Geospatial Chain of Thought Reasoning for Enhanced Visual Question Answering on Satellite Imagery

TL;DR

This paper introduces a geospatial chain of thought reasoning framework combined with direct preference optimization to significantly improve the interpretability, robustness, and accuracy of visual question answering on satellite imagery for climate-related applications.

Contribution

It presents a novel VQA framework that integrates CoT reasoning with DPO, enhancing reasoning capabilities and performance on complex geospatial tasks.

Findings

CoT supervision improves accuracy by 34.9% over baselines

DPO further enhances accuracy and reasoning quality

Enables richer geospatial reasoning for climate applications

Abstract

Geospatial chain of thought (CoT) reasoning is essential for advancing Visual Question Answering (VQA) on satellite imagery, particularly in climate related applications such as disaster monitoring, infrastructure risk assessment, urban resilience planning, and policy support. Existing VQA models enable scalable interpretation of remote sensing data but often lack the structured reasoning required for complex geospatial queries. We propose a VQA framework that integrates CoT reasoning with Direct Preference Optimization (DPO) to improve interpretability, robustness, and accuracy. By generating intermediate rationales, the model better handles tasks involving detection, classification, spatial relations, and comparative analysis, which are critical for reliable decision support in high stakes climate domains. Experiments show that CoT supervision improves accuracy by 34.9\% over direct baselines, while DPO yields additional gains in accuracy and reasoning quality. The resulting system advances VQA for multispectral Earth observation by enabling richer geospatial reasoning and more effective climate use cases.