A Physically Driven Interpretable Machine Learning Framework for Early Forecasting of Summer Hypoxia in the Semi-Enclosed Bohai Sea Using Remote Sensing Data
文献类型：期刊论文
| 作者 | Jin, Yong3,4; Guo, Jie4; Liu, Shanwei3; Li, Tao3; Yue, Hansen2; Ji, Diansheng1; Hou, Chawei1; Tang, Haitian1 |
| 刊名 | REMOTE SENSING
 |
| 出版日期 | 2026-04-07 |
| 卷号 | 18期号:7页码:27 |
| 关键词 | hypoxia prediction remote sensing data Bohai Sea machine learning SHAP |
| DOI | 10.3390/rs18071097 |
| 通讯作者 | Guo, Jie(jguo@yic.ac.cn) |
| 英文摘要 | Highlights What are the main findings? An interpretable machine learning model integrating multi-source remote sensing and reanalysis data successfully forecasted the August hypoxia probability in the central Bohai Sea with a lead time of two months (using June predictors), achieving high accuracy (F1 = 0.76, AUC = 0.92) and capturing interannual variability under contrasting climate modes (e.g., El Ni & ntilde;o vs. La Ni & ntilde;a). SHAP analysis revealed that hypoxia is primarily controlled by physical suppression of vertical mixing (via low wind speeds and strong stratification), which restricts oxygen replenishment to the bottom layer. What are the implications of the main findings? The framework demonstrates that physical variables alone-derived entirely from satellite and reanalysis products-can effectively forecast coastal hypoxia, offering a practical, low-cost early-warning tool for data-scarce regions where in situ biogeochemical monitoring is limited. By linking remote sensing observations to mechanistic oceanographic processes through physics-informed features and explainable AI, this study provides a transferable paradigm for environmental risk prediction that balances predictive performance, physical plausibility, and operational transparency.Highlights What are the main findings? An interpretable machine learning model integrating multi-source remote sensing and reanalysis data successfully forecasted the August hypoxia probability in the central Bohai Sea with a lead time of two months (using June predictors), achieving high accuracy (F1 = 0.76, AUC = 0.92) and capturing interannual variability under contrasting climate modes (e.g., El Ni & ntilde;o vs. La Ni & ntilde;a). SHAP analysis revealed that hypoxia is primarily controlled by physical suppression of vertical mixing (via low wind speeds and strong stratification), which restricts oxygen replenishment to the bottom layer. What are the implications of the main findings? The framework demonstrates that physical variables alone-derived entirely from satellite and reanalysis products-can effectively forecast coastal hypoxia, offering a practical, low-cost early-warning tool for data-scarce regions where in situ biogeochemical monitoring is limited. By linking remote sensing observations to mechanistic oceanographic processes through physics-informed features and explainable AI, this study provides a transferable paradigm for environmental risk prediction that balances predictive performance, physical plausibility, and operational transparency.Abstract Hypoxia has become increasingly frequent in the semi-enclosed Bohai Sea since the early 2000s, posing significant risks to marine ecosystems. To address the limitations of existing dissolved oxygen models-particularly their poor predictive ability and lack of interpretability-we developed a two-month lead probabilistic forecasting framework for summer hypoxia using only multi-source remote sensing and reanalysis data, supplemented by in situ observations for validation. Environmental conditions in June were used to predict hypoxia probability in August via machine learning; among the seven algorithms tested, the optimized Random Forest model achieved the best performance (F1 = 0.76 and AUC = 0.92 on the independent test set). The model successfully reproduced observed hypoxia patterns in 2019 (validated against numerical simulations) and 2022 (validated against field measurements), capturing an increase in hypoxic area from 8229 km2 to 13,866 km2, which is consistent with intensifying thermal stratification under climate warming. SHAP-based interpretability analysis identified reduced wind speed and enhanced thermal stratification as the dominant physical drivers, highlighting the critical role of suppressed vertical mixing in limiting bottom-water oxygen supply. This study demonstrates that a physics-informed, interpretable machine learning approach based solely on satellite and reanalysis data can deliver reliable, early, and physically consistent hypoxia forecasts, offering a scalable solution for environmental monitoring of data-limited coastal seas. |
| WOS关键词 | GULF-OF-MEXICO ; DISSOLVED-OXYGEN ; COASTAL ; CONSEQUENCES ; ESTUARIES ; DYNAMICS ; EVENTS ; BOTTOM ; INLAND |
| WOS研究方向 | Environmental Sciences & Ecology ; Geology ; Remote Sensing ; Imaging Science & Photographic Technology |
| 语种 | 英语 |
| WOS记录号 | WOS:001738926600001 |
| 资助机构 | National Natural Science Foundation of China |
| 源URL | [http://ir.yic.ac.cn/handle/133337/42356]  |
| 专题 | 烟台海岸带研究所_海岸带信息集成与综合管理实验室 |
| 通讯作者 | Guo, Jie |
| 作者单位 | 1.Minist Nat Resources, Yantai Marine Ctr, Yantai 264006, Peoples R China 2.Chinese Acad Sci, Inst Oceanol, Qingdao 266000, Peoples R China 3.China Univ Petr East China, Coll Oceanog & Space Informat, Qingdao 266580, Peoples R China 4.Chinese Acad Sci, Yantai Inst Coastal Zone Res, Shandong Key Lab Coastal Zone Environm Proc & Ecol, Yantai 264003, Peoples R China |
| 推荐引用方式 GB/T 7714 | Jin, Yong,Guo, Jie,Liu, Shanwei,et al. A Physically Driven Interpretable Machine Learning Framework for Early Forecasting of Summer Hypoxia in the Semi-Enclosed Bohai Sea Using Remote Sensing Data[J]. REMOTE SENSING,2026,18(7):27. |
| APA | Jin, Yong.,Guo, Jie.,Liu, Shanwei.,Li, Tao.,Yue, Hansen.,...&Tang, Haitian.(2026).A Physically Driven Interpretable Machine Learning Framework for Early Forecasting of Summer Hypoxia in the Semi-Enclosed Bohai Sea Using Remote Sensing Data.REMOTE SENSING,18(7),27. |
| MLA | Jin, Yong,et al."A Physically Driven Interpretable Machine Learning Framework for Early Forecasting of Summer Hypoxia in the Semi-Enclosed Bohai Sea Using Remote Sensing Data".REMOTE SENSING 18.7(2026):27. |
入库方式： OAI收割
来源：烟台海岸带研究所
浏览0
下载0
收藏0
其他版本
除非特别说明，本系统中所有内容都受版权保护，并保留所有权利。