Correction of Sun-View Angle Effect on Normalized Difference Vegetation Index (NDVI) With Single View-Angle Observation

Normalized difference vegetation index (NDVI) is one of the most widely used vegetation indices (VIs) to retrieve vegetation parameters such as fractional vegetation cover (FVC) and leaf area index (LAI). Due to the bidirectional reflectance distribution function (BRDF) effect on the surface, NDVI is greatly affected by the Sun-view angle, leading to considerable uncertainty in the parameters derived from NDVI. The angle effect can be corrected using the BRDF model. Nevertheless, the majority of satellites are unable to collect sufficient multiangle data in near real time to retrieve the parameters of the BRDF model. In this study, we proposed a correction model for the Sun-view angle effect of NDVI (SVAC) that only needed single view-angle observation for implementation. The SVAC model was developed based on a published cosine correction model (CCM) that corrected NDVI’s Sun angle effect. The simulated data and MODerate-resolution Imaging Spectroradiometer (MODIS) product data were used in validation. The results showed that the SVAC model performed well for all simulated scenes, where the NDVI’s uncertainty originating from the Sun-view angle was reduced by over 70% and 30% for homogeneous and nonhomogeneous vegetation, respectively. The validation with real data at the VAlidation of Land European Remote Sensing Instrumentations (VALERI) sites and the ImagineS sites demonstrated a percentage decrease of root mean square error (RMSE) of approximately 30%. The SVAC model can effectively reduce the NDVI’s sensitivity to Sun-view angles with high applicability and simplicity and is expected to facilitate the acquisition of vegetation parameters using angular-independent NDVI.