Earth observation embeddings are effective sub-grid descriptors for probabilistic weather downscaling

2026-08-12

AI Summary

This study addresses the challenge that coarse-resolution reanalysis data inadequately represent local near-surface meteorology due to unresolved terrain and surface characteristics. To overcome this limitation, the authors propose a probabilistic downscaling framework that leverages high-resolution land surface embeddings derived from the TESSERA foundation model as subgrid descriptors, integrated with convolutional conditional neural processes to refine coarse-grid datasets such as ERA5. They demonstrate for the first time that long-term surface embeddings significantly enhance short-term weather downscaling performance by systematically correcting subgrid biases through persistent surface attributes. The approach consistently improves both point and probabilistic predictions of 2-meter temperature and 10-meter wind speed across five climate zones, reducing the Continuous Ranked Probability Score (CRPS) by 11.5% and 6.2%, respectively, with robust gains maintained at unseen stations and under varying input data sources.

Abstract

Global weather reanalyses and forecasts resolve the evolving atmospheric state on coarse grids, but site-specific applications require predictions at arbitrary locations where near-surface conditions also depend on unresolved terrain and land-surface properties. Existing probabilistic downscalers address this gap using hand-crafted topographic descriptors. We ask instead whether Earth observation foundation models can provide transferable sub-grid surface representations for probabilistic weather downscaling. We augment a convolutional conditional neural process that downscales coarse ERA5 reanalysis fields at ~25 km resolution with a learned local surface descriptor, obtained by compressing a patch of TESSERA embeddings at 10 m resolution. Although these embeddings summarise surface conditions over annual timescales, they improve downscaling of instantaneous 2 m temperature and 10 m wind speed by encoding persistent surface properties that capture a location's departure from the coarse-grid atmospheric state. Across five climatically diverse regions, the embedding improves point and probabilistic skill at stations held out in both space and time, overall improving CRPS skill by 11.5% for 2 m temperature and 6.2% for 10 m wind speed. We further analyse how its contribution differs by variable, finding that topography explains more of temperature's sub-grid structure, while TESSERA provides additional surface information for wind speed. These improvements persist when the coarse input is changed from ERA5 to forecasts from the Aurora AI forecasting model, and when predicting at newly deployed stations with no regional history. To our knowledge, this is the first evidence that long-timescale Earth-observation embeddings can support short-timescale weather downscaling where sub-grid departures are systematically structured by persistent surface properties.

Problem

• weather downscaling
• sub-grid representation
• Earth observation embeddings
• surface properties
• probabilistic forecasting

Innovation

• Earth observation embeddings
• probabilistic downscaling
• sub-grid representation
• foundation models
• TESSERA