Global warming, mainly caused by the rapid increase of anthropogenic greenhouse gasemissions, has been in the spotlight in recent years and has profoundly affected sea level, globaland regional weather, food security, and ecosystem structure and biodiversity on the Earth.Vegetation, which plays an essential role in the carbon and water cycles of global terrestrialecosystems, is changing rapidly in response to the globally rising temperature reported by bothfield and satellite observations. In return, vegetation can also mitigate global warming throughincreased carbon sequestration via photosynthesis.However, a deep global understanding of the impact of warming and/or human landmanagement-induced changes in vegetation composition on the vegetation growth cycle(phenology) and the adaptation of vegetation to global warming is still insufficient. Based onthese research gaps, this thesis focuses on three research questions: 1) What historical andprojected atmospheric aridity changes may threaten the global terrestrial ecosystems? 2) Whatis the role of change in vegetation composition in the response of land surface phenology toclimate change? 3) Is the adaptation of vegetation to increasing temperature a globalphenomenon? These three questions are proposed to be answered in this thesis, which canimprove our understanding of the response of the global carbon cycle to climate warming.The first study evaluated the quality of the vapour pressure deficit (VPD) products derivedfrom three gridded climate datasets (ERA5, MERRA2, and CRU) against the in situobservations. Historical and projected changes in VPD were also estimated from the mean ofthree gridded climate datasets and CMIP6 simulations under different scenarios, respectively.The results showed that the estimated VPD from three gridded datasets (ERA5, MERRA2, andCRU) is consistent with the VPD derived from in situ observations (R2 = 0.915–0.954). Lowerconsistency was found mainly in areas with few station observations. Higher agreements werefound mainly in the summer and lower in the winter, related to the temperature level.Furthermore, global VPD has increased significantly (p < 0.05) from 1981 to 2020 at a rate of0.025 hPa year-1, and the increase in VPD derived from CMIP6 will continue in the future,particularly under high emission scenarios. This increase in VPD indicates an increase inatmospheric aridity in the future, which could threaten global terrestrial ecosystems, especiallycrops in drylands.The second study estimated the impact of tree cover (TC) and short vegetation cover (SVC)change on the length of growing season (LOS) using long-term satellite-observed NDVI dataand a Vegetation Continuous Fields (VCF) dataset covering from 1982 to 2015. We found that“PT” (permanent tree) has a substantially higher proportion (47.9%) of increasing LOS than“PSV” (permanent short vegetation) (19.6%). Furthermore, a prolonged growing season wasfound in 36.6% of the areas with an increasing trend in tree cover, while only 20.1% showedextended LOS with a decreasing trend in tree cover. In addition, areas with a larger proportionof tree cover are likely to show a more prolonged LOS (especially in boreal forests).The third study examined global temporal trends in optimal temperature for ecosystemproductivity (Topt), which was estimated as the monthly mean temperature that corresponds tothe maximum monthly gross primary productivity (GPP) in a given spatiotemporal “window”(5-year, 1°) based on satellite observations from 1982 to 2016. The potential drivers for thechanging Topt were also studied using simulations of the LPJ-GUESS model under differentenvironmental conditions. Finally, the CMIP6 outputs were used to estimate the projectedchanges in Topt from 2017 to 2100 under different scenarios of CO2 emissions. Globallysignificant (p < 0.05) increase in Topt was found at a rate of 0.019 ± 0.002 ° C y-1 from 1982to 2016, mainly driven by rising temperature and co-regulated by other factors, such as CO2level, rainfall, solar radiation, and nitrogen deposition. Furthermore, the projected Topt willlikely continue to increase as the temperature increases from 2017 to 2100.