Thunderstorms are among the most devastating weather phenomena, mostly because of hail, heavy rainfall and strong winds. Although significant progress in convective scale numerical weather prediction models have been done these last years, the forecast of thunderstorms from a few hours to a few days remains difficult. It is partly because of the low predictability of these small scale phenomena but also due to deficiencies in the models, in particular in their microphysical schemes that describe the evolution of hydrometeor concentrations and sizes. Dual-polarization radar observations are very useful for microphysical schemes evaluation (Ryzhkov et al., 2020) as they provide information related to the size, shape, composition and orientation of the hydrometeors.
Recent studies (Kumjian and Ryzhkov, 2008; Kumjian et al., 2014; Johnson et al., 2016; Snyder et al., 2017a, b) have shown that supercell thunderstorms are associated with recurrent distinctive polarimetric signatures, such as the ZDR columns that consist of quasi-vertical continuous columns of enhanced ZDR extending above the environmental 0°C level and provide information about the location and intensity of a storm’s updraft (Snyder et al., 2015).
This work aims to statistically evaluate the ability of the AROME convective scale model (Brousseau et al., 2016) to replicate this signature when associated with different microphysics schemes : ICE3 (Pinty and Jabouille, 1998, one-moment) or LIMA (Vié, 2016; two-moment for rain, cloud water and ice cristals) which both can explicitely predict hail.
To do so, an automated detection of ZDR columns will be first implemented following Snyder et al. (2015) and Starzec et al. (2017). The algorithm will be applied on a few convective cases that occurred in France in 2022. Simulations with the AROME model will be performed on the same cases with different microphysics options. Simulated ZDR columns will be obtained thanks to the Augros et al. (2016) polarimetric radar forward operator applied to the AROME simulations. Finally, the characteristics (width, height) of the observed and simulated ZDR columns will be compared to assess which microphysics scheme is able to best reproduce the observed signatures. Besides evaluating the model microphysics, this study will enable to better understand the discrepancies between simulated and observed ZDR columns, a necessary first step towards using theses signatures in an assimilation context to improve storm-scale analysis and forecast.
