CMIP6年代际试验对中国气温预测能力的初步评估  

Preliminary assessment on CMIP6 decadal prediction ability of air temperature over China

作　　者：汤秭晨 李清泉[1,2] 王黎娟[1] 伍丽泉 TANG Zi-Chen;LI Qing-Quan;WANG Li-Juan;WU Li-Quan(Key Laboratory of Meteorological Disaster,Ministry of Education/Joint International Research Laboratory of Climate and Environment Change/Collaborative Innovation Center on Forecast and Evaluation of Meteorological Disasters,Nanjing University of Information Science&Technology(NUIST),Nanjing 210044,China;Laboratory for Climate Studies,National Climate Center,China Meteorological Administration,Beijing 100081,China;Guangxi Climate Center,Nanning 530022,China)

机构地区：[1]南京信息工程大学气象灾害教育部重点实验室/气候与环境变化国际合作联合实验室/气象灾害预报预警与评估协同创新中心,南京210044 [2]国家气候中心中国气象局气候研究开放实验室,北京100081 [3]广西壮族自治区气候中心,南宁530022

出　　处：《气候变化研究进展》2021年第2期162-174,共13页Climate Change Research

基　　金：国家重点基础研究发展计划(2016YFA0602200);国家“第二次青藏高原综合科学考察研究”(2019QZKK0208);中国科学院战略性先导科技专项(XDA20100304);国家自然科学基金重大项目(41790471);国家自然科学基金青年基金项目(41376030)。

摘　　要：利用参加第六次国际耦合模式比较计划(CMIP6)年代际气候预测计划(DCPP)的加拿大CanESM5模式和日本MIROC6模式的结果,评估了模式对中国近地面气温的预测能力。在年代际尺度上,两个模式年代际试验对近地面气温的回报技巧均高于历史试验的模拟能力,采用海温初始化可以提高模式对中国近地面气温的年代际预报技巧。对年代际回报试验的进一步分析表明,两个模式均能较好地预测年平均气温的变化;对季节平均气温,模式在秋季的回报技巧最高,在冬季较低。就区域平均气温而言,两个模式对中国各个地区年平均和季节平均气温都有较高的回报技巧,其中我国南方和西部地区的气温回报技巧比北方高。年平均以及春季、冬季的气温年代际回报技巧总体随提前时间的增加而降低,夏季和秋季的气温回报技巧随提前时间的增加提高。各区域内年代际预测技巧随提前时间的变化特征与全国整体基本一致。The prediction ability of air temperature over China is evaluated in this paper based on the outputs of the Canadian CanESM5 model and the Japanese MIROC6 model participating in the Decadal Climate Prediction Project(DCPP) of the sixth Coupled Model Intercomparison Project(CMIP6). Comparing decadal prediction with historical simulation, both models’ decadal hindcasts show higher prediction skill for surface air temperature(SAT), which proves oceanic initialization improves the prediction skill of SAT in China on decadal scale. The models can capture the variation of annual mean temperature, and the prediction skill of seasonal mean temperature is the highest in autumn and comparatively lower in winter. Although both models have good performances in predicting annual and seasonal mean temperatures in various regions of China, the prediction skills are higher in the southern and western China than the northern China. As the lead time increases, the prediction skills of annual, spring and winter mean temperatures decrease, while those of summer and autumn mean temperature increase. The prediction skills of subregions share the same characteristics with the whole country.

关 键 词：CMIP6 气候系统模式 年代际预测 中国近地面气温 

分 类 号：P423[天文地球—大气科学及气象学]