5 May 2016 Semi-empirical modelling for forest above ground biomass estimation using hybrid and fully PolSAR data
Author Affiliations +
Abstract
Forests act as sink of carbon and as a result maintains carbon cycle in atmosphere. Deforestation leads to imbalance in global carbon cycle and changes in climate. Hence estimation of forest biophysical parameter like biomass becomes a necessity. PolSAR has the ability to discriminate the share of scattering element like surface, double bounce and volume scattering in a single SAR resolution cell. Studies have shown that volume scattering is a significant parameter for forest biophysical characterization which mainly occurred from vegetation due to randomly oriented structures. This random orientation of forest structure causes shift in orientation angle of polarization ellipse which ultimately disturbs the radar signature and shows overestimation of volume scattering and underestimation of double bounce scattering after decomposition of fully PolSAR data. Hybrid polarimetry has the advantage of zero POA shift due to rotational symmetry followed by the circular transmission of electromagnetic waves. The prime objective of this study was to extract the potential of Hybrid PolSAR and fully PolSAR data for AGB estimation using Extended Water Cloud model. Validation was performed using field biomass. The study site chosen was Barkot Forest, Uttarakhand, India. To obtain the decomposition components, m-alpha and Yamaguchi decomposition modelling for Hybrid and fully PolSAR data were implied respectively. The RGB composite image for both the decomposition techniques has generated. The contribution of all scattering from each plot for m-alpha and Yamaguchi decomposition modelling were extracted. The R2 value for modelled AGB and field biomass from Hybrid PolSAR and fully PolSAR data were found 0.5127 and 0.4625 respectively. The RMSE for Hybrid and fully PolSAR between modelled AGB and field biomass were 63.156 (t ha-1) and 73.424 (t ha-1) respectively. On the basis of RMSE and R2 value, this study suggests Hybrid PolSAR decomposition modelling to retrieve scattering element for AGB estimation from forest.
© (2016) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
Kiledar Singh Tomar, Kiledar Singh Tomar, Shashi Kumar, Shashi Kumar, Valentyn A Tolpekin, Valentyn A Tolpekin, Sushil Kumar Joshi, Sushil Kumar Joshi, "Semi-empirical modelling for forest above ground biomass estimation using hybrid and fully PolSAR data", Proc. SPIE 9877, Land Surface and Cryosphere Remote Sensing III, 987729 (5 May 2016); doi: 10.1117/12.2223639; https://doi.org/10.1117/12.2223639
PROCEEDINGS
11 PAGES


SHARE
Back to Top