Historical Terrain Inference: Reconstructing Undocumented Landscapes of History

Historical Terrain Inference is a pilot-stage historical landscape reconstruction project designed to help organize, evaluate, and explain evidence-supported historical inference using GIS. The project focuses on how historical activity may have related to terrain, water, movement routes, settlement patterns, land cover, and documented historical references. It is designed to support research into undocumented or poorly documented historical activity areas by combining source evidence with landscape logic. The project uses an AI-assisted but human-reviewed workflow. AI may assist with extracting, organizing, comparing, and summarizing historical source material, but it does not make final conclusions. Human review, source criticism, GIS interpretation, uncertainty tracking, and historical judgment remain central to the process. The project separates historical evidence from interpretation. Source claims, historical map references, terrain observations, and candidate-area interpretations are treated as separate layers of information. This structure is intended to make the reasoning process more transparent, reviewable, and defensible. Historical Terrain Inference is currently focused on Civil War-era landscape reconstruction in selected pilot regions in Arkansas and North Mississippi. The long-term goal is to develop a careful, explainable framework that can support historical research, preservation planning, education, cultural-resource review, and public-history interpretation. This project does not claim to have discovered or verified any archaeological site. Candidate areas produced through the workflow are treated as reviewable interpretations that require further source review, expert evaluation, and, where appropriate, lawful and ethical field verification.