A Deep Learning Model To Predict Rapid Intensification Of Cyclones – Asian Scientist Magazine

AsianScientist reports on the challenges of predicting rapid intensification (RI) periods in tropical cyclones due to their rarity and severe risks. A new model, RITCF-contrastive, developed by researchers from Institute of Oceanology Chinese Academy of Sciences, uses contrastive learning to improve RI TC forecasting accuracy by analyzing atmospheric and oceanic conditions alongside satellite imagery. Testing with operational data showed a 92.3% prediction rate for RI periods, indicating potential for real-time forecasting to enhance early warning systems and disaster response.

Full Article