Extreme precipitation modeling and Probable Maximum Precipitation (PMP) estimation in Chile

Study region: Maipo River Basin (MRB) in Chile Study focus: Aging infrastructures and climate-change emphasize an urgent need to improve extreme precipitation modeling and ultimately update Probable Maximum Precipitation (PMP) estimates. Motivated by this background, this study first examined the optimal physics parameterizations for modeling extreme precipitation. We then estimated PMP using the model-based method, which is compared with traditional methods-based PMP estimates. Finally, the exceedance probability of the estimated PMP was evaluated based on precipitation amount and its atmospheric drivers. New hydrological insights: Among the 48 patterns of physical parameterization combinations in the Weather Research and Forecasting (WRF) model, the combination of Stony-Brook University microphysics, BouLac PBL, and Grell-Freitas cumulus schemes resulted in the lowest RMSE of 72-hr MRB-average precipitation. We used the WRF with optimized parameterizations to estimate PMP, employing the ensemble runs of moisture amplification and storm transposition. The 72-hr MRB-average PMP was estimated to be 323.7 mm in the model-based method, whereas it was 348.3 mm in the moisture maximization method. The statistical method computed the station-average 72-hr PMP as 515.8 mm (standard deviation: 147.4 mm) within the MRB. Our analysis focusing on physical mechanisms showed that the integrated water vapor trasnport (IVT) change strongly drove the precipitation change on a basin scale and on hourly and event timescales for the target events. The IVT magnitude driving the PMP scenario was found to fall in return periods of similar to 44.5 years, indicating that the estimated PMP was driven by moisture flux within the range of historical observations, suggesting its physical credibility for practical applications.