RainSeer: Fine-Grained Rainfall Reconstruction via Physics-Guided Modeling

2025-10-02

AI Summary

Existing spatial interpolation methods often over-smooth high-resolution rainfall fields, failing to preserve frontal structures and local extremes. To address this, we propose a physics-guided two-stage modeling framework: (1) leveraging radar reflectivity as a physically consistent structural prior, we design a bidirectional structure-to-point mapping mechanism to semantically translate aloft hydrometeor distributions into surface rainfall; (2) we integrate a geography-aware rainfall decoder with a causal spatiotemporal attention module to jointly fuse volumetric radar scans and sparse ground observations. Evaluated on the RAIN-F and MeteoNet datasets, our method reduces mean absolute error by 13.31% compared to state-of-the-art baselines, significantly improving spatial structural fidelity—particularly for sharp gradients and frontal systems—as well as the representation of localized extreme precipitation events.

Technology Category

• Machine Learning: Structured Learning
• Planning, Routing, and Scheduling: Optimization of Spatio-temporal Systems
• Knowledge Representation and Reasoning: Geometric, Spatial, and Temporal Reasoning

Application Category

• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• User Modeling, Personalization and Recommendation: Explainable and interpretable methods for personalization

Abstract

Reconstructing high-resolution rainfall fields is essential for flood forecasting, hydrological modeling, and climate analysis. However, existing spatial interpolation methods-whether based on automatic weather station (AWS) measurements or enhanced with satellite/radar observations often over-smooth critical structures, failing to capture sharp transitions and localized extremes. We introduce RainSeer, a structure-aware reconstruction framework that reinterprets radar reflectivity as a physically grounded structural prior-capturing when, where, and how rain develops. This shift, however, introduces two fundamental challenges: (i) translating high-resolution volumetric radar fields into sparse point-wise rainfall observations, and (ii) bridging the physical disconnect between aloft hydro-meteors and ground-level precipitation. RainSeer addresses these through a physics-informed two-stage architecture: a Structure-to-Point Mapper performs spatial alignment by projecting mesoscale radar structures into localized ground-level rainfall, through a bidirectional mapping, and a Geo-Aware Rain Decoder captures the semantic transformation of hydro-meteors through descent, melting, and evaporation via a causal spatiotemporal attention mechanism. We evaluate RainSeer on two public datasets-RAIN-F (Korea, 2017-2019) and MeteoNet (France, 2016-2018)-and observe consistent improvements over state-of-the-art baselines, reducing MAE by over 13.31% and significantly enhancing structural fidelity in reconstructed rainfall fields.

Problem

• Reconstructing high-resolution rainfall fields from sparse observations
• Capturing sharp transitions and localized extremes in rainfall data
• Bridging physical disconnect between radar reflectivity and ground precipitation

Innovation

• Physics-guided two-stage architecture for rainfall reconstruction
• Structure-to-Point Mapper enables spatial alignment
• Geo-Aware Decoder captures hydro-meteor transformations