1,203 reservoirs, 5,154 complete reservoir-drought events, 2,501 strict meteorological–reservoir matches and 2,152 positive recovery-debt events. The archive is designed to reproduce the reported analysis-ready results, numerical claims, source-data tables and figures. The archive does not recreate every drought index from the original provider records. Raw third-party datasets, including GRanD, GOODD, GeoDAR, GROWL and meteorological inputs, are not redistributed and must be obtained from their original providers under the applicable licences and terms. Code is licensed under the MIT License. Original data products and documentation created for this archive are licensed under the Creative Commons Attribution 4.0 International License.