On Global Applicability and Location Transferability of Generative Deep Learning Models for Precipitation Downscaling

2025-12-01

AI Summary

Generative deep learning models exhibit limited global generalizability in precipitation downscaling, particularly across unseen geographical regions. Method: We propose a geography-aware hierarchical data partitioning strategy to systematically evaluate cross-regional transferability across 15 geographically distinct, held-out areas. Leveraging both generative adversarial networks (GANs) and diffusion models, we integrate ERA5 reanalysis data with IMERG high-resolution precipitation observations to achieve global 0.1° precipitation downscaling. Contribution/Results: This work establishes the first benchmark framework for cross-regional transferability in climate downscaling. Empirical evaluation reveals that while most regions retain strong performance, significant degradation occurs in areas with pronounced topographic or climatic divergence—highlighting fundamental limits of generative model generalization in climate applications. The study provides a reproducible evaluation paradigm and a concrete technical pathway toward developing universally applicable climate downscaling models.

Technology Category

• Machine Learning: Deep Generative Models & Autoencoders
• Computer Vision: Generative Adversarial Networks (GANs) for Vision
• Natural Language Processing: Generation

Application Category

• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• Search and Retrieval-Augmented AI: Vertical and domain-specific search
• Economics, Online Markets and Human Computation: Economic ramifications for generative AI infrastructure and applications

Abstract

Deep learning offers promising capabilities for the statistical downscaling of climate and weather forecasts, with generative approaches showing particular success in capturing fine-scale precipitation patterns. However, most existing models are region-specific, and their ability to generalize to unseen geographic areas remains largely unexplored. In this study, we evaluate the generalization performance of generative downscaling models across diverse regions. Using a global framework, we employ ERA5 reanalysis data as predictors and IMERG precipitation estimates at $0.1^circ$ resolution as targets. A hierarchical location-based data split enables a systematic assessment of model performance across 15 regions around the world.

Problem

• Evaluates generalization of generative precipitation downscaling models globally
• Assesses model transferability across unseen geographic regions systematically
• Tests hierarchical location-based performance using ERA5 and IMERG data

Innovation

• Global framework for precipitation downscaling
• Hierarchical location-based data split
• Generative deep learning models evaluation