arXiv subjects

Seok-Woo Son

Publications and source records attributed to Seok-Woo Son.

2 recordsLinked to original sources

arXiv2026

Did recent sea surface temperature warming reinforce the extreme East Asian summer monsoon precipitation in 2020?

We analyzed the possible effects of recent sea surface temperature (SST) warming on the extraordinary East Asian summer monsoon (EASM) precipitation in 2020 summer. The dynamic and thermodynamic impacts of SST are examined by conducting regional climate model experiments with observed SST and cold SST where the 22-year SST trend is removed. In the presence of warm SST, precipitation increases in low latitudes but decreases in the EASM region. This dipolar precipitation change pattern opposes the precipitation anomalies in 2020 summer, indicating that the extraordinary 2020 EASM precipitation is not likely driven by recent SST warming. The warm SST suppresses the western North Pacific subtropical high expansion and weakens the southwesterly from the South China Sea toward the EASM region. In terms of large-scale atmospheric circulations, SST-induced wind changes strengthen the local Walker circulation in the South China Sea and the Philippines and the local Hadley circulation across the EASM region. These support the reduced EASM rainfall in the control experiment compared to the cold SST experiment and imply that the precipitation reduction by dynamical effects could exceed the precipitation increase by thermodynamic effects in the EASM region under warm SST.

physics.ao-ph↗

arXiv2022

Effective Training Strategies for Deep-learning-based Precipitation Nowcasting and Estimation

Deep learning has been successfully applied to precipitation nowcasting. In this work, we propose a pre-training scheme and a new loss function for improving deep-learning-based nowcasting. First, we adapt U-Net, a widely-used deep-learning model, for the two problems of interest here: precipitation nowcasting and precipitation estimation from radar images. We formulate the former as a classification problem with three precipitation intervals and the latter as a regression problem. For these tasks, we propose to pre-train the model to predict radar images in the near future without requiring ground-truth precipitation, and we also propose the use of a new loss function for fine-tuning to mitigate the class imbalance problem. We demonstrate the effectiveness of our approach using radar images and precipitation datasets collected from South Korea over seven years. It is highlighted that our pre-training scheme and new loss function improve the critical success index (CSI) of nowcasting of heavy rainfall (at least 10 mm/hr) by up to 95.7% and 43.6%, respectively, at a 5-hr lead time. We also demonstrate that our approach reduces the precipitation estimation error by up to 10.7%, compared to the conventional approach, for light rainfall (between 1 and 10 mm/hr). Lastly, we report the sensitivity of our approach to different resolutions and a detailed analysis of four cases of heavy rainfall.

cs.CV↗