Data-driven Seasonal Hydrologic Prediction Using Earth Observing Satellites

利用地球观测卫星进行数据驱动的季节性水文预测

• 批准号: 18KK0117
• 负责人: 金 炯俊
• 金额: $ 11.4万
• 依托单位: The University of Tokyo
• 依托单位国家: 日本
• 项目类别: Fund for the Promotion of Joint International Research (Fostering Joint International Research (B))
• 财政年份: 2018
• 资助国家: 日本
• 起止时间: 2018-10-09 至 2024-03-31
• 项目状态: 已结题
• 来源: https://kaken.nii.ac.jp/en/grant/KAKENHI-PROJECT-18KK0117/
• 关键词: GPM; Multi-task learning; Precipitation; lake surface area; remote sensing; water big data; データ駆動型モデリング; 水文季節予報; 衛星観測; テレコネクション; 人工知能; 衛星高度計; 河川の水位; 海水面温度; 陸域水貯留; ニューラルネットワーク; 長期リードタイム予測; data-driven modeling; seasonal prediction; satel

Timely prediction of flood and drought greatly minimize the related losses. It requires precise precipitation estimates as a predictor of terrestrial processes across spatiotemporal scales. In this fiscal year, we proposed a novel precipitation retrieval framework in which regression and classification tasks are simultaneously trained using multi-task learning approach. Satellite-based precipitation estimations provide frequent, large-scale measurements. Recently, deep learning has shown significant potential for improving estimation accuracy. In this project, we designed a novel network architecture and loss function to maximize the benefits of multi-task learning. The proposed multi-task (i.e., two-task) model successfully achieved a better performance than the conventional single-task model possibly due to efficient knowledge transfer between tasks. The product intercomparison showed that our product outperformed existing products in rain rate retrieval and also yielded better skills in the rain/no-rain retrieval task.

及时对洪水和干旱的预测极大地减少了相关损失。它需要精确的降水估计作为跨时空尺度的陆生过程的预测指标。在这个财政年度，我们提出了一个新颖的降水检索框架，其中使用多任务学习方法同时训练回归和分类任务。基于卫星的降水估计提供了频繁的大规模测量。最近，深度学习显示了提高估计准确性的巨大潜力。在这个项目中，我们设计了一种新颖的网络体系结构和损失功能，以最大程度地提高多任务学习的好处。提议的多任务（即两任任务）模型成功地取得了比传统的单任务模型更好的性能，这可能是由于任务之间有效的知识传输所致。该产品比较表明，我们的产品在降雨率检索中的表现优于现有产品，并且在雨水/无雨的检索任务中也获得了更好的技能。

项目成果

Sensitivity of Global Hydrological Simulations to Groundwater Capillary Flux Parameterizations

• DOI: 10.1029/2018wr023434
• 发表时间: 2019-01
• 期刊: Water Resources Research
• 影响因子: 5.4
• 作者: Sujan Koirala;Hyungjun Kim;Y. Hirabayashi;S. Kanae;T. Oki

南米大陸における水ストレス下の陸域生態系の光合成動態

南美洲水分胁迫下陆地生态系统的光合动态

• 发表时间: 2019
• 作者: 藤森 慎太郎;Kim Hyungjun
• 通讯作者: Kim Hyungjun

Observed changes in dry-season water availability attributed to human-induced climate change

• DOI: 10.1038/s41561-020-0594-1
• 发表时间: 2020-06
• 期刊: Nature Geoscience
• 影响因子: 18.3
• 作者: Ryan S. Padrón;L. Gudmundsson;B. Decharme;A. Ducharne;D. Lawrence;J. Mao;D. Peano;G. Krinner

Recurrent pattern of extreme fire weather in California

加州极端火灾天气反复出现

• DOI: 10.1088/1748-9326/ac1f44
• 发表时间: 2021
• 期刊: Environmental Research Letters
• 影响因子: 6.7
• 作者: Son, Rackhun;Wang, S-Y Simon;Kim, Seung Hee;Kim, Hyungjun;Jeong, Jee-Hoon;Yoon, Jin-Ho
• 通讯作者: Yoon, Jin-Ho

0.5° more warming changes compound extreme risks

变暖变化增加 0.5° 会加剧极端风险