Following severe wildfires across New Mexico fire scars—including Mora-San Miguel, Rociada, and Cerro Pelado—land and resource managers needed a rapid, repeatable way to locate burnt and hazardous trees along a 13-mile critical infrastructure corridor.

We built an automated point-cloud processing pipeline that fuses airborne Lidar with high-resolution multispectral imagery across 700 acres, normalizing data against pre-fire terrain models and executing deep-learning tree segmentation.

The project successfully identified and geolocated over 5,000 damaged trees, generating interactive web map layers for utility crews, foresters, and mitigation teams to prioritize hazard tree removal, suppression repair, and erosion prevention.