A random sample of points were generated within fire perimeters (1984-2010) with a minimum nearest neighbor distance of 500m. We used the method developed by Davis et al. (2017): the number of points generated wthin each fire perimeter was equal to the square root of the perimeter area divided by 40. We then extracted points that fell within high-severity burn areas only. Points were randomly split between model training (80%) and testing (20%).
We provide advanced conservation science, technology, and planning to empower our partners in solving the world’s critical ecological challenges