判断大气要素场相似度的相似面积比方法
|
摘要
目前常用的相似量一般是对样本差异的一个平均意义上的描述,在判断样本相似时难免受到个别过大或过小的样本因子的影响。针对这个问题,提出一种新的相似性度量方法——相似面积比方法,试图从全局的角度来考虑样本相似的选取问题。首先将原始数据标准化,根据对样本间距离的统计分析确定一个判断相似的阈值,然后计算两个样本中距离小于阈值的面积与总面积的比来判定两个样本的相似程度,最后利用2001 2011年冬季00:00(世界时)NCEP FNL温度、高度和风场的全球分析资料对该方法进行了验证。结果表明,数据标准化后可以得到值域范围大致相同的无量纲变量场,有利于多层次多变量的综合性相似判据的建立;与海明距离、相关系数相比,相似面积比选取的相似样本和基准样本在分布形态上较为接近,但相似面积比对高纬地区的相似判断较弱。基于上述三种相似性方法,利用1948 2013年NCEP月平均温度再分析资料进行长期天气预报试验表明,三种方法在相似样本的选取上有一定数量的重合率;从相似预报效果来看,相似面积比方法最好,海明距离方法次之,相关系数方法较差。
The common similarity method describes the difference of data samples in terms of an average. It can suffer from the effect of several extreme sample factors. For this reason,a new similarity method called similarity area ratio is developed to determine the similarity in a global view. In this method,the non-dimensional variables are generated by using a normalization procedure on atmospheric data. The threshold in the criterion of similarity is estimated from the analysis on the distance of samples. And the similarity degree is determined by the ratio of the area where the distance between sample factors is less than the threshold to total area of sample factors.The data including temperature,height,and wind at 00: 00 UTC in the winter from 2001 to 2011 is used to test the validity and efficiency of the similarity-area-ratio method. The results show the range of non-dimensional variables is consistent with each other. It is helpful to build a combined criterion of similarity that combines many atmospheric elements at all pressure levels. Different from the results from Hamming distance and correlation coefficient,the similarity sample selected by the similarity-area-ratio method is optimal. However,this similarity is not better than the other two similarity methods in high-latitude region. The long-term predictions based on these above three similarity methods are performed with monthly mean temperature data from 1948 to 2013 to further test their efficiency. The results show that the similar sample selected by these methods is same for some references. The frequency of the best forecasts with similarity-area-ratio method is higher than that with the other similarity methods.
The common similarity method describes the difference of data samples in terms of an average. It can suffer from the effect of several extreme sample factors. For this reason,a new similarity method called similarity area ratio is developed to determine the similarity in a global view. In this method,the non-dimensional variables are generated by using a normalization procedure on atmospheric data. The threshold in the criterion of similarity is estimated from the analysis on the distance of samples. And the similarity degree is determined by the ratio of the area where the distance between sample factors is less than the threshold to total area of sample factors.The data including temperature,height,and wind at 00: 00 UTC in the winter from 2001 to 2011 is used to test the validity and efficiency of the similarity-area-ratio method. The results show the range of non-dimensional variables is consistent with each other. It is helpful to build a combined criterion of similarity that combines many atmospheric elements at all pressure levels. Different from the results from Hamming distance and correlation coefficient,the similarity sample selected by the similarity-area-ratio method is optimal. However,this similarity is not better than the other two similarity methods in high-latitude region. The long-term predictions based on these above three similarity methods are performed with monthly mean temperature data from 1948 to 2013 to further test their efficiency. The results show that the similar sample selected by these methods is same for some references. The frequency of the best forecasts with similarity-area-ratio method is higher than that with the other similarity methods.
引文
[1]Keogh E,Chakrabarti K,Pazzani M,et al.Dimensionality reduction for fast similarity search in large time series databases[J].Knowledge and Information Systems,2000,3(3):263-286.
[2]刘跃虎,王飞,刘晓东,等.一种特征矩阵的相似性度量方法及其在图像检索中的应用[J].模式识别与人工智能,2006,19(4):497-502.
[3]李桂玲,王元珍,杨林权,等.基于SAX的时间序列相似性度量方法[J].计算机应用研究,2012,29(3):893-896.
[4]邱明.语义相似性度量及其在设计管理系统中的应用[D].杭州:浙江大学,2012.
[5]许炳南,周颖.贵州春季冰雹短期预报的高空温压场相似法[J].高原气象,2003,22(4):426-430.
[6]Barnett T P,Preisendorfer R W.Multifield analog prediction of short-term climate fluctuations using a climate state vector[J].J Atmos Sci,1978,35(10):1771-1787.
[7]Hamill T M,Whitaker J S.Probabilistic quantitative precipitation forecasts based on reforecast analogs:Theory and application[J].Mon Wea Rev,2006,134:3209-3229.
[8]张培群,李维京,丑纪范,等.基于多个参考态更新的动力相似预报方法及应用[J].物理学报,2006,55(8):4388-4396.
[9]郑志海,任宏利,黄建平.基于季节气候可预报分量的相似误差订正方法和数值试验[J].物理学报,2009,58(10):7359-7367.
[10]王启光,封国林,郑志海,等.长江中下游汛期降水优化多因子组合客观定量化预测研究[J].大气科学,2011,35(2):287-297.
[11]朱海峰,王式功,尚可政,等.切比雪夫判据与常用相似判据的比较及应用[J].高原气象,2013,32(3):848-855,doi:10.7522/j.issn.1000-0534.2012.00078.
[12]刘扬,韦志刚,李振朝,等.中国北方地区降水变化的分区研究[J].高原气象,2012,31(3):638-645.
[13]邱崇践,丑纪范.天气预报的相似—动力方法[J].大气科学,1989,13(1):22-28.
[14]覃志年,李维京,何慧,等.广西6月区域性暴雨过程的延伸预测试验[J].高原气象,2009,28(3):688-693.
[15]Huang J P,Yi Y H,Wang S W,Chou J F.An analogue-dynamical long-range numerical weather prediction system incorporating historical evolution[J].Quart J Roy Meteor Soc,1993,119:547-565.
[16]郑庆林,杜行远.使用多时刻观测资料的数值天气预报新模式[J].中国科学(A辑),1973,16(3):289-297.
[17]陈伯民,纪立人,杨培才,等.改善月动力延伸预报水平的一种新途径[J].科学通报,2003,48(5):513-520.
[18]丑纪范,任宏利.数值天气预报——另类途径的必要性和可行性[J].应用气象学报,2006,17(2):240-244.
[19]李开乐.相似离度及其使用技术[J].气象学报,1986,44(2):174-183.
[20]阎慧芳,李社宗,黄跃清,等.常用相似性判据的检验和综合相似系数的使用[J].气象科技,2003,31(4):211-215.
[21]万日金,何溪澄,林刚.用动力相似方法预报广东省区域暴雨预报试验[J].热带气象学报,2006,22(2):198-202.
[22]陈磊,翟宇梅,王力维.K近邻非参数回归技术常用相似性测度的对比分析[J].军事气象水文,2010,2:24-27.
[23]黄建平,郜吉东,丑纪范.北半球月平均环流异常演变的相似韵律现象[J].高原气象,1990,9(1):88-92.
[24]罗阳.一种新的相似性度量——高分辨率相似系数[J].空军气象学院学报,1996,17(1):23-32.
[25]钟元.天气与短期气候的多元判据综合相似预测方法[M].第1版.北京:气象出版社,2003年11月.
[26]钱维宏.天气尺度瞬变扰动的物理分解原理[J].地球物理学报,2012,55(5):1439-1448,doi:10.6038/j.issn.0001-5733.2012.05.002.
[2]刘跃虎,王飞,刘晓东,等.一种特征矩阵的相似性度量方法及其在图像检索中的应用[J].模式识别与人工智能,2006,19(4):497-502.
[3]李桂玲,王元珍,杨林权,等.基于SAX的时间序列相似性度量方法[J].计算机应用研究,2012,29(3):893-896.
[4]邱明.语义相似性度量及其在设计管理系统中的应用[D].杭州:浙江大学,2012.
[5]许炳南,周颖.贵州春季冰雹短期预报的高空温压场相似法[J].高原气象,2003,22(4):426-430.
[6]Barnett T P,Preisendorfer R W.Multifield analog prediction of short-term climate fluctuations using a climate state vector[J].J Atmos Sci,1978,35(10):1771-1787.
[7]Hamill T M,Whitaker J S.Probabilistic quantitative precipitation forecasts based on reforecast analogs:Theory and application[J].Mon Wea Rev,2006,134:3209-3229.
[8]张培群,李维京,丑纪范,等.基于多个参考态更新的动力相似预报方法及应用[J].物理学报,2006,55(8):4388-4396.
[9]郑志海,任宏利,黄建平.基于季节气候可预报分量的相似误差订正方法和数值试验[J].物理学报,2009,58(10):7359-7367.
[10]王启光,封国林,郑志海,等.长江中下游汛期降水优化多因子组合客观定量化预测研究[J].大气科学,2011,35(2):287-297.
[11]朱海峰,王式功,尚可政,等.切比雪夫判据与常用相似判据的比较及应用[J].高原气象,2013,32(3):848-855,doi:10.7522/j.issn.1000-0534.2012.00078.
[12]刘扬,韦志刚,李振朝,等.中国北方地区降水变化的分区研究[J].高原气象,2012,31(3):638-645.
[13]邱崇践,丑纪范.天气预报的相似—动力方法[J].大气科学,1989,13(1):22-28.
[14]覃志年,李维京,何慧,等.广西6月区域性暴雨过程的延伸预测试验[J].高原气象,2009,28(3):688-693.
[15]Huang J P,Yi Y H,Wang S W,Chou J F.An analogue-dynamical long-range numerical weather prediction system incorporating historical evolution[J].Quart J Roy Meteor Soc,1993,119:547-565.
[16]郑庆林,杜行远.使用多时刻观测资料的数值天气预报新模式[J].中国科学(A辑),1973,16(3):289-297.
[17]陈伯民,纪立人,杨培才,等.改善月动力延伸预报水平的一种新途径[J].科学通报,2003,48(5):513-520.
[18]丑纪范,任宏利.数值天气预报——另类途径的必要性和可行性[J].应用气象学报,2006,17(2):240-244.
[19]李开乐.相似离度及其使用技术[J].气象学报,1986,44(2):174-183.
[20]阎慧芳,李社宗,黄跃清,等.常用相似性判据的检验和综合相似系数的使用[J].气象科技,2003,31(4):211-215.
[21]万日金,何溪澄,林刚.用动力相似方法预报广东省区域暴雨预报试验[J].热带气象学报,2006,22(2):198-202.
[22]陈磊,翟宇梅,王力维.K近邻非参数回归技术常用相似性测度的对比分析[J].军事气象水文,2010,2:24-27.
[23]黄建平,郜吉东,丑纪范.北半球月平均环流异常演变的相似韵律现象[J].高原气象,1990,9(1):88-92.
[24]罗阳.一种新的相似性度量——高分辨率相似系数[J].空军气象学院学报,1996,17(1):23-32.
[25]钟元.天气与短期气候的多元判据综合相似预测方法[M].第1版.北京:气象出版社,2003年11月.
[26]钱维宏.天气尺度瞬变扰动的物理分解原理[J].地球物理学报,2012,55(5):1439-1448,doi:10.6038/j.issn.0001-5733.2012.05.002.