Synergy of thermal and optical remote sensing signatures with biophysical models for estimating dynamic change of biomass and CO_2 flux in agricultural ecosystems

摘要

The objective of this study was to investigate the potential of the synergy between the biophysical/ecophysiological models and remote sensing signatures for dynamic estimation of key biophysical variables at the ecosystems-atmosphere interface. We obtained a long-term and comprehensive data set of micrometeorological, plant, and remote sensing (optical and thermal domains) measurements over well-managed uniform agricultural fields. The net ecosystem CO_2 flux (NEECO2) was measured by the eddy covariance method (ECM). A soil-vegetation-atmosphere transfer (SVAT) model was used to describe the energy balance, water budget, and physiological processes in the soil-vegetation-atmosphere system that allowed simulating the seasonal change of CO_2 and water fluxes as well as biomass, photosynthesis, soil water, and surface temperatures. Both remotely sensed surface temperature and spectral reflectance were useful to effectively tune the process-based model, so that biomass, evapotranspiration, and CO_2 flux were accurately simulated. Simulated NEECO2 agreed nicely with those measured by ECM, while simulated biomass agreed well with independent measurements. The synergy of remote sensing and process-based modeling was quite effective in utilizing infrequent and multi-source remote sensing data. This approach would have great potential for quantitative and dynamic assessment of multiple variables in terrestrial ecosystems.