The Impact Of Land Use And Cover Change On Temperature Extremes And The Possible Mechanisms

Land use and cover change(LUCC)is an important anthropogenic forcing of the climate system.In this paper,we attempt to discuss the possible effects of LUCC on temperature extremes from the historical period to the different RCP scenarios of the future,based on multiple LUCC scenarios and multiple numerical simulations.Furthermore,we also explored the possible causes and mechanisms of the LUCC-induced changes in temperature extremes from the mean state shifting and changing variability perspectives.The main conclusions are shown as follows:(1)The idealized LUCC exert robust region-specific impacts on the multiple extreme temperature indices.Results of the idealized experiment regarding current versus historical LUCC scenario show significant increases(e.g.,0.2-0.7 ?)in extreme cold indices(e.g.,TNn)over central and eastern China,India and mainland Southeast Asia in response to such robust LUCC.Broad and significant decreases(e.g.,approximately-1?)in extreme warm indices(TXx)are mainly observed in Eastern Europe and western Siberia.Moreover,the effects of LUCC on high and low-percentile extreme indices of minimum temperature(Tmin)are generally symmetrical,but those for maximum temperature(Tmax)indices are asymmetrical,which are characterized by stronger influences on high-percentile indices rather than low ones.Further analyses suggest that LUCC-induced changes in temperature extremes are mostly influenced by shifts in the mean state of temperatures.Furthermore,the responses and sensitivities of Tmin/Tmax to LUCC are remarkably distinct among regions.This result mainly occurs because of different LUCC-induced changes in surface energy components,which depend on the region-specific climatology of each energy component and the dominant plant functional types.(2)Inconsistent responses of hot extremes to LUCC are identified among the selected models from the fifth phase of the Coupled Model Intercomparison Project(CMIP5).Regarding the past 50 years,four state-of-art CMIP5 models are selected to explore the multimodel performance of simulating LUCC-induced changes in hot extremes.In general,all the cases of the model from the Geophysical Fluid Dynamics Laboratory(GFDL)show robust asymmetric responses between the 90th(TX90P)and 10th percentiles(TX10P)of Tmax,mainly due to cropland expansions,especially over India,the Sahel and some parts of North America.However,weak and insignificant responses are shown for both the TX90P and TX10P in other models.The different changes in the Tmax variability among the models are primarily responsible for the occurrence of asymmetric features.Furthermore,by decomposing the Tmax changes over three typical regions,we analyze the potential causes for the inconsistencies among these models' results and find two crucial processes,i.e.,the repartitioning of the turbulent heat fluxes and the changes of the diurnal cycle variability due to LUCC.Whether these processes are pronounced determines the occurrence of the asymmetric Tmax responses.(3)Short time-scale(i.e.,20-year)of LUCC can also significantly affect the regional compound temperature extremes.Concerning the last 20 years,we discuss the potential effects of LUCC on the independent and compound temperature extremes over China,by using the RegCM4.4-CLM4.5 regional climate model with two experiments under 1980 and 2005 land cover maps.Results show that such LUCC scenario in China caused considerable decreases(increases)in the independent hot day(independent cold day),especially over deforestation regions(e.g.,northeast China).Moreover,the LUCC-induced robust increases(decreases)of independent hot night(independent cold night)are also detected over southern China(e.g.,Yunnan-Guangxi-Guizhou).However,the responses of compound temperature extremes to LUCC are rather weak relative to independent temperature extremes.The only observable LUCC signals in compound temperature extremes are detected in the compound cold events over northeast China.Furthermore,we also explore the mechanisms of the corresponding LUCC-induced changes in the Tmax and Tmin from the mean state shifting and the changing variability.In the daytime,the LUCC-induced changes of SRnet,SH and G affect not only the mean state but also the variability of temperature.But at nighttime,the LUCC-induced evolution of SH and G only dominate the mean state changes of temperature.(4)The projections show that future LUCC exert more robust impacts on joint temperature-humidity extremes than single temperature-based extremes.Based on the comparison of CMIP5 and LUCID-CMIP5 multimodel output,we find that LUCC exert significant coolings over the mid-to-high latitude regions(e.g.,northern U.S.),but warmings in the tropical regions(e.g.,central Africa)in the RCP2.6 scenario.Furthermore,the significance and the magnitude of the LUCC-induced changes in the joint temperature-humidity indices(i.e.,TME and THD)are more robust than those of the mean temperature in RCP2.6,particularly over the tropical regions.However,the results of all three temperature indices(i.e.,mean temperature,TME and THD)in RCP8.5 show insignificant responses globally.Furthermore,we also explore the extreme state(i.e.,10th and 90th percentiles)of changes in three temperature indices and find similar patterns in all cases,which suggest that the mean state changes of these temperatures dominate the changes in the extreme state.In addition,the projections of the LUCC-induced tendency in joint temperature-humidity under different RCP scenarios are also discussed.In summary,warm-humidity(cold-dry)tendencies are detected in the tropical(high latitude)regions in RCP2.6,whereas no significant tendency is shown in RCP8.5 scenario.