Skip to Main Content
Many current models of ecosystem carbon exchange based on remote sensing such as the MODIS17 product are complex and require considerable input variables from ground-based meteorological measurements. They can introduce substantial errors into the carbon exchange estimates because these data are often not available at the same spatial scale as the remote sensing imagery. Here we propose a new net ecosystem carbon exchange (NEE) model solely based on MODIS data. Presumed that NEE can be simulated based only on the enhanced vegetation index (EVI), this model, termed the Temperature and Greenness (TG) model, also includes the land surface temperature (LST) product and land surface water index (LSWI) from MODIS. Site-specific data from the deciduous-dominated Harvard Forest AmeriFlux site were used. We analyzed 6 years (2001-2006) of CO2 flux data, the first four years used for model building and the others as validated set. The research showed that combination of the three parameters in the model could well reflect the correlation between predicted and measured NEE by eddy covariance technique at Harvard forest site. Furthermore, the TG model provided substantially better predictions of seasonal dynamics of NEE. Although it may be possible to improve the precision of various satellite-based production efficiency models through improved parameterization, our study suggests simpler empirical model based entirely on MODIS data could reliably estimate NEE.