首页|极端温度指数在中国地区基于ETCCDI指数新结果的CMIP5模式输出

极端温度指数在中国地区基于ETCCDI指数新结果的CMIP5模式输出

扫码查看
原文链接
  • 国家科技期刊平台
  • NETL
  • NSTL
  • 万方数据
使用新的数据集对由气候变化与监测指数专家团队(ETCCDI)定义的温度极端指数TX10p、TN10p、TX90p、TN90p、CSDI、WSDI的时空分布特点进行了研究.这6个指数都是基于百分位数的指数,通过半参数分位数回归的方法计算而得.我们在超出率序列中去除了自相关的影响,从而保证了计算分位数时的独立性条件.我们利用参与多模型比较计划第5阶段(CMIP5)的19种全球气候模型,在历史和3种典型浓度路径RCP 2.6、RCP4.5、RCP 8.5模拟了6个指数的变化.根据新的结果,中国的高温事件在未来有持续上升趋势,而低温事件有减少趋势.到21世纪末,在RCP26、RCP4.5、RCP8.5情景下,TX90p、TN90p和WSDI会显著上升,而TX10p、TN10p和CSDI会下降.在中国,低温事件频率下降最大的区域是东北地区,而高温事件频率上升最显著的区域是华南和西南地区.
Projected temperature extremes in China based on new results of ETCCDI indices for CMIP5 output
This study provides a view of temporal evolutions and spatial variations in temperature extremes indices (TX10p,TN10p,TX90p,TN90p,CSDI and WSDI) defined by the Expert Team of Climate Change and Detection Indices (ETCCDI) using a new dataset.All of the six indices are percentile-based and derived through semi-parametric quantile regression.The autocorrelation is removed from annual exceedance rate series,independence condition is guaranteed in estimation of percentiles.We use 19 global climate models,in the Couple Model Inter-comparison Project Phase 5,to simulate six indices for historical and three Representative Concentration Pathways (RCP2.6,RCP4.5 and RCP8.5) scenarios.The frequency of warm events in China shows an increasing trend in the future,while frequency of cold events tends to be lower in contrast.By the end of 21st century,TX90p,TN90p and WSDI will significantly increase,but TX10p,TN10p and CSDI will go through a reduction under RCP2.6,RCP4.5,and RCP8.5,respectively.The largest slump of cold events is projected in northeast China,the strongest escalation of warm events is projected in south and southwest China.

ETCCDI indicesCMIP5semiparametric quantile regressionpercentile-based indicesrepresentative concentration pathways

李灵珊

展开 >

北京师范大学全球变化与地球系统科学研究院,100875,北京

ETCCDI指数 CMIP5 半参数分位数回归 百分位数指数 典型浓度路径

Supported by The National Natural Science Foundation of China

41675044

2019

北京师范大学学报(自然科学版)
北京师范大学

北京师范大学学报(自然科学版)

CSTPCDCSCD北大核心
影响因子：0.505
ISSN：0476-0301
年,卷(期)：2019.55(1)
  • 3