【作者】 薛志磊；

【导师】 张书余；

【作者基本信息】 兰州大学 ， 大气物理与大气环境， 2013， 硕士

【摘要】 根据甘肃省81个地面气象站1982-2011年逐日极端气温资料,分析了甘肃省日极端气温的时空分布特征和突变状况,同时初步探讨了极端气温产生突变的原因。另外,选取甘肃省东南部、西北部两个地区,分别分析在这两个地区WRF模式极端气温模拟对三种陆面参数化方案的敏感性,试验模拟时段为2005年1月25日08时到29日02时(北京时),共15个6小时气温预报。所做工作得到如下主要结论：(1)EOF分析表明,甘肃省夏季极端高温频数、冬季极端低温频数和冬季极端高温频数有着非常高的一致性变化特征,其中夏季和冬季的极端高温频数EOF第一特征向量的方差贡献率达到70%以上,二者的时间系数变化也呈现明显的上升趋势；而冬季极端低温频数方差贡献率为59.3%,一致性特征小于极端高温频数,其时间系数变化呈波动上升趋势。甘肃省夏季日极端高温在80年代末期出现明显突变,冬季日极端低温和日极端高温突变现象不明显,夏季极端高温突变与西风环流指数呈正相关关系。(2)WRF模式三种陆面方案在不同地区对气温的模拟都有比较好的一致性,对气温的数值范围和分布形式也都有合理统一的模拟效果。整体上,三种方案对甘肃东南地区气温的模拟一致好于甘肃西北地区气温的模拟效果,对14时气温模拟一致好于对02时气温的模拟效果。对于西北地区,NOAH方案表现最好,RUC方案次之,SLAB方案再次之；对于东南地区,SLAB方案最优,RUC方案次之,NOAH方案再次之。(3)三种陆面方案对14时气温模拟的差异主要来自于不同陆面方案对地表温度和感热、潜热通量模拟的不同,对地表温度的模拟基本上决定了气温模拟的走势,感热通量的模拟亦是影响气温模拟的重要因素。更多还原

【Abstract】 Based on the daily extreme temperature (ET) data observed by81stations in GanSu Province from1982to2011, the spatio-temporal evolution characteristics of extreme air temperature were analyzed. The moving cut data-approximate entropy(MC-ApEn) and the moving t test to be used to detect the climate abrupt change of the high-ET in summer and low-ET in winter, as well as discussing the probable reasons of it. At the same time three land surface schemes(SLAB,NOAH and RUC)in WRF model are used to simulate2m temperature of the Southeast and Northwest in Gansu.The results shows that:(1)The first EOF eigenvector of the frequency of the high-ET in summer and winter shows the consistent change all over Qansu, and over70%of total variance is explained. The time coefficient indicates that the frequency of extreme maximum temperature increases obviously;The first EOF eigenvector of the frequency of low-ET in winter explains59.3%of total variance, which also shows consistent change.The time coefficient indicates that the extreme minimum temperature frequency increases. The high-ET in summer have climate abrupt change and it occurred at the end of1980s.At the same time We can proved the change of zonal index is the primary cause of abrupt change in high-ET.(2)The simulated2m temperature generally is sensitive to different land surface schemes and agree with the scope and the number of observations in different regions.Generally, three schemes are more reasonable in simulating the High-ET than simulating he Low-ET,and more reasonable in Southeast of Gansu.SLAB scheme is better than the other land surface scheme in the Southeast, NOAH scheme presents best in northwest of Gansu,and there are no more different in simulating the high-ET at daytime and the low-ET at nighttime.(3)The reasons of the model performance difference with difference Land surface scheme is the prediction of meteorological variables--surface skin temperature, upward heat flux and latent heat flue at the surface.The distributed of surface skin temperature decided the tendency of2m temperature,the value of upward heat flux is a very important factor to decide the temperature.更多还原