From excavation to virtualization: reconstruction of the pre-Hispanic city of Oxtankah in Chetumal Bay (Quintana Roo, Mexico)

Abstract:

This study develops a comprehensive three-dimensional virtual reconstruction of the pre-Hispanic city of Oxtankah, located in the Chetumal Bay, Quintana Roo, Mexico. The primary objective is to enhance archaeological research, conservation, interpretation, and public dissemination by rigorously integrating traditional archaeological data with advanced digital technologies. This approach modernises interpretative infrastructure, museography and visitor information systems currently available at the archaeological site. The central objective is to apply virtual archaeology as a scientific methodology rather than merely as an illustrative resource. The reconstruction aims to produce a coherent, transparent and verifiable interpretative model of Oxtankah during its pre-Hispanic occupation. It is grounded in archaeological evidence, spatial data and comparative architectural analysis. The study explicitly adheres to the Seville Principles. This ensures methodological rigour, interdisciplinary collaboration, the traceability of interpretative decisions and the recognition of virtual reconstructions as hypothetical models subject to ongoing review. The research design was structured into two main phases: field-based data acquisition and laboratory-based digital processing. Fieldwork involved high-precision topographic and architectural surveys using GNSS RTK systems, total stations, and close-range photogrammetry. These methods enabled the generation of Digital Elevation Models (DEM), Digital Terrain Models (DTM) and orthomosaics with spatial resolutions of 1 cm or finer, providing a robust geometric framework for 3D modelling.

A total of 233 archaeological structures were documented through direct fieldwork, along with associated features such as albarradas (dry-stone walls), pathways, hydraulic elements and anthropogenic natural landscape modifications. Each element was recorded using standardised digital forms and integrated into a Geographic Information System (GIS), facilitating spatial analysis and data interoperability. For architectural remains that had not undergone archaeological excavation, the methodology combined surface observation, volumetric interpretation derived from topographic data and carefully justified architectural analogies, based on comparable Maya sites with similar chronology and construction techniques. The laboratory phase involved data processing using specialised software, including AutoCAD, QGIS, Surfer, and Agisoft Metashape. 3D modelling, texturing, lighting, and rendering were carried out primarily in Blender. The workflow followed established practices in virtual archaeology, including planimetric referencing, polygonal modelling, Unwrapping (UV mapping), application of Physically Based Rendering (PBR) materials, High Dynamic Range Imaging (HDRI) lighting environments and photorealistic rendering engines. Additional tools such as BlenderGIS, MakeHuman, and Adobe Photoshop were used for terrain modelling, human figure generation and post-production. All reconstruction decisions were documented and explicitly treated as interpretative hypotheses, open to revision as new archaeological evidence becomes available.

An extensive 3D virtual model of Oxtankah was the principal outcome the researchers created; it comprises approximately 440 architectural structures, including those extrapolated from topographical data evidence and analogue reconstructions for unexcavated contexts. This model demonstrates the potential of virtual archaeology to produce rigorous interpretative reconstructions that support archaeological research and heritage management. Detailed site reconstructions (including monumental, elite, domestic and perishable structures, as well as hydraulic infrastructure and surrounding landscape) enable both quantitative and qualitative analysis of urban form and historical development. The virtual environment integrates the natural landscape settings of the Chetumal Bay and Tamalcab Island, reconstructed using topographical data and satellite imagery; it includes agricultural areas based on the Mesoamerican milpa system. Native vegetation, procedurally generated endemic vegetation, enhances the realism of the environment. Seventy-five human figures, modelled using bioanthropological and ethnohistorical criteria, represent the daily, ritual and pre-Hispanic Maya society productive activities, contributing to a contextualised interpretation of urban life. As an applied outcome, twenty-seven interpretative panels were created directly from the 3D models and implemented in site signage and the museum exhibition programme. These visual resources provide scientifically grounded reconstructions of individual buildings and broad settlement overviews, greatly enhancing visitor understanding of the spatial layout, architectural complexity and historical development.

The study demonstrates that virtual archaeology is a robust and scientifically valid methodological approach to the integrated archaeological heritage management. The virtual reconstruction of Oxtankah supports digital preservation, facilitates spatial and architectural analysis, and enables a clear, reproducible interpretative hypotheses visualisation. The model also helps to disseminate research findings, connecting specialised archaeological knowledge to a wider public. The research acknowledges certain limits. Reconstructions of unexcavated structures remain hypothetical and depend on architectural analogies where direct evidence is lacking. These limitations are addressed thanks to virtual archaeology, which regards digital reconstructions as dynamic, revisable and well-documented academic products. This project offers a methodological framework that can be replicated in other archaeological contexts. It demonstrates that integrating advanced digital technologies with rigorous archaeological research can significantly enhance heritage management, museography and community engagement.

Show more Show less