Search papers, labs, and topics across Lattice.
MAGPIE-Net introduces a novel approach to predicting short-duration heavy rainfall events by integrating multitemporal observations from the FY-4A AGRI satellite with a geographically adaptive grid-to-station mapping. This method directly supervises the satellite-to-station learning pathway, enabling more accurate local rainfall warnings compared to traditional gridded precipitation models. In tests over central and eastern China, MAGPIE-Net achieved a critical success index of 0.371 at the 0-1 hour mark, significantly outperforming the best baseline by detecting 65.1% of events with a lead time of 64.6 minutes.
MAGPIE-Net outperforms traditional models by directly linking satellite data to localized rainfall predictions, achieving a 65.1% detection rate for heavy rainfall events.
Short-duration heavy-rainfall warning determines whether 1 h rainfall will exceed a threshold within a target-station neighborhood over the next few hours. Multitemporal infrared and water-vapor observations from the Fengyun-4A Advanced Geostationary Radiation Imager (FY-4A AGRI) capture cloud-top cooling, moisture evolution, and cloud expansion before substantial surface rainfall develops. However, most deep-learning nowcasting methods convert these signals into local warnings by post-processing gridded precipitation predictions, preventing station-neighborhood event targets from directly supervising the satellite-to-station learning pathway. We propose MAGPIE-Net, which embeds a geographically adaptive, differentiable grid-to-station mapping in a pathway combining convection-initiation features, multiscale encoding, and auxiliary gridded precipitation diagnosis. Station-neighborhood event losses thereby constrain the satellite representation and its mapping to irregular station locations for 0-3 h event prediction. In independent 2023 warm-season tests over central and eastern China, critical success index (CSI) values under the primary 40 km/20 mm h-1 definition were 0.371, 0.304, and 0.238 at 0-1, 1-2, and 2-3 h. Across episodes, MAGPIE-Net achieved a detection rate of 65.1% and a mean lead time of 64.6 min, compared with 23.6% and 18.3 min for the best gridded-output baseline, and remained superior for smaller neighborhoods and the 50 mm h-1 threshold. During the critical early-warning stage, when antecedent 1 h rainfall within 40 km remained below 1 mm, MAGPIE-Net detected 51.9% of episodes with a mean lead time of 38.5 min. These results show that event-oriented satellite-to-station modeling converts multitemporal geostationary cloud and moisture observations into local heavy-rainfall warnings more effectively than gridded-precipitation modeling.
