Composite flood generation driven by a warm atmospheric River in a southern Andean basin, Chile

Study region Ñuble Alto basin, Chile. Study focus Analyze and characterize the interaction among meteorological variables, hydrological state variables, and hydrological flux variables before, during, and after the landfall of a Zonal Atmospheric River (ZAR), that caused flooding events between June 22 and 26, 2023, in Ñuble Alto basin in south-central Chile (71.25° W, 36.68° S). We analyzed and characterized the simulated hydrological state and flux variables using a physically-based distributed hydrological model. To improve the initial conditions (ICs), we enhanced the spatiotemporal characterization of precipitation by using a Random Forest model to merge observed data and precipitation products. To increase the reliability of our results, we performed a sensitivity analysis on the input data, ICs, and parameters. In addition, we quantified the uncertainty in the ICs. Finally, we developed a workflow for forecasting streamflow up to 10 days in advance. New Hydrological Insights for the Region The results showed that the event was dominated mainly by liquid precipitation (snowmelt contribution = 0.03), with 75% of the precipitation on the day of the peak event converted to runoff and a soil moisture ratio of 0.97, confirming that extreme liquid precipitation and highly saturated soils jointly controlled the magnitude of the June 2023 floods. In addition, our forecast was able to predict peak streamflow five days in advance.