GeoPhysAdapter: Scale-Matched Geophysical Adaptation for Cross-Domain Landslide Mapping with Vision Foundation Models

2026-08-10

AI Summary

This study addresses the problem of high-confidence false positives in cross-regional landslide mapping by vision foundation models, which arises from a lack of geographic context. To mitigate this issue, the authors propose a dual-level bounded adaptive framework that integrates multi-scale geoscientific priors—encompassing topography, material properties, and rainfall triggers. The method enables synergistic decision-making at both pixel and landslide-candidate levels, with a novel alignment of geoscientific priors to candidate units to achieve scale-matched adaptation. By incorporating dense spatial guidance, regional modulation, and temporal constraints, the approach significantly reduces false alarms across 55 global landslide events: candidate-level adaptation yields a 23.99% error reduction, a relative IoU improvement of 14.2% (+0.031), and corrects 9.92 false positive pixels for every sacrificed true positive pixel.

Abstract

Newly triggered landslides rarely carry immediate annotations, so cross-domain transferability determines the value of landslide mapping for emergency response and regional risk assessment. Vision foundation models have strengthened representational transfer, yet on unseen regions, events, and data sources they still generate high-confidence false alarms. Terrain, material, and rainfall triggering can constrain such errors, but their supports are local, regional, and event-scale, so that resampling onto a 10~m grid misaligns them with the segmentation decision unit and compounds the uncertain geographic context problem (UGCoP). We propose GeoPhysAdapter, which anchors on a frozen vision foundation model, restricts terrain, material, and triggering to dense spatial guidance, regional modulation, and event-timing forcing, and applies bounded adaptation at two decision units, the pixel and the candidate landslide body, reverting exactly to the visual prediction where support is insufficient. On an event-isolated PILD dataset of four public sources, 55 global landslide events, and 7,890 test samples, 70.3% of cross-domain false-positive mass lies in near-pure spurious bodies of median equivalent diameter 207m, matching coarse-prior support rather than the pixel. Pixel-level adaptation removes a net 507,817 erroneous pixels and reduces error by 7.76%, whereas raising the decision unit to the candidate body, under identical samples, anchor, and baseline, increases error reduction to 23.99%, approximately 3.1 times the pixel-level effect, improves IoU by 0.031 (14.2% relative), and corrects 9.92 pixels per pixel harmed. The data and code are publicly available at: https://github.com/Liu-Zhihang/geophysadapter.

Problem

• cross-domain landslide mapping
• uncertain geographic context problem
• scale mismatch
• false alarms
• vision foundation models

Innovation

• GeoPhysAdapter
• cross-domain adaptation
• vision foundation models
• geophysical priors
• landslide mapping