Abstract
Semantic search of Earth observation archives remains challenging. Visual foundation models such as CLAY produce rich embeddings of satellite imagery but lack the natural-language grounding needed for intuitive query, and full contrastive training of a remote-sensing CLIP-style model requires paired data and compute that are unavailable at global scale. We present GeoQuery, a zero-shot retrieval system that sidesteps this constraint through prompt-aligned text proxies. Rather than training a joint encoder, we generate language descriptions for a 100k proxy subset of global Sentinel-2 tiles and optimise the description-generation prompt so that distances in the resulting text-embedding space correlate with distances in the frozen CLAY visual-embedding space. Queries are resolved in two stages, with a text-similarity search over the proxy subset followed by a visual nearest-neighbour search over worldwide CLAY embeddings. On 76 disaster-location queries covering UK floods, US wildfires, and US droughts, GeoQuery achieves 31.6% accuracy within 50 km, with the strongest performance on floods (50% within 50 km) where terrain features are well captured by RGB embeddings. Deployed within ECHO, a crisis response system using Agentic Action Graphs, GeoQuery identified vulnerable areas during Brisbane's 2025 Cyclone Alfred, with downstream flood simulations reproducing historical patterns. Prompt-aligned proxies offer a practical bridge between EO foundation models and operational retrieval when full contrastive training is out of reach.
Abstract (translated)
URL
https://arxiv.org/abs/2605.05405
