Modeling gross primary production of an evergreen needleleaf forest using modis and climate data

Forest canopies are composed of photosynthetically active vegetation (PAV, chloroplasts) and nonphotosynthetic vegetation (NPV, e.g., cell wall, vein, branch). The fraction of photosynthetically active radiation (PAR) absorbed by the canopy (FAPAR) should be partitioned into FAPAR(PAV) and FAPAR(NPV). Gross primary production (GPP) of forests is affected by FAPAR(PAV). In this study we developed and validated a satellite-based vegetation photosynthesis model (VPM; GPP = epsilon(g) X FAPAP(PAV) X PAR) that incorporates improved vegetation indices derived from the moderate resolution imaging spectroradimeter (MODIS) sensor. Site-specific data from the CO, flux tower site (evergreen needleleaf forest) at Howland, Maine, USA, were used. The enhanced vegetation index (EVI) better correlated with the seasonal dynamics of GPP than did the normalized difference vegetation index (NDVI). Simulations of the VPM model were conducted, using both daily and eight-day composites of MODIS images (500-m spatial resolution) and climate data (air temperature and PAR), respectively. Predicted GPP values in 2001 agree reasonably well with estimated GPP from the CO2 flux tower site. There were no significant differences in VPM-predicted GPP (from eight-day MODIS composites) among one pixel (similar to 500-m resolution), 3 X 3 pixel block (similar to 1.5-km resolution), and 5 X 5 pixel block (similar to 2.5-km resolution). The differences between VPM-predicted and observed GPP were smaller for simulations using eight-day MODIS composites than for simulations using daily MODIS images. The results of this study have shown the potential of MODIS data (both daily and eight-day composites) and the VPM model for quantifying seasonal and interannual variations of GPP of evergreen needleleaf forests.