Forest biomass estimation from the fusion of C-band SAR and optical data using wavelet transform

Abstract

Forest biomass estimation is essential for greenhouse gas inventories, terrestrial carbon accounting and climate change modeling studies. Although a lot of efforts have been made in estimating biomass using both field-based and remote sensing techniques, no universal and transferable technique has been developed so far to quantify biomass carbon sources and sinks due to the complexity of the environmental, topographic and biophysical characteristics of forest ecosystems. This study investigated the potential of SAR (RADARSAT-2 dual polarizations) and optical (AVNIR-2) image fusion for biomass estimation using wavelet transform. Six different types of wavelets (haar, daubechies, symlet, coiflet, biorthogonal and discrete meyer) were tested with different rules and three decomposition levels for four different image combinations of SAR and optical data. The highest accuracy (r) of 0.84 was obtained from the fusion of NIR and HV polarization data, compared to 0.70 (r) from the NIR band alone. The results indicated a substantial improvement of biomass estimation from the fused images, and this accuracy is very promising, especially when using only one fused image in the high biomass situation of the study area, and gives a clear message to the research community that biomass estimation can be improved using the fusion of SAR and optical data due to their complementary information. Furthermore this fusion process can significantly reduce the saturation problem of optical and SAR data for biomass estimation.