Understanding the Spectral Signature Characterization of Cropland Burning

Abstract:

Accurate and reliable cropland burned area products are essential for aiding estimates of trace atmospheric gases and particles which influence atmospheric chemistry dynamics and air quality. Cropland residue burning has been estimated to account for approximately 8-11% of all global fires (Korontzi et al. 2006). Cropland residue burning is also a significant fraction of atmospheric emissions at the local level. Timely and accurate assessments of cropland fire activity are needed by local and regional administrators involved in carbon emission, human health, and air quality regulations. Remote sensing systems used for monitoring fire activity have significant detection uncertainty due to ambiguous spectral characteristics of cropland burning. The present research is focused on the quantitative characterization of the spectral signature of cropland burning by monitoring the spectral signal throughout the full growing, harvest, and burning periods. Field data collection was conducted in the southern Palouse region of Idaho and Washington. Reflectance was measured on a test area of spring wheat fields with an analytical spectral device (ASD) spectrometer. All the dates for field data collections are coincident with Landsat 8 overpasses, and the spatial sampling of field measurements were performed following the VALERI protocol for upscaling to 30 meter pixels (Baret et al. 2005).