Seeking for a Trade-Off Between Accuracy and Timeliness in Meteo-Hydrological Modeling Chains

摘要

The level of detail achieved by operational General Circulation Models (e.g., the HRES 9km resolution forecast recently launched by the ECMWF) raises questions about the most appropriate use of Limited Area Models, which provide for further dynamical downscaling of the weather variables. The two main objectives targeted in hydro-meteorological forecasts, i.e. accuracy and timeliness, are to some extent conflicting. Accuracy and precision of a forecast can be evaluated by proper statistical indices based on observations, while timeliness mainly depends on the spatial resolution of the grid and the computational resources used. In this research, several experiments are set up applying the Advanced Research Weather Research and Forecasting (WRF-ARW) Model to a weather event occurred in Southern Italy in 2018. Forecast accuracy is evaluated both for the HRES ECMWF output and that provided by WRF dynamical downscaling at different resolutions. Furthermore, timeliness of the forecast is assessed adding to the time needed for GCM output availability the time needed for Limited Area simulations at different resolutions and using varying core numbers. The research provides useful insights for the operational forecast in the study area, highlighting the level of detail required and the current weaknesses hindering correct forecast of the hydrological impact of extreme weather events.