Extreme meteo-marine events in the Mediterranean: numerical modeling approaches for early-warning

The presented work analyses the results of an innovative numerical simulation system over the Mediterranean Sea basin developed in the context of the PROMETO project, aimed at producing early-warning indicators for coastal protection and navigation safety. 

The numerical system is exploited to simulate the evolution of extreme weather and sea events in the Mediterranean Sea over the decade 2010-2020, using both the downscaling of the global ERA5 model by means of the WRF (Weather Research and Forecasting) model at 5 km horizontal resolution and the reforecast with the same model at 3 km resolution (with sea surface temperature updates every 6 hours). Resulting atmospheric fields are used to  force the SHYFEM hydrodynamic model coupled to the WAVEWATCHIII wave model at very high resolution and to produce the relevant environmental variables for early-warning indicators over the entire Mediterranean basin.

To test the system in an operational early-warning context, using the Ensemble reforecast approach based on 50 ECMWF members, two case studies of extreme weather-sea events are simulated, namely the high-impact storms 'Vaia' (2018) and 'Detlef' (2019). For each event, we can evaluate the model uncertainty of wind speed, rain and ocean-wave fields and we can assess the impact of the uncertainty provided by the ensemble approach on the predictions of wind, waves, sea level and derived early-warning indicators. In addition, a ‘member selection’ technique is used in order to assess how the selection of a few, potentially, more significant ensemble members is impactful in statistical and forecasting terms, allowing also to reduce the computational load of high-resolution ensemble meteo-marine forecasts at regional scale.