Global prediction of extreme floods in ungauged watersheds

• 成果类型: Article
• 署名作者: Nearing, Grey; Cohen, Deborah; Dube, Vusumuzi; Gauch, Martin; Gilon, Oren; Harrigan, Shaun; Hassidim, Avinatan; Klotz, Daniel; Kratzert, Frederik; Metzger, Asher; Nevo, Sella; Pappenberger, Florian; Prudhomme, Christel; Shalev, Guy; Shenzis, Shlomo; Tekalign, Tadele Yednkachw; Weitzner, Dana; Matias, Yossi
• 署名单位: Alphabet Inc.; Google Incorporated; European Centre for Medium-Range Weather Forecasts (ECMWF); Helmholtz Association; Helmholtz Center for Environmental Research (UFZ); RAND Corporation
• 刊物名称: Nature
• ISSN/ISSBN: 0028-6755
• DOI: 10.1038/s41586-024-07145-1
• 发表日期: 2024-03-21
• 关键词: negative density-dependence; janzen-connell hypothesis; latitudinal gradient; natural enemies; plant diversity; mortality; survival; distance; coexistence; pathogens

摘要

Floods are one of the most common natural disasters, with a disproportionate impact in developing countries that often lack dense streamflow gauge networks 1 . Accurate and timely warnings are critical for mitigating flood risks 2 , but hydrological simulation models typically must be calibrated to long data records in each watershed. Here we show that artificial intelligence-based forecasting achieves reliability in predicting extreme riverine events in ungauged watersheds at up to a five-day lead time that is similar to or better than the reliability of nowcasts (zero-day lead time) from a current state-of-the-art global modelling system (the Copernicus Emergency Management Service Global Flood Awareness System). In addition, we achieve accuracies over five-year return period events that are similar to or better than current accuracies over one-year return period events. This means that artificial intelligence can provide flood warnings earlier and over larger and more impactful events in ungauged basins. The model developed here was incorporated into an operational early warning system that produces publicly available (free and open) forecasts in real time in over 80 countries. This work highlights a need for increasing the availability of hydrological data to continue to improve global access to reliable flood warnings. Artificial intelligence-based forecasting improves the reliability of predicting extreme flood events in ungauged watersheds, with predictions at five days lead time that are as good as current systems are for same-day predictions.