Providing a physical basis to SPPT through the identification of weather features as coherent objects of potential vorticty anomalies: application to Mediterranean cyclogenesis

The Mediterranean Basin is a region characterized by intense cyclonic activity associated with events of extreme and adverse weather, often having high social and economic impact. The genesis and further development of these cyclones is strongly dependent to preceding Rossby wave breaking over the Atlantic Ocean. However, the accurate reproduction of such Rossby wave breaking is a challenging issue for numerical weather prediction (NWP) limiting the predictability of Mediterranean cyclones to lead times of typically no more than 5 days.

To better understand the sensitivity of forecast systems in predicting Mediterranean cyclogenesis and for further advancing with adequate ensemble approaches for NWP, we have developed a new scheme of stochastic parameter perturbation of physical tendencies (SPPT). The new SPPT scheme has been implemented into WRF and consists of random perturbations which are applied solely to coherent 3D objects of potential vorticity (PV) components that are produced by physical parametrisations. Given the invertibility trait of PV, these 3D objects are expected to significantly affect the forecasted atmospheric state with inherent uncertainties to physical parametrisations and consequent model errors.

A scenario tested for the needs of the study is the Mediterranean tropical-like cyclone (medicane) Zorbas that took place in September 2018; an event characterized by difficulties in adequately forecasting the cyclogenesis stage. Overall, our results quantify the sensitivity of NWP of Mediterranean cyclogenesis and highlight our scheme as an attractive alternative to classic approaches producing significant ensemble spread.