Per la serie di seminari in meteorologia ambientale dell’Università di Trento giovedì 7 marzo alle ore 14:30 Claudia Acquistapace, dell’Institute for Geophysics and Meteorology – University of Cologne, terrà un seminario dal titolo: Exploiting machine learning to improve our understanding of severe storms over the Alps: EXPATS project’s story and future.
Il seminario si terrà in presenza nell’aula 1P (primo piano) al DICAM – Università di Trento, in via Masiano 77, e sarà possibile seguirlo in streaming su Zoom al seguente link:
https://unitn.zoom.us/j/88611750340
(Meeting ID: 8861175 0340, Passcode: 615808)
Abstract Severe thunderstorms are crucial in the expected damages due to climate change and floods that the EU will face in the 2050s. These systems, especially in the Alps, produce large hail, heavy rain and cause landslides and flash floods. The project “Exploiting spatiotemporal cloud patterns to advance severe storms process understanding and detection (EXPATS, https://expats-ideas4s.com)” was recently funded by DWD within the IDEAS4S German-Italian cooperation to improve our understanding of extreme precipitating events over the Alps. EXPATS will exploit Meteosat second and third-generation satellite data (MSG, MTG) and ML self-supervised approaches to identify spatiotemporal cloud evolution patterns that can lead to extreme events and to characterize large hail spatiotemporal evolution across multi-year records. ML self-supervised embeddings showed exciting spatial cloud pattern identification developments from satellite images. In this talk, we will introduce the main research questions tackled by the EXPATS research group and introduce you to the tools we use in our research. We will show some of the challenges we face and some preliminary results and feasibility tests we performed to identify hail with the self-supervised machine learning approach. Bio Claudia’s path starts in Pisa, where she got her bachelor degree in general Physics, and then went on in Bologna, where she completed her master in physics focusing on atmosphere with a thesis on satellite remote sensing of light precipitation in Europe. Then, she moved to Cologne for completing a PhD within the Marie Curie initial training network focusing on drizzle detection but this time from the ground, using cloud radar Doppler spectra, for which she was awarded with the Reinhard-Süring-Stiftung 2019 Research award. She employed her first post-doc to use her observations to evaluate the performance of the ICON-LEM model. She was then actively taking part in the EUREC4A measurement campaign as PI on the research vessel Maria S. Merian, conducting radar ship-based observations of clouds and rain in the tropics. Currently she is completing a master in science communication at the University of Trento and she will soon start her new position as junior research group leader in Cologne, coming back to satellite remote sensing, this time for better understanding extreme precipitation events using machine learning methods. She is the science communication manager of the PROBE Cost Action and she also got funding for outreach projects, like the videodocumentary on women in science “Wetoo: what they don’t tell you [1]” and for the realization of an outreach video on atmospheric boundary layer [2], which was awarded by the AISAM association with a communication price.