Methodologies for integrating multiple high-level remotely sensed land products

Abstract

Various high-level biophysical and geophysical products have been generated from the same or different satellite data. Each individual product may have variable qualities and accuracies, but the integration of multiple products of the same land surface variable can reduce the uncertainty and improve their reliability. This chapter presents our recent studies on integrating multiple high-level land products using different techniques, such as optimal interpolation and empirical orthogonal functions (EOF) for the LAI products, the multiresolution tree (MRT) method for broadband albedo, the fraction of absorbed photosynthetically active radiation by green vegetation (FAPAR), land surface emissivity products, and the Bayesian model averaging (BMA) method for longwave downward radiation and surface latent heat flux products.