Fire Emission Estimation Informed by Spectral Data

Environmental and remote sensing scientists at UC Berkeley are utilizing airborne and satellite spectral imaging data in a geospatial context for revising and improving the Berkeley Emission Estimation System (EES).

The Emission Estimation System (EES) adapts the USDA Forest Service First Order Fire Effects Model (version 4.0) to assess the fuels that contributed to a fire, the combustion amount and efficiency, and the resultant emissions.

Researchers have acquired AVIRIS hyperspectral data along with thermal data over a forested site in UC Berkeley’s Blodgett Research Forest to determine the spectral characteristics of the fuel load that may be utilized for calibrating MODIS data. They also plan to use UAVSAR data along with thermal and field data for assessing fuel moisture. Visible and near infrared spectral data was also collected on the ground using the SOC710-VP hyperspectral imaging system, for comparison with the satellite and AVIRIS (airborne) data.

“We will explore SAR, AVIRIS, MODIS, and TM/ETM+ data individually and together to determine their contributions in providing the inputs to the Berkeley Emission Estimation System (EES) models,” stated Dr. Greg Biging, head of UC Berkeley’s Biometrics and Spatial Analysis Lab. “The information gained from the higher resolution of hyperspectral data will provide insight into the analysis of lower spectral and spatial resolution data such as MODIS and TM/ETM+. The combined analysis of AVIRIS, UAVSAR, and thermal data will lead to new ways to determine fuel moisture and fire severity. Both fuel moisture and fire severity are needed for the improvement of the current EES model”.