NEX-GDDP数据具有分辨率高且其中所有模式分辨率统一的优点, 其对青藏高原地区极端气候的模拟能力如何, 鲜见报道。使用1986 -2005年青藏高原地面气象台站逐日观测资料, 选取了最能体现高原地区自然生态环境与社会经济活动受气候变化影响的10个极端气候指数: 霜冻日数(FD)、 结冰日数(ID)、 最低气温极小值(TNn)、 最高气温极大值(TXx)、 暖日持续日数(WSDI)、 冷日持续日数(CSDI)、 年降水量(PRCPTOT)、 连续无雨日数(CDD)、 连续有雨日数(CWD)和日最大降水量(RX1day), 从极端降水与极端温度两方面, 全面评估了NEX-GDDP中21个模式对青藏高原极端气候的模拟能力。结果表明: (1)TNn和TXx的模式模拟结果平均值小于观测值, 而其余8个极端气候指数的模式模拟平均值大于观测值。从变化趋势看, FD、 ID、 CSDI和CDD的观测值与模式模拟值变化趋势一致性强, 其余6个极端气候指数的一致性较弱。(2)21个模式对所选极端气候指数的空间模拟能力和时序模拟能力差异较大, 就相关系数而言, 多模式产品的空间模拟能力好于时序模拟能力, 就中心化均方根误差而言, 时序模拟能力又优于空间模拟能力。(3)按照极端气候指数识别所用数据不同将其分为三类: 日最低气温类极端气候指数(FD、 TNn和CSDI)、 日最高气温类极端气候指数(ID、 TXx和WSDI)和日降水量类极端气候指数(PRCPTOT、 CDD、 CWD和RX1day)。基于21个模式对极端气候指数时序与空间模拟能力的评估, 综合评选了三类极端气候指数的5个最优模式, 各自依次分别为: ①日最低气温类GFDL-ESM2G、 GFDL-CM3、 CCSM4、 MIROC5和ACCESS1-0; ②日最高气温类CanESM2、 BNU-ESM、 MIROC-ESM-CHEM、 inmcm4、 和CCSM4; ③日降水量类BNU-ESM、 CanESM2、 CSIRO-Mk3-6-0、 MIROC-ESM和MIROC-ESM-CHEM; 在利用NEX-GDDP研究青藏高原未来极端气候事件变化时, 建议将它们作为优选模式。
NEX-GDDP data has the advantages of high resolution and uniform resolution with all the models.However, there are few studies on evaluating the NEX-GDDP’s ability to simulate the extreme climate on the Qinghai-Xizang Plateau.Based on the daily observation dataset of meteorological stations for the period of 1986-2005 over the Qinghai-Xizang Plateau, this study selected ten extreme climate indices, i.e., Frost days (FD), Ice days (ID), Min Tmin (TNn), Max Tmax (TXx), Warm spell duration indicator (WSDI), Cold spell duration indicator (CSDI), annual Total wet-day precipitation (PRCPTOT), Consecutive dry days (CDD), Consecutive wet days (CWD) and Max 1-day precipitation amount (Rx1day), which can directly reflect the influence of climate change on social and economic activities and geographical landforms in the plateau area, and comprehensively evaluated the abilities of 21 models that participate the NASA Earth Exchange/Global Daily Down-scaled Projection(NEX-GDDP)in simulating extreme climate indices.The main conclusions are drawn as follows: (1) Except for the average values of TNn and TXx calculated by all models are lower than the average values calculated from observation dataset, the average values of other extreme climate indices calculated by all models are higher than the average values calculated by observation dataset.With respect to the variation trend of extreme climate indices, the FD, ID, CSDI and CDD trends calculated by observation shows strong consistency with those calculated by all models while weak consistency was seen for others extreme climate indices.(2) Large differences in spatial simulation ability can be seen among the 21 models, and in terms of correlation coefficient (r), the spatial simulation ability of all the models is better than that of time sequential simulation ability, while in terms of root mean square error (RMSE), the ability of sequential simulation ability is better than that of spatial simulation.(3) According to the data used to identify the extreme climate indices, the extreme climate indices is divided into three categories: daily minimum temperature category (FD, TNn and CSDI), daily maximum temperature category (ID, TXx and WSDI) and daily precipitation category (PRCPTOT, CDD, CWD and RX1day).Based on the 21 models’ abilities to simulate the spatio-temporal variations of the extreme climate indices, five optimal modes for three extreme climate indices were selected as follows: ① daily minimum temperature category: GFDL-ESM2G, GFDL-CM3, CCSM4, MIROC5 and ACCESS1-0.② daily maximum temperature category: CanESM2, BNU-ESM, MIROC-ESM-CHEM, inmcm4 and CCSM4.③ daily precipitation category: BNU-ESM, CanESM2, CSIRO-Mk3-6-0, MIROC-ESM and MIROC-ESM-CHEM.Based on the above statements, the above optimal models are recommended when the NEX-GDDP is used to investigate the extreme climate change over the Qinghai-Xizang Plateau in the future.
[1]Allen S K, Plattner G K, Nauels A, al et, change Climate2013: The physical science basis.An overview of the working group 1 contribution to the fifth assessment report of the Intergovernmental Panel on Climate Change (IPCC) [M].AGU Fall Meeting Abstracts.2013: GC51A-0949.
[2]Bao Y, Wen X Y, 2017.Projection of China's near-and long-term climate in a new high-resolution daily downscaled dataset NEX-GDDP[J].Journal of Meteorological Research, 31(1): 236-249.
[3]Chen H P, Sun J Q, 2015.Assessing model performance of climate extremes in China: an intercomparison between CMIP5 and CMIP3[J].Climatic Change, 129(1): 197-211.
[4]Chen H P, Sun J Q, Li H X, 2017.Future changes in precipitation extremes over China using the NEX-GDDP high-resolution daily downscaled data-set[J].Atmospheric and Oceanic Science Letters, 10(6): 403-410.
[5]Li W, Jiang Z H, Xu J J, al et, 2016.Extreme precipitation indices over China in CMIP5 Models.Part II: Probabilistic projection[J].Journal of Climate, 29(24): 8989-9004.
[6]Liu J, Shangguan D H, Liu S Y, al et, 2019.Evaluation and comparison of CHIRPS and MSWEP daily-precipitation products in the Qinghai-Tibet Plateau during the period of 1981-2015[J].Atmospheric Research, 230: 104634.
[7]Mishra S K, Jain S, Salunke P, al et, 2019.Past and future climate change over the Himalaya-Tibetan Highland: Inferences from APHRODITE and NEX-GDDP data[J].Climatic Change, 156(3): 315-322.
[8]Orlowsky B, Seneviratne S, 2012.Global changes in extreme events: regional and seasonal dimension[J].Climatic Change, 110(3): 669-696.
[9]Thrasher B, Xiong J, Wang W L, al et, 2013.Downscaled climate projections suitable for resource management[J].Eos, Transactions American Geophysical Union, 94(37): 321-323.
[10]Taylor K E, Stouffer R J, Meehl G A, al et, 2012.An overview of CMIP5 and the experiment design [J].Bulletin of the American Meteorological Society, 93(4): 485-498.
[11]Xu Y, Gao X J, Shen Y, al et, 2009.A daily temperature dataset over China and its application in validating a RCM simulation[J].Advances in Atmospheric Sciences, 26(4): 763-772.
[12]Zhang S J, Wang D H, Qin Z K, al et, 2018.Assessment of the GPM and TRMM precipitation products using the rain gauge network over the Tibetan Plateau[J].Journal of Meteorological Research, 32(2): 324-336.
[13]陈说, 叶涛, 刘苇航, 等, 2020.NEX-GDDP和CMIP5对青藏高原地区近地面气象场历史和未来模拟的评估与偏差校正[J/OL].高原气象, 1-15.DOI: 10.7522/j.issn.1000-0534.2020.00019. [2020-11-05]..
[14]程志刚, 刘晓东, 赵林, 2007.青藏高原未来气候变化及其对高原多年冻土分布的影响[C]//青藏高原资源·环境·生态建设学术研讨会暨中国青藏高原研究会2007学术年会论文摘要汇编.16-17.
[15]崔鹏, 贾洋, 苏凤环, 等, 2017.青藏高原自然灾害发育现状与未来关注的科学问题[J].中国科学院院刊, 32(9): 985-992.
[16]丁永建, 秦大河, 2009.冰冻圈变化与全球变暖: 我国面临的影响与挑战[J].中国基础科学, 11(3): 4-10.
[17]郭晓宁, 2010.青海高原近50a来雪灾特征研究 [D].兰州: 兰州大学.
[18]胡浩林, 2013.CMIP5模式集合对中国区域性低温事件的模拟及预估 [D].兰州: 兰州大学.
[19]蒋帅, 江志红, 李伟, 等, 2017.CMIP5模式对中国极端气温及其变化趋势的模拟评估[J].气候变化研究进展, 13(1): 11-24.
[20]李金洁, 王爱慧, 郭东林, 等, 2019.高分辨率统计降尺度数据集NEX-GDDP对中国极端温度指数模拟能力的评估[J].气象学报, 77(3): 579-593.
[21]李韧, 赵林, 丁永建, 等, 2012.青藏公路沿线多年冻土区活动层动态变化及区域差异特征[J].科学通报, 57(30): 2864-2871.
[22]马伟东, 2019.青藏高原极端降水特征及洪涝灾害临界雨量估算 [D].西宁: 青海师范大学.
[23]秦大河, 丁永建, 2009.冰冻圈变化及其影响研究——现状、 趋势及关键问题[J].气候变化研究进展, 5(4): 187-195.
[24]沈雨辰, 2014.CMIP5模式对中国极端气温指数模拟的评估及其未来预估 [D].南京: 南京信息工程大学.
[25]吴佳, 高学杰, 2013.一套格点化的中国区域逐日观测资料及与其他资料的对比[J].地球物理学报, 56(4): 1102-1111.
[26]王玉琦, 鲍艳, 南素兰, 2019.青藏高原未来气候变化的热动力成因分析[J].高原气象, 38(1): 29-41.DOI: 10.7522/j.issn. 1000-0534.2018.00066.
[27]肖林鸿, 高艳红, FEI C, 等, 2016.青藏高原极端气温的动力降尺度模拟[J].高原气象, 35(3): 574-89.DOI: 10.7522/j.issn. 1000-0534.2016.00039.
[28]杨志刚, 建军, 洪建昌, 2014.1961-2010年西藏极端降水事件时空分布特征[J].高原气象, 33(1): 37-42.DOI: 10.7522/j.issn. 1000-0534.2013.00147.
[29]姚檀栋, 陈发虎, 崔鹏, 等, 2017.从青藏高原到第三极和泛第三极[J].中国科学院院刊, 32(9): 924-31.
[30]姚遥, 罗勇, 黄建斌, 2012.8个CMIP5模式对中国极端气温的模拟和预估[J].气候变化研究进展, 8(4): 250-256.
[31]于海英, 许建初, 2009.气候变化对青藏高原植被影响研究综述[J].生态学杂志, 28(4): 747-54.
[32]于灏, 周筠珺, 李倩, 等, 2020.基于CMIP5模式对四川盆地湿季降水与极端降水的研究[J].高原气象, 39(1): 68-79.DOI: 10.7522/j.issn.1000-0534.2019.00007.
[33]赵志龙, 张镱锂, 刘峰贵, 等, 2013.青藏高原农牧区干旱灾害风险分析[J].山地学报, 31(6): 672-84.
[34]周莉, 兰明才, 蔡荣辉, 等, 2018.21世纪前期长江中下游流域极端降水预估及不确定性分析[J].气象学报, 76(1): 47-61.
[35]周天军, 邹立维, 陈晓龙, 2019.第六次国际耦合模式比较计划(CMIP6)评述[J].气候变化研究进展, 15(5): 445-456.
[36]钟水新, 2020.地形对降水的影响机理及预报方法研究进展[J].高原气象, 39(5): 1122-1132.DOI: 10.7522/j.issn.1000-0534. 2019.00083.