摘要
用全球格点分析数据集(CRU TSv4.0)月降水资料和24个CMIP5(Coupled Model Intercomparison Project Phase 5)模式历史模拟数据以及RCP4.5情景下的预估数据,分析了多模式集合平均降水的偏差特征并进行了扣除模式气候漂移和一元对数差分回归订正。结果表明,模式降水在西部和北部明显偏多,东南沿海偏少;冷季(11月至次年4月)在全国大部分地区模式降水偏多,暖季(5~10月)东南沿海季风区偏少。1956~2005年多模式集合平均历史模拟降水偏差中84%属于气候漂移,其余是偏差的非定常模态。扣除气候漂移后,RCP4.5情景下2006~2015年中国模式降水预估偏差减小90%以上,大部分地区降水偏差百分率分布在±5%以内,仅在青藏高原西部和西北中部等地区模式降水偏多10%~40%;暖季降水偏差分布与年降水量类似;冷季偏差较大,北方降水偏多,南方偏少。检验表明,一元线性对数差分回归方程订正后,模式降水对于2006~2015年期间西南和江南中部的干旱少雨气候均能再现,且距平同号率高于多模式集合平均和扣除气候漂移的结果。用该方法对RCP4.5情景下2016~2035年模式预估降水进行订正,结果显示,南方(淮河以南)降水减少5%~20%,河套、内蒙古和华北北部减少20%~40%,东北南部、淮河流域、西北大部增加10%~40%及以上,东南沿海和台湾省降水增加10%~20%。以上降水预估结果说明,在RCP4.5情景下,21世纪前期持续十年的西南干旱会略有缓解,但南方降水偏少格局变化不大,淮河流域和三江源区及其以西等地降水可能明显增加。中国降水异常分布总体呈现南北少、中间多的格局,但北方和西部高山地带的降水预估存在较大的不确定性。
Based on precipitation data from CRU TS v4.0(Climatic Research Unit Time series 4.0) and 24 CMIP5(Coupled Model Intercomparison Project Phase 5) historical experimental models and the RCP4.5(Representative Concentration Pathway 4.5) scenario, this paper analyzed the bias and its correction method in precipitation projected by the ensemble of the CMIP5 multi-models. Bias of precipitation estimated over China for 2006-2015 under the RCP4.5 scenario was corrected by removing its climate drift and univariate logarithmic difference regression equation. Results showed that the model precipitation was usually overestimated in western and northern parts of China, while it was underestimated in the southeast coastal zone. In the warm season(May-October), the model precipitation was underestimated in the southeast coastal monsoon zone, while it was significantly overestimated in the cold season(November-April). 84% precipitation deviations simulated by the historical experiment from 1956 to 2005 belonged to climate drift, the rest to unsteady modes. Bias of model precipitation was decreased by about 90% after removal of climate drift, and the anomaly percentage in most areas was within ±5%. Model precipitation was overestimated 10%-40% in the western Tibetan Plateau and the middle northwestern region. Distribution of precipitation anomaly was similar to annual precipitation in warm season. The deviation of the cold season was larger, with more precipitation in the north and less in the south. The model was improved in its coincidence rate of anomaly sign after correction by using single linear logarithmic increment regression. The drought and reduced rain in the southwestern and central parts of the regions south of the Yangtze River reappeared during the period from 2006 to 2015, although the standard deviation was larger than the correction of climate drift. After the correction by SR-Log-Increment, the projected precipitation for 2016-2035 under RCP4.5 scenario showed that the precipitation would increase in the northeastern part of the south, the Huaihe River basin, and most parts of the northwest by about 10%-40%. Over the southeast coast and Taiwan,precipitation increased by about 10%-20%. It decreased in the south of the Huaihe River by about 5%-20%, while in Hetao, Inner Mongolia, and north of Northern China, the decrease was about 20%-40%. Results implied that under the RCP4.5 scenario, the southwest drought would be slightly relieved in the early 21 st century, while there would be little change in the pattern of less precipitation in the south. Also, precipitation would increase significantly in the Huaihe River basin, the source of the three rivers, and its western region. Distribution of the precipitation anomaly over China's Mainland would present a pattern of less precipitation in the south and north, with more in the middle. Nevertheless, the projection of precipitation in the northern and western alpine regions had clear uncertainty.
作者
张蓓
戴新刚
杨阳
ZHANG Bei;DAI Xingang;YANG Yang(Tianshui Meteorology Bureau,Tianshui,Gansu 741000;Key Laboratory of Regional Climate-Environment for Temperate East Asia,Institute of Atmospheric Physics,Chinese Academy of Sciences,Beijing 100029;College of Atmospheric Sciences,L anzhou University,Lanzhou 730000)
出处
《大气科学》
CSCD
北大核心
2019年第6期1385-1398,共14页
Chinese Journal of Atmospheric Sciences
基金
国家自然科学基金项目41475075、41675087
国家重点研究发展计划项目2016YFA0601901~~
