Since both remote sensing and modeling methods have inherent limitations in estimating ecosystem dynamics, this study investigated the potential of synergy of remote sensing and process-based modeling. A comprehensive data set was constructed, which included ground-based and airborne remote sensing data in optical and thermal wavelength domains as well as CO2 flux by eddy covariance method, and plant and micrometeorological data over 8 years. Result showed that an approach to tune a soil-vegetation-atmospheric transfer model (SVAT model) using remotely-sensed data could allow accurate and dynamic assessment of important ecosystem variables such as biomass, CO2 flux, and soil moisture content.