Spectroscopic analysis for mapping wildland fire effects from remotely sensed imagery

Abstract

Wildfire burns 1.5–4 million ha of land across the United States each year, contributing to postfire erosion, ecosystem degradation, and loss of wildlife habitat. Unmanned aircraft systems and sensor miniaturization offer a new paradigm, providing an affordable, safe, and responsive on-demand tool for monitoring fire effects at a much finer spatial resolution than is possible with current technology. Using spectroscopic analysis of a variety of live and combusted vegetation samples to identify the spectral separability of vegetation classes, an optimal set of spectra was selected to be utilized by machine-learning classifiers. This approach allows high-resolution mapping of wildland fire severity and extent.