Assessing multisensor integration for geological modeling: UAV photogrammetry, LiDAR, and hyperspectral data in a visualization workflow

Accurate geological interpretation depends on integrating multiple remote sensing datasets, as single-sensor methods often miss the full spatial and spectral complexity of outcrops. This study introduces a multi-sensor workflow that combines UAV photogrammetry, close-range smartphone LiDAR, and VNIR – SWIR hyperspectral imaging. The datasets are fused during preprocessing, stored in a relational SQL database, and embedded in a 3D visualization environment for interactive analysis. Within this platform, users can perform real-time rendering and point-based spectral interrogation of georeferenced data at centimeter-to-meter scales. We applied this workflow to the Vaca Muerta Formation, identifying four stratigraphic units and resolving vertical mineralogical gradients, including the distribution of illite, smectite, gypsum, and chlorite. Classification accuracy was comparable to traditional hand specimen analysis, while integrated visualization revealed structural and mineralogical boundaries missed by single-sensor methods. We assess how spatial resolution, mesh simplification, and sensor alignment influence geometric fidelity, uncertainty, and interpretive reproducibility. By embedding spectral data in a georeferenced 3D framework, our workflow enables reproducible stratigraphic segmentation and facies differentiation in fine-grained, low-contrast sedimentary systems. Though developed for geology, the approach is transferable to other fields requiring precise multi-sensor integration, such as urban planning, civil engineering, and digital twin development.