ABOVEGROUND BIOMASS ESTIMATION OF SALT CEDAR IN CIBOLA NATIONAL WILDLIFE REFUGE

Remote sensing has become an essential tool in biomass inventory and estimation. This paper presents the biomass estimation of salt cedar, an invasive species in the riparian area of Lower Colorado River Basin, through the biomass destructive sampling and using of Landsat TM5 imagery. In the summers of 2009 and 2011, destructive samplings were conducted in Cibola National Wildlife Refuge, Arizona to study the allometric equation of salt cedar given no existing allometric equation available for salt cedar. The allometric relationship was explored between total aboveground biomass and average diameter, average height, and total basal area. The validity and the strength of the allometric models were examined with adjusted coefficient of determination (r²), root mean square error (RMSE), and Akaike Information Criterion (AIC). The regression model with all three predictors, basal area, tree height, and diameter, had the adjusted r² value equal to 0.924; and the AIC and RMSE values were the smallest among all the models. The Landsat TM5 imagery was obtained and corrected with 6S radiotive code for atmospheric attenuation. Then, by making use of the obtained allometric equation and the 6S corrected TM5 imagery, the estimation of biomass of Cibola was conducted and explored with original bands, Normalized Difference Vegetation Index (NDVI), Principal component analysis (PCA), and wavelet analysis. The efficiency and accuracy of the four methods were compared and discussed.