Modeling MODIS NBAR time series of vegetated surfaces and its use in LAI recursive estimation

The inconsistent data quality of remote sensing observation, which is mainly caused by atmospheric conditions, presents problems in the application of these data. For the land cover types that cycle yearly, the variations in surface reflectance usually have temporal periodic characteristics. In this study, we modeled the temporal feature of Moderate-Resolution Imaging Spectroradiometer (MODIS) Nadir BRDF-adjusted reflectance (NBAR) time series data of the typical vegetated area using the season-trend statistical method. The fitting values of season-trend model were applied to the recursive estimation of leaf area index (LAI) time series based on nonlinear autoregressive exogenous (NARX) neural network. The results of MODIS NBAR modeling indicate that the season-trend method is effective to model the NBAR time series of the vegetation surface. The NARX neural network works well using the improved NBAR time series as input, and the estimated LAI time series is more continuous than the MODIS LAI.