TerraScope: Pixel-Grounded Visual Reasoning for Earth Observation

TL;DR

TerraScope is a novel vision-language model designed for precise pixel-level geospatial reasoning in earth observation, capable of handling multiple modalities and temporal sequences, supported by a large dataset and benchmark.

Contribution

It introduces TerraScope, a unified model with modality-flexible and multi-temporal reasoning, along with the Terra-CoT dataset and TerraScope-Bench for evaluation.

Findings

Outperforms existing VLMs in pixel-grounded geospatial reasoning

Provides interpretable visual evidence

Achieves high accuracy on multiple sub-tasks

Abstract

Vision-language models (VLMs) have shown promise in earth observation (EO), yet they struggle with tasks that require grounding complex spatial reasoning in precise pixel-level visual representations. To address this problem, we introduce TerraScope, a unified VLM that delivers pixel-grounded geospatial reasoning with two key capabilities: (1) modality-flexible reasoning: it handles single-modality inputs (optical or SAR) and adaptively fuses different modalities into the reasoning process when both are available; (2) multi-temporal reasoning: it integrates temporal sequences for change analysis across multiple time points. In addition, we curate Terra-CoT, a large-scale dataset containing 1 million samples with pixel-level masks embedded in reasoning chains across multiple sources. We also propose TerraScope-Bench, the first benchmark for pixel-grounded geospatial reasoning with six sub-tasks that evaluates both answer accuracy and mask quality to ensure authentic pixel-grounded reasoning. Experiments show that TerraScope significantly outperforms existing VLMs on pixel-grounded geospatial reasoning while providing interpretable visual evidence.