CMCC Webinar
8 February 2024, 11:30 CET
To join the webinar, register here

Speaker:
Gabriele Accarino, CMCC – ASC Division

Moderator:
Enrico Scoccimarro, CMCC – CSP Division

Abstract
Tropical Cyclones (TCs) rank among the most formidable natural phenomena, posing substantial threats due to their destructive potential. The frequency of TCs is on the rise globally, with an annual average of 90 occurrences over tropical waters, exacerbated by the intensifying effects of global warming. Precise localization and tracking of these events have become crucial in weather and climate science. The webinar introduces a novel Machine Learning (ML) ensemble approach for locating TC center coordinates. The ensemble integrates TC center estimates from various ML models in agreement regarding the presence of a TC in the input data. Moreover, starting from the individual TC center locations detected by the ML ensemble approach, a deterministic tracking algorithm was used for reconstructing TC trajectories, leading to a hybrid approach. The talk summarizes the experimental setup, the main outcomes in terms of localization and tracking, as well as the comparison of the approach against four deterministic trackers reported in literature.

WATCH THE VIDEO