基于CMIP6多模型的北太平洋海表温度模拟能力评估  

ASSESSMENT OF CMIP6 MODELS IN REPRODUCING OBSERVED SEA SURFACE TEMPERATURE IN THE NORTH PACIFIC OCEAN

作　　者：王帅奇 孙群 国文宁 WANG Shuai-Qi;SUN Qun;GUO Wen-Ning(School of Ocean and Environment,Tianjin University of Science and Technology,Tianjin 300457,China)

机构地区：[1]天津科技大学海洋与环境学院,天津300457

出　　处：《海洋与湖沼》2025年第2期262-274,共13页Oceanologia Et Limnologia Sinica

基　　金：崂山实验室科技创新项目,LSKJ202202104号。

摘　　要：北太平洋海域是反映全球气候变化的敏感区域,特别是对紧邻西北太平洋的东亚地区的天气与气候变化有重大影响。本文选取第六次国际耦合模式比较计划(Coupled Model Intercomparison Project Phase 6,CMIP6)中的28个模型结果,计算得到其多模式集合平均(multi-model ensemble,MME)模型结果;并基于多种评估指标,与再分析资料ERSST v5比较,评价了29个模型对北太平洋海表面温度(sea surface temperature,SST)的季节、年际和年代际变化的模拟能力,利用综合性指数(comprehensive rating metrics,MR)标准方法量化了模型结果的整体性能。结果表明,MME能够准确地模拟全年和四季变化,模型的偏差大部分集中在1~3月和7~9月,FGOALS-g3与NorCPM1的偏差高于其他模型。北太平洋SST在1915~2014年有明显的上升趋势,MME结果为每10年0.055℃,与再分析数据每10年0.059℃结果相近。两者相关系数为0.99、相对偏差为0.06、相对均方根误差为0.25。通过泰勒图的综合评估,不同的模型对同一的海域表现不同,对区域的模型选择不可随机。MR方法评价结果表明,有9个模型评分为A级,其中MME为A级中第三,证明了采用模型集合平均的方法可有效减少单个模型系统误差。The North Pacific Ocean is a sensitive region reflecting global climate change,especially having a significant impact on weather and climate change in East Asia,which is immediately adjacent to the Northwest Pacific Ocean.We selected 28 model results from the Coupled Model Intercomparison Project Phase 6(CMIP6)and computed their multi-model ensemble averaging(MME)model results,and evaluated the seasonal,inter-annual,and inter-decadal variability of the North Pacific sea surface temperature(SST)of 29 models based on a variety of assessment metrics and in comparison with reanalyzed information ERSST v5,using the integrated scoring criteria method.The overall performance of the model results was quantified using a system of comprehensive rating metrics(MR).Results show that the MME is able to accurately simulate year-round and seasonal variations,in which most of the model biases concentrated in January-March and July-September,and the biases of FGOALS-g3 and NorCPM1 are higher than those of other models.The North Pacific SST has a clear upward trend in 1915~2014,with an MME result of 0.055℃per decade,which is similar to the result of the reanalysis data(0.059℃per decade).The correlation coefficient of the two is 0.99,the relative deviation is 0.06,and the relative root-mean-square error is 0.25.Through the comprehensive assessment of the Taylor diagrams,different models perform differently for the same sea area,and the model selection of the region cannot be random.The evaluation results of the MR method show that nine models can be rated as Level A,and the MME rank the third place in Level A,which proved that the model ensemble averaging method could be used to effectively reduce the systematic errors of a single model.

关 键 词：第六次国际耦合模式比较计划 北太平洋 海表面温度 模式评估 

分 类 号：P732.6[天文地球—海洋科学]