Using machine learning to downscale coarse-resolution environmental variables for understanding the spatial frequency of convective storms

Hungjui Yu, Lander Ver Hoef, Kristen L. Rasmussen, Imme Ebert-Uphoff

Artificial Intelligence for the Earth Systems (2026) · 10 September 2025 · Colorado State University (incl. CIRA)

In one sentence. Pixel-based neural networks map coarse environmental variables to convection-permitting storm frequency with SSIM above 0.8, capturing diurnal and orographic patterns.

Abstract

Global climate models (GCMs), typically run at ~100-km resolution, capture large-scale environmental conditions but cannot resolve convection and cloud processes at kilometer scales. Convection-permitting models offer higher-resolution simulations that explicitly simulate convection but are computationally expensive and impractical for large ensemble runs. This study explores machine learning (ML) as a bridge between these approaches. We train simple, pixel-based neural networks to predict convective storm frequency from environmental variables produced by a regional convection-permitting model. The ML models achieve promising results, with structural similarity index measure (SSIM) values exceeding 0.8, capturing the diurnal cycle and orographic convection without explicit temporal or spatial coordinates as input. Model performance declines when fewer input features are used or specific regions are excluded, underscoring the role of diverse physical mechanisms in convective activity. These findings highlight ML potential as a computationally efficient tool for representing convection and as a means of scientific discovery, offering insights into convective processes. Unlike convolutional neural networks, which depend on spatial structure and grid size, the pixel-based model treats each grid point independently, enabling value-to-value prediction without spatial context. This design enhances adaptability to resolution changes and supports generalization to unseen environmental regimes, making it particularly suited for linking environmental conditions to convective features and for application across diverse model grids or climate scenarios.

Labels

Labels and summary are generated by Claude and reviewed weekly. Wrong? Edit this entry.