Potential impacts of land-use on climate variability and extremes

This study aims at exploring potential impacts of land-use vegetation change (LUC) on regionalclimate variability and extremes. Results from a pair of Australian Bureau of Meteorology Research Centre(BMRC) climate model 54-yr (1949--2002) integrations have been analysed. In the model experiments, twovegetation datasets are used, with one representing current vegetation coverage in China and the otherapproximating its potential coverage without human intervention. The model results show potential impactsof LUC on climate variability and extremes. There are statistically significant changes of surface interannualclimate variability simulated by the model. Using different vegetation datasets, significant changes incorrelation coefficients between tropical Pacific Ni~no3.4 SST and precipitation and surface temperatureover East Asia are identified, which indicate that changes in vegetation coverage may alter ENSO impactson regional climate variability. Because of the lack of slowly varying surface processes when forests areremoved and less rainfall is received following LUC, the ENSO signal simulated by the model becomes stronger.Results furthermore show that land-use could modulate characteristics of decadal variations in this region.When using current vegetation coverage, the model gives better simulation of observed climate variations inthe region than the case using potential vegetation coverage. In addition, results suggest that land-use couldbe a potential factor contributing to the prolonged drought in central-west China. Changes in local climateextremes, including precipitation and surface temperature maxima and minima, are also identified. Overall,this study has illustrated the importance of further investigation of such important issues in future land-usestudies.