I have been honored to lead a community-based article published in BAMS (the Bulletin of the American Meteorological Society) presenting the next decade's vision to enhance research-to-operations (R2O) hydrological predictions across scales and time horizons.
Artificial Intelligence is rapidly reshaping climate science—bringing powerful tools for forecasting floods, droughts, and heatwaves, but also raising important concerns about trust, transparency, and scientific validity.
In our recent publication in Nature Communications Earth & Environment we show how combining process-based models with Machine Learning and statistical post-processing significantly boosts the accuracy of local streamflow predictions across Europe—even from large-scale models like SMHI's E-HYPE.