| Terrestrial ecosystem is one of the pools in the carbon and water fluxes at the globe scale, and also the one which has the highest sensitivity to the human disturbances. Therefore, it is rather essential to present the precise estimates on the carbon and water fluxes for terrestrial ecosystem, so as to deepen our understanding concerned on the functions of terrestrial ecosystem, as well as the carbon and water circulations. In this work, we used the moderate-resolution imaging spectroradiometer (MODIS) products and the data from the ground-based weather station and ground-based eddy covariance measurements to estimate the gross primary productivity (GPP) and ET for the vegetated regions in China over 2003-2009, via the light use efficiency model (LUE model) and Priestley-Taylor model respectively which are both used widely due to the great abilities to integrate with RS, and then analyzed the spatio-temperal patterns of water use efficiency (WUE), one indicator describing the water status of ecosystem which is calculated by the division between GPP and ET. The results was expected to provide some insight in the future managements on carbon and water regulations for vegetation ecosystem.The analysis based on the site measurements showed that both monthly summed GPP and monthly mean LUE have significant seasonal variabilities, which was coincident with the climate inner-annual variation. Although different climate-environmental factors were found to exert great influences on GPP and LUE at monthly interval, it is the monthly air temperature and the evaporative fraction (EF) that showed significant relationship with monthly summed GPP and monthly average LUE. Based on this finding, the GPP estimates from various GPP model, including vegetation production model (VPM) with the replacement of EF (VPM-EF) for LSWI which is used to calculate the water scalar in the original VPM, VPM, eddy covariance-light use efficiency model (EC-LUE), EC-LUE with the replacement of LSWI (EC-LUE-LSWI) for the water scalar in the original EC-LUE, vegetation index model (VI) and the MOD17A2 product, were compared using various statistical indicators. Results showed that both VPM-EF and VPM presented more precise estimates rather than other models, largely attributed to the enhanced vegetation index (EVI) instead of the normalized differential vegetation index (NDVI) used to parameterize the fraction of photosysthetically solar radiation absorbed by plant (FPAR). VPM-EF and EC-LUE both had better performances compared to VPM and EC-LUE-LSWI respectively, showing that it is suitable to using the EF to be the indicator of the water scalar in LUE model.GPP estimates from VPM-EF showed that the annual mean GPP over 2003-2009 was 5.48 Pg C, the carbon density was approximately 540.31 gC m-2. ET estimates from Priestley-Taylor model showed annual mean ET was 534 mm. The zonal analysis was found for the spatial distributions of both GPP and ET, that is, large values were found in the southern and coastal regions, while the small values in the northern and inland regions. Among the various vegetation types, the forest had largest GPP, while the shrub and grassland had the less GPP. Slightly different from GPP, the largest ET were found in the evergreen broadleaf forest, mixed forest and cropland, while the less ET in the shrub and grassland. Nevertheless, significant differences were also found among distinguishing vegetation regionalization regions for each vegetation types, suggesting that the tangled effects of plant and climate-environmental factors.WUE was further analyzed based on the GPP and ET estimated as above in the vegetated regions. Results showed that the annual mean WUE was large in the Great Khingan, Changbai Mountain, Qinling Mountains and the southeast hilly area, with a value of 2.0 gC kg-1 H2O, while the small value were found in the northwest and the Qinghai-Tibet Plateau, with a value below 1.0 gC kg-1 H2O. The forest had the largest WUE value among various vegetation types, contrary to this, the shrub and grassland had the smallest WUE. The seasonal variability of monthly mean WUE was largely governed by the monthly mean temperature in most regions except in temperate steppe zone (TSZ) where the monthly precipitation exert greater influences on seasonal variability of WUE for various vegetation types. Spatially, the regions with the highest sensitivity of monthly mean WUE to climate were located in the transitional regions of natural geographical areas. Inter-annually, the changes of WUE was largely regulated by GPP, rather than ET during the study period. |
PDF Full Text Request