Monitoring the seasonal and interannual variation of the carbon sequestration in a temperate deciduous forest with MODIS time series data

Abstract Understanding the seasonal and interannual variation of the ecosystem carbon balance and the mechanisms that control it is crucial for assessing the vulnerability of the terrestrial carbon pools under future changing climate conditions. Expanding on the previous work, an improved predictive model was developed to estimate the carbon sequestration of a temperate deciduous forest exclusively from remotely sensed time series data, including MODIS land surface temperature (LST), Terra nighttime LST′, enhanced vegetation index (EVI), land surface water index (LWSI), the fraction of absorbed photosynthetically active radiation (FPAR) and leaf area index (LAI). The objectives of this study are to relate remotely sensed proxies of environmental drivers to both seasonal and interannual carbon sequestration observed from seven years’ net ecosystem carbon exchange (NEE) data, and to extrapolate long-term trends in forest carbon balance. Our results suggest that the improved model provided good estimates of NEE and well reflected the seasonal dynamics of the temperate deciduous forest, though with limited skill during the spring and autumn transitions. Correlation analysis showed that these MODIS products have different relationships to NEE at 8-day and interannual time scales. The mature forest ecosystem acted as a strong carbon sink during the entire decade; however, a trend of decreasing carbon sequestration in the forest was observed as time goes on.