RAPID LIDAR ASSESSMENT AND VALIDATION IN THE BLADEN NATURE RESERVE

Over a decade of Lidar data acquisition in Belize has informed how past humans interacted with Belize’s diverse environments. However, diverse environments can pose challenges for Lidar data collection and analysis. In the Bladen Nature Reserve, Toledo District, the dynamic terrain and rainforest vegetation lend to greater noise and less distinct anthropogenic features in the Lidar dataset, making feature identification via a traditional interpreter-led approach untenable. Instead, we developed an iterative approach that utilizes a common model input, a local relief model, and the mounded characteristics of Maya structures to guide an area-based survey around the site of Ek Xux. This rapid assessment informs settlement distribution patterns while simultaneously elucidating limitations in the Lidar dataset that can be accounted for in future models. The method detailed here, or a variant of it, can be applied across Lidar datasets to situate model training data in its environmental context and create more robust feature maps through the comprehensive consideration of error.