Interannual And Interdecadal Variability Of The Surface Air Temperature And Precipitation In East Asia Based On CMIP5

Using the CMIP5 runs for global climate system decadal hindcast of the surface temperature and precipitation, with the NCEP\NCAR monthly reanalysis data as observation data, assessment of interannual variability of surface air temperature and precipitation was conducted by each model、Bayesian Model Averaging and multi-model ensemble mean. Then the variability of future temperature and precipitation are estimated in East Asia based on this. At the same time, BMA exhibited high performance in hindcasting the mean of temperature and precipitation, so in the RCP4.5 scenario, the projection of four seasons was used to estimate during the global warming. And the uncertainties of seasonal change projection are addressed.The results showed that, in terms of the variance distribution, the spatial and temporal distribution of the surface air temperature variation as well as the periodic oscillation characteristics, the skills of single models, EMN, BMA hindcasts are reasonably good by using EOF analysis and Morlet wavelet analysis.10 models、 EMN and BMA can hindcast the variance distribution quite well for the surface air temperature at the period of 1981-2010. And among them BMA provides the best hindcast while GFDL-CM2p1 perform best for precipitation. Considering the correlation coefficient of the space and time、the standard deviation and the root mean square, BMA and IPSL-CM5A-LR show better in the first mode of surface temperature. MIROC5 and CanCM4 do better in in the second mode of surface temperature. MPI-ESM-LR and GFDl-CM2p1 do well in the first mode of precipitation. Ensemble and CCSM4 do better in second mode of precipitation. BMA can not only hindcast the trend, but also the detrended variabilities of the second EOF mode of fairly well. Wavelet analysis suggests that CMCC-CM model systems can hindcast the quasi-quadrennial oscillation reasonably well, which is associated with that of the NCEP data. BMA can provide the uncertainties of surface temperature and precipitation well. For the hindcast of the temperature, the uncertainty over the oceans is less than over the continents, and precipitation is vice. Overall Bayesian model average temperature and precipitation for return better results.In the RCP4.5 scenario, The results show that, annual temperature fluctuation of 2006-2035 is less than 1981-2010, while precipitation fluctuation of 2016-2025 is greater than 2001-2010. Temperature in East Asia is decreased firstly and then increased, the turning point is 2024. The temperature of 2017-2035 is given priority to with 5-6 years cycle. Temperature and precipitation are found to be more affected by seasonal variability, with the increase in temperature over East Asia more intense in autumn (mainly), while the change of precipitation is obviously less than temperature.