Identification of the Hurricane Katrina Impacted Forest Region Using MODIS Measurements

Wanting Wang, George Mason University, wwang@gmu.edu (Presenting)
John J Qu, George Mason University, jqu@cos.gmu.edu
Xianjun Hao, George Mason University, xhao@gmu.edu
Yongqiang Liu, USDA Forest Service, Forestry Sciences Laboratory, yliu@fs.fed.us

Severe hurricanes can extensively influence the composition, structure and natural succession of forests. From the perspective of fire management, the most immediate impact of hurricanes is that a massive amount of living forest biomass turns to dead fuel, and the consequent increase of fuel bed depth and decrease of dead fuel moisture. Previous researches have indicated that occurrence and intensity of forest fires substantially increases in hurricane-impacted areas in the succeeding years after landfall. This phenomenon has made estimations of fuel loading change important to evaluate fire danger in the post-hurricane region. In this poster, a vegetation index was selected to evaluate the impacted area and identify severity categories. The MODIS Nadir BRDF and Atmospheric Corrected Reflectance product was preprocessed to eliminate or significantly reduce inter- and intra-annual vegetation phenology mediated errors and anomalous high and low values. An image differencing method was then employed to detect the hurricane impacted region. The range of hurricane impacted areas and severity categories are identified using the thresholds derived by statistical analysis.