Assessment of red-edge vegetation descriptors in a modified water cloud model for forward modelling using Sentinel – 1A and Sentinel – 2 satellite data

ABSTRACT

This study investigates the potential of different vegetation descriptors (V) in the modified water cloud model (MWCM), with a focus on comparing the Red – Edge vegetation indices (VI) based and other vegetation descriptors using Sentinel – 1A and Sentinel − 2 satellite data. In order to reduce the influences of vegetation and roughness effectively, a soil geometrical model of equivalent roughness was coupled with a MWCM for the simulation in forward modelling. The optimum value of vegetation extinction coefficient (K_VI) and dense vegetation indices ($V{I}_{\mathrm{\infty }}$) of modified Beer’s law were calculated using non-linear least square optimization technique. After the parameterization of MWCM, the five different types of V for wheat crop were tested for the simulation of backscattering coefficients (${\sigma }^0\left(\mathrm{d}\mathrm{B}\right)$) at VV polarization. The higher statistical performance indices like coefficient of determination (R² = 0.96), root-mean-square error (RMSE = 0.17 (dB)) and Nash sutcliffe efficiency (NSE = 0.93) were found for the case of Red-Edge-based vegetation descriptors (for VI = NDVI_RE) than others in MWCM. Therefore, this methodology reveals that the VI = NDVI_RE by modified Beer’s law can be used effectively as the vegetation parameter in the MWCM for accurate simulation in forward direction over vegetation-covered areas.

Acknowledgments

The authors would like to thank ESA (European Space Agency) for providing free access to the Sentinel – 1A and Sentinel – 2 satellite data and SNAP tool used in the present study.