Ecosystem Respiration over a Mixed Forest Plantation of Northern India using MODIS-Derived Vegetation Indices and Flux-Tower Observations.

Abstract:

Accurate estimation of terrestrial ecosystem respiration (R_e) is crucial in assessment of regional to global carbon budget. In this study, continuous measurement of carbon dioxide exchange in 2014 was made over a young mixed forest plantation in Nainital district, India using the eddy covariance technique. The temperature dependence of night-time respiration were derived using non-linear optimization model between night-time NEE and corresponding night-time temperature and subsequently used in the model to estimate the daytime respiration. The model explained about 32.04% to 85.59% variation in Re (p=<0.05) throughout the season. Seasonally, the biophysical parameters such as leaf area index (LAI) and the leaf chlorophyll content explained majority of the R_e variations in the plantation. The study also attempted to examine the direct relationships between R_e and MODIS (MOD09) derived vegetation indices such as NDVI, EVI, SAVI, WDRVI, VARI, GI. The linear relationships between various vegetation indices and 8-day R_e was found to be significant. Among the selected indices, the EVI showed stronger relationship with ecosystem respiration (R² = 0.67, SEE = 1.42 F = 63.7, p =0.001) than other indices. This may be due to saturation of NDVI at high vegetation cover and more sensitivity to background reflectance. The seasonal course of R_e in forest plantation begins in March and continued to increase with increase in air temperature. The average rate of R_e was found to be 4.31 gC/m²/day, reaching to a maximum of 9.45 gC/m²/day during growing period. The seasonal effect of temperature (Q10) on ecosystem respiration was also examined and showed that there was an overall increasing trend in (Q10) from beginning of the growing season i.e. February to November. The Q10 value ranged from 1.19 to 4.78 throughout the year with an average of 2.51. This may be due to enhanced maintenance cost of respiration with increased biomass. The results demonstrated that simple remote sensing-based vegetation indices can be a proxy for R_e and could be helpful for the development of future R_e models over a large spatial scale. Also a new realistic model could be developed with higher accuracy by considering the sensitivity of air temperature and the soil moisture.

 

Keywords: Ecosystem respiration, eddy covariance, MODIS, Q10, vegetation indices.