地理科学进展 ›› 2010, Vol. 29 ›› Issue (7): 818-826.doi: 10.11820/dlkxjz.2010.07.007

• 气候与水文过程 • 上一篇    下一篇

IPCC-AR4全球气候模式在华东区域气候变化的 预估能力评价与不确定性分析

顾问1,2, 陈葆德2, 杨玉华2, 董广涛3   

  1. 1. 南京信息工程大学,南京 210044|
    2. 中国气象局上海台风研究所/台风预报技术重点开放试验室,上海200030|
    3. 上海市气候中心,上海 200030
  • 收稿日期:2009-09-01 修回日期:2010-03-01 出版日期:2010-07-25 发布日期:2010-07-25
  • 通讯作者: 陈葆德(1964-),男,研究员,博导。研究领域:气候变化,数值模拟。E-mail:baode@mail.typhoon.gov.cn. E-mail:baode@mail.typhoon.gov.cn
  • 作者简介:顾问(1985-),女,硕士。研究领域:气候变化。 E-mail:guw@mail.typhoon.gov.cn.
  • 基金资助:

    中国气象局气候变化专项(CCSF-09-10).

Simulation Evaluation and Uncertainty Analysis for Climate Change Projections in East China Made by IPCC-AR4 Models

GU Wen1,2, CHEN Baode2, YANG Yuhua2, DONG Guangtao3   

  1. 1. Nanjing University of Information Science and Technology, Nanjing 210044, China|
    2. Shanghai Typhoon Institute/Key Open Laboratory of Typhoon Forecast Technique, CMA, Shanghai 200030, China|
    3. Shanghai Climate Center, Shanghai 200030, China
  • Received:2009-09-01 Revised:2010-03-01 Online:2010-07-25 Published:2010-07-25

摘要:

本文分析了IPCC第四次评估报告中的全球气候模式在华东区域的气候预估能力与不确定性。以均方根误差作为衡量预估能力的标准,比较了IPCC-AR4中21个气候模式在中等排放情景下的预估能力,结果表明气候模式对华东区域气候变化的模拟能力差异较大,模式NCAR-CCSM3和MRI_CGCM2_3.2在年平均气温和年降水2个要素的均方根误差均较小,说明它们对华东区域的气候预估能力比其他模式强。在中等排放情景下,气候模式能够模拟出接近观测实况的年平均气温与降水的空间分布特征,但是由于空间分辨率较低,模式不能模拟出局部细致的结构;多模式集合平均对华东区域气温预估存在明显系统偏差,比观测实况偏低1.6℃以上,偏低幅度超过了不确定性(一倍的模式间标准偏差)能涵盖的范围;华东区域年降水模式间标准偏差占模式集合平均百分比为26.7%,表明直接用AR4多模式集合平均的结果难以准确反映华东区域该要素的未来变化。

关键词: 不确定性, 华东区域, 气候模式, 情景预估

Abstract:

In this paper, the climate change projections in East China made by the climate models in the IPCC-AR4 were assessed by simulation evaluation and uncertainty analysis. By comparing individual simulation of the 21 IPCC AR4 models with the observations and with each other as well, it is demonstrated that the simulation abilities of different models vary widely. Only models of NCAR-CCSM3 and MRI-CGCM2.3.2 have small root mean square errors for both temperature and precipitation simulations in East China. It has been shown that, under the scenario A1B, multi-model ensemble mean can fairly well illustrate the spatial patterns of annual mean temperature and precipitation. Nevertheless, it can hardly reflect local fine structure of the distributions because of their low spatial resolutions. Moreover, there is a significant systematic deviation in multi-model ensemble mean for temperature projection. It underestimates the annual mean temperature by more than 1.6℃, and the difference between simulation and observation exceeds the extent of the uncertainty which is defined by one-fold standard deviation of inter-models’ simulations. The standard deviation of annual precipitation is up to 26.7% of the multi-model ensemble mean. Therefore it would be quite questionable if IPCC-AR4 multi-model ensemble means of temperature and precipitation in East China are directly used as climate change projection.

Key words: climate models, East China, scenarios projection, uncertainty