The Multi-sensor Approach for Satellite Hail Advection (MASHA): a new technique for nowcasting applications

The Multi-sensor Approach for Satellite Hail Advection (MASHA) is a new satellite hybrid technique conceived for the real time detection and advection of hail clouds. MASHA is based on a machine learning algorithm able to identify hail clouds from satellite measurements and predict the evolution of hail-bearing systems every 5 min. The machine Learning techniques represent a valuable tool to address this problem. In particular, the use of deep learning model allows to automatically combine low level data and providing accurate predictions. Operationally, MASHA combines the strengths of the MWCC-H method to detect hail through the whole GPM constellation (Laviola et al., 2020a-b) with the high temporal rate of the Meteosat Rapid Scan Service (MSG-RSS). The novelty of this approach is offering the unprecedented possibility to advect hail-bearing systems in real-time and at very high spatial resolution. This opens the way to the operational applications of MASHA method by offering an unprecedented support to the nowcasting of hailstorms and to regional numerical weather predictions. Recent applications experimented the ingestion of lightning strikes and radar hail indices in order to improve the reconstruction of hail fields when the GPM-C overpasses are missing. The result is a near-real time, more consistent, high-resolution hail map.

References

Laviola S., V. Levizzani, R. R. Ferraro, and J. Beauchamp: Hailstorm Detection by Satellite Microwave Radiometers. Remote Sens. 2020a, 12(4), 621; https://doi.org/10.3390/rs12040621

Laviola S., G. Monte, V. Levizzani, R. R. Ferraro, and J. Beauchamp: A new method for hail detection from the GPM constellation. A prospective for a global hailstorm climatology. Remote Sens. 2020b, 12(21), 3553; https://doi.org/10.3390/rs12213553