WaveC2R: Wavelet-Driven Coarse-to-Refined Hierarchical Learning for Radar Retrieval

2025-11-13

AI Summary

Existing satellite radar precipitation retrieval methods often rely on single-source data and oversimplified spatial modeling, limiting their ability to accurately reconstruct complex precipitation structures and resolve sharp meteorological boundaries—particularly in topographically complex regions with sparse ground-radar coverage. To address this, we propose a frequency-aware coarse-to-fine two-stage learning framework. First, wavelet decomposition disentangles low-frequency precipitation intensity from high-frequency boundary features. Second, a frequency-specific loss function is designed, and a conditional diffusion model is introduced to enable frequency-guided detail enhancement. The method jointly models spectral and spatial information by fusing multi-source remote sensing observations. Evaluated on the SEVIR dataset, our approach achieves state-of-the-art performance, significantly improving fidelity for heavy precipitation events and sharpening meteorological boundaries.

Technology Category

• Machine Learning: Multimodal Learning
• Computer Vision: Multi-modal Vision
• Natural Language Processing: Language Grounding & Multi-modal NLP

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking
• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs

Abstract

Satellite-based radar retrieval methods are widely employed to fill coverage gaps in ground-based radar systems, especially in remote areas affected by terrain blockage and limited detection range. Existing methods predominantly rely on overly simplistic spatial-domain architectures constructed from a single data source, limiting their ability to accurately capture complex precipitation patterns and sharply defined meteorological boundaries. To address these limitations, we propose WaveC2R, a novel wavelet-driven coarse-to-refined framework for radar retrieval. WaveC2R integrates complementary multi-source data and leverages frequency-domain decomposition to separately model low-frequency components for capturing precipitation patterns and high-frequency components for delineating sharply defined meteorological boundaries. Specifically, WaveC2R consists of two stages (i)Intensity-Boundary Decoupled Learning, which leverages wavelet decomposition and frequency-specific loss functions to separately optimize low-frequency intensity and high-frequency boundaries; and (ii)Detail-Enhanced Diffusion Refinement, which employs frequency-aware conditional priors and multi-source data to progressively enhance fine-scale precipitation structures while preserving coarse-scale meteorological consistency. Experimental results on the publicly available SEVIR dataset demonstrate that WaveC2R achieves state-of-the-art performance in satellite-based radar retrieval, particularly excelling at preserving high-intensity precipitation features and sharply defined meteorological boundaries.

Problem

• Filling coverage gaps in ground-based radar systems
• Capturing complex precipitation patterns accurately
• Delineating sharply defined meteorological boundaries

Innovation

• Wavelet decomposition separates frequency components
• Intensity-boundary decoupled learning optimizes separately
• Detail-enhanced diffusion refines precipitation structures