Computational Advocacy for Imperiled Landscapes: A Temporal Exploration of Mountaintop Removal's Ecological Impact

This project explores the environmental impact of mountaintop removal (MTR) in the Appalachian Mountains, focusing on the significant ecological damage caused by this coal extraction method. Utilizing high-resolution satellite imagery and LiDAR data, the project aims to analyze the landscape changes and habitat destruction resulting from MTR. Methodologies such as species distribution and habitat modeling are employed to assess the suitability of areas for local ecology, while LiDAR classification and segmentation techniques offer more detailed spatial analysis than traditional methods. The research specifically targets the habitat needs of five critically endangered species in Appalachia, aiming to develop a comprehensive framework that integrates LiDAR and satellite imagery for precise habitat classifications. Four MTR sites of various operational statuses were examined, including the Hobet 21 coal mine, Keppler Mine Impoundment, Compass Mine, and areas within the Cranberry Wilderness and Big Ugly managed wilderness area. By leveraging Quality Level 2 LiDAR data and Sentinel-2 satellite imagery, the project enhances the classification of landscape features, enabling a nuanced analysis of habitat variations relevant to threatened species. Challenges in integrating LiDAR and satellite data were encountered, notably in training a PointNet++ model for 3D classification, underscoring the complexity of merging different data types. The project has generated a dataset of 3,000 classified and labeled point clouds, laying the groundwork for more refined ecological classifications.