摘要
气象变点检测的准确性对于气象数据的均一化处理尤为重要,然而气候因素以及非气候因素等原因给气象数据变点检测带来了诸多不便。文章基于面板数据分析方法对气象数据的结构性突变进行检测和估计,区分了造成气象数据发生非均一性变化的非气候因素变点;实证分析了安徽省1957—2015年5个气象台站的逐月平均气温序列的变点问题,并结合现有的元数据资料探究了各个台站变点发生的主要原因。
The accuracy of change-point detection is vital to the climate homogenization procedure while it also has many problems for climate and non-climate factors. In this paper, panel data analysis model is developed to estimate the position of change-point and non-climate factors change-point for the inhomogeneity in meteorological data is distinguished. Then the problem of change-point in monthly mean temperature series is empirically analyzed according to the five meteorological stations of Anhui Province from 1957 to 2015 and the major cause of change-points is also explored with the metadata records.
引文
[1] LI Z,YAN Z W,TU K,et al.Changes in wind speed and extremes in Beijing during 1960-2008 based on homogenized observations [J].Advances in Atmospheric Sciences,2011,28(2):408-420.
[2] 梁小文,杨梅学,万国宁,等.青藏高原气温序列的均一性研究[J].冰川冻土,2015,37(2):275-285.
[3] WANG X L,WEN Q H,WU Y H.Penalized maximal t-test for detecting undocumented mean change in climate data series[J].Journal of Applied Meteorology and Climatology,2007,46(6):916-931.
[4] BUISHAND T A.Some methods for testing the homogeneity of rainfall records[J].Journal of Hydrology,1982,58:11-27.
[5] ALEXANDERSSON H,MOBERG A.Homogenization of Swedish temperature data:part 1:homogeneity test for linear trends [J].International Journal of Climatology,1997,17:25-34.
[6] 吴利红,康丽莉,陈海燕,等.地面气象站环境变化对气温序列均一性影响[J].气象科技,2007,35(1):152-155.
[7] ZHAO X,CHU P.Bayesian change point analysis for extreme events (typhoons,heavy rainfall,and heat waves):an RJMCMC approach[J].Journal of Climate,2010,23(5):2416-2436.
[8] LI Y B,LUND R.Multiple changepoint detection using metadata[J].Journal of Climate,2014,28(10):4199-4218.
[9] 朱慧明,黄仁存,游万海.贝叶斯隐马尔可夫异质面板模型及应用研究[J].统计研究,2014,31(7):97-104.
[10] LIANG F M.Annealing stochastic approximation Monte Carlo algorithm for neutral network training[J].Mach Learn,2007,68:201-233.
[11] KIM J,CHEON S.Bayesian multiple change-point estimation with annealing stochastic approximation Monte Carlo [J].Computational Statistics,2010,25:215-239.
[12] HORVATH L,HUSKOVA M.Change-point detection in panel data[J].Journal of Time Series Analysis,2012,33(4):631-648.
[13] ALFRED K,ARNUT P.Shapiro-Francia test compared to other normality test using expected p-value[J].Journal of Statistical Computation and Simulation,2015,85(15):3002-3016.
[14] 周建平,孙照渤,倪东鸿,等.中国气象台站迁移对年平均气温均一性的影响[J].大气科学学报,2013,36(2):139-146.
[15] 吴增祥.气象台站历史沿革信息及其对观测资料序列均一性影响的初步分析[J].应用气象学报,2005,16(4):461-467.
[16] 宋超辉,刘小宁,李集明.气温序列非均一性检验方法的研究[J].应用气象学报,1995,6(3):289-296.