Abstract
Geologic map, as a fundamental diagram in geology science, provides criticalinsights into the structure and composition of Earth's subsurface and surface.These maps are indispensable in various fields, including disaster detection,resource exploration, and civil engineering. Despite their significance,current Multimodal Large Language Models (MLLMs) often fall short in geologicmap understanding. This gap is primarily due to the challenging nature ofcartographic generalization, which involves handling high-resolution map,managing multiple associated components, and requiring domain-specificknowledge. To quantify this gap, we construct GeoMap-Bench, the first-everbenchmark for evaluating MLLMs in geologic map understanding, which assessesthe full-scale abilities in extracting, referring, grounding, reasoning, andanalyzing. To bridge this gap, we introduce GeoMap-Agent, the inaugural agentdesigned for geologic map understanding, which features three modules:Hierarchical Information Extraction (HIE), Domain Knowledge Injection (DKI),and Prompt-enhanced Question Answering (PEQA). Inspired by theinterdisciplinary collaboration among human scientists, an AI expert group actsas consultants, utilizing a diverse tool pool to comprehensively analyzequestions. Through comprehensive experiments, GeoMap-Agent achieves an overallscore of 0.811 on GeoMap-Bench, significantly outperforming 0.369 of GPT-4o.Our work, emPowering gEologic mAp holistiC undErstanding (PEACE) with MLLMs,paves the way for advanced AI applications in geology, enhancing the efficiencyand accuracy of geological investigations.