用多模态数据提升印度季风降雨1公里精度预测
Learning Regional Monsoon Patterns with a Multimodal Attention U-Net
- 融合七类地理数据，用注意力U-Net捕捉区域降雨模式
- 在极端降雨类别上表现显著优于单模态模型
- 适合气候韧性、农业规划与地理空间AI研究者参考
准确的季风降雨预测对印度农业、水资源管理和气候风险规划至关重要，但受限于稀疏地面观测和复杂的区域变异。本文提出一种多模态深度学习框架，用于高分辨率降水分类，整合卫星与地球观测数据。不同于以往基于5-50公里网格的模型，我们构建了覆盖印度五个邦、1公里分辨率的新数据集，包含七类关键地理模态：地表温度、植被（NDVI）、土壤湿度、相对湿度、风速、高程和土地利用，时间范围为2024年6月至9月季风期。方法采用注意力引导的U-Net架构，捕捉多模态间的时空依赖关系，并结合焦点损失和Dice损失函数处理印度气象局（IMD）定义的降雨类别不平衡问题。实验表明，该多模态框架持续优于单模态基线及现有深度学习方法，尤其在极端降雨类别中表现突出。本工作贡献了一个可扩展框架、基准数据集及区域季风预报的最先进结果，适用于气候韧性与地理空间人工智能应用。
原文摘要 · Abstract (English)
Accurate monsoon rainfall prediction is vital for India's agriculture, water management, and climate risk planning, yet remains challenging due to sparse ground observations and complex regional variability. We present a multimodal deep learning framework for high-resolution precipitation classification that leverages satellite and Earth observation data. Unlike previous rainfall prediction models based on coarse 5-50 km grids, we curate a new 1 km resolution dataset for five Indian states, integrating seven key geospatial modalities: land surface temperature, vegetation (NDVI), soil moisture, relative humidity, wind speed, elevation, and land use, covering the June-September 2024 monsoon season. Our approach uses an attention-guided U-Net architecture to capture spatial patterns and temporal dependencies across modalities, combined with focal and dice loss functions to handle rainfall class imbalance defined by the India Meteorological Department (IMD). Experiments demonstrate that our multimodal framework consistently outperforms unimodal baselines and existing deep learning methods, especially in extreme rainfall categories. This work contributes a scalable framework, benchmark dataset, and state-of-the-art results for regional monsoon forecasting, climate resilience, and geospatial AI applications in India.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。