Simulated Global Carbon Budget And Its Interactions With Water Cycle Under Heterogeneous Atmospheric CO₂Enrichment Effect

Climate change is one of the biggest major concerns in our time, with great impacts on the carbon and water cycle of global terrestrial ecosystems. As atmospheric CO2substantially increase, understanding the CO2dynamic and the carbon sources and sinks at global scale is a prerequisite for a climate change control strategy. This study aims to discuss the global carbon budget and its interactions with water cycle under climate change and heterogeneous atmospheric CO2enrichment effect by using Scanning Imaging Absorption Spectrometer for Atmospheric Cartography (SCIAMACHY) retrievals of column-averaged dry air mole fractions of CO2(denoted XCO2) and Integrated Biosphere Simulator (IBIS). For these purposes, this paper investigated the spatial-temporal patterns of atmospheric XCO2and its relationship with ground-based background monitoring network (GLOBALVIEW), establishing a linear regression model between XCO2and near surface CO2. Several global datasets with spatial resolution of0.5degree were built up as input of the model such as climate datasets, land surface characteristic datasets, and environmental factors. Numerous field data were collected for model validation, including the results of Free Air CO2enrichment experiment (FACE) across major ecosystems. Based on the combination of the near surface CO2variations under heterogeneous scenario which were established from SCIAMACHY XCO2retrievals and the simulated results with reasonable validation, several conclusions were drawn as below:1) The seasonal amplitudes of atmospheric XCO2retrieved from SCIAMACHY have relationship with local vegetation types which determine the carbon sinks and sources capability. The correlation coefficient between XCO2and ground CO2were larger than0.75, indicating the high near surface sensitivity of SCIAMACHY XCO2, albeit of the undetectability in Amazon regions. Latitudinal distribution of XCO2has three differences with comparison to ground CO2:(1) substantial high concentration in equatorial regions;(2) extreme high value occurred in mid-and high latitude (30°N-60°N) regions of Northern Hemisphere;(3) disturbances from anthropogenic emission, dust, ice cover, and aerosol. From the relationship between GLOBALVIEW and SCIAMACHY XCO2, the linear regression model had been established. The precision of the model was validated by independent ground observations with the accuracy of2.7±2.0ppm. The processes of spatial extrapolation have a relative mean bias of0.74%and standard error of4.05ppm.2) As to carbon balance, the model simulated NPP and NEP well in most regions of the world. The amount of annual carbon sinks from terrestrial ecosystem in IBIS was higher than other studies’results. This is because the land cover and land use change did not considered into model simulations. The simulations of evaportranspiration captured the yearly and monthly variations well. The simulated carbon budget results showed reasonable variety and spatial patterns compared with other works.3) Under the spatial heterogeneity of atmospheric CO2scenario, the estimation of global mean annual NPP and NEP were0.5%and7%differ from traditional C sequestration assessments, respectively. The Amazon, Southeast Asia, and Tropical Africa showed higher C sequestration than traditional assessment with about15Gg C/yr/grid, and the rest of areas around the world showed slightly lower C sequestration than traditional assessment with about0-3Gg C/yr/grid. Climate change had great negative effects on global NPP and NEP with0.3Tg C/yr decline respectively, while spatial heterogeneous CO2had positive effects on NPP and NEP with0.1Tg C/yr and0.07Tg C/yr respectively. The spatial variation of NPP is associated with temperature in60to70%of the area of nonpolar terrestrial ecosystems, while70to80%of terrestrial land is colimited by precipitation. The spatial variations of atmospheric CO? have limited effects on spatial patterns of NPP, indicating that anthropogenic emission did not lead to significant effets on carbon budget at global scale.4) The Water Use Efficiency (WUE) showed a decreasing order for terrestrial vegetation types:Warm Temperate Forest>Temperate Forest>Tropical Forest> Boreal Forest>Savanna>Dense Shrubland>Grassland>Tundra>Desert.The high WUE distributed in southeastern US,eastern Canada, and Eurpoe and Russia where boreal forest mainly covered.In contrast,the low WUE distributed in Greenland, North Africa, and Central Asia. The estimation of global total WUE was0.3%lower than traditional C sequestration assessments, indicating larger response of global carbon budget to extreme event such as drought. The Amazon, Southeast Asia, and Tropical Africa showed higher WUE (～5mgC mm-1m-2yr-1) than traditional assessment, and the rest of areas around the world showed slightly lower C sequestration than traditional assessment within about1mgC mm-1m-2yr-1. Climate change, spatial heterogeneous CO2and uniform CO2scenarios had positive effects on global total WUE with3.9%,0.36%, and0.7%, respectively. The spatial pattern of WUE is associated with temperature and precipitation in27%and30%areas of terrain respectively. The spatial variations of atmospheric CO2did not have considerable effects on spatial patterns of WUE.