论文部分内容阅读
利用惩罚最大T检验(Penalized MaximalTtest,PMT)方法,选取均一的邻近气象站为参考站,基于月平均地面气温(SAT)资料,利用相关系数权重平均方法构建参考序列,同时结合元数据信息,对1960-2011年中国沿海27个海洋观测站月平均海表温度(SST)进行了均一性检验与订正,并分析了造成海表温度序列非均一的主要原因.结果表明,中国沿海海洋台站海表温度资料存在较为严重的非均一性问题,几乎所有的台站都存在断点,仪器变更(包括人工观测转自动观测) (占总断点数的52.4%)和迁站(占总断点数的33.3%)是造成序列非均一的主要原因.整套资料负订正量所占比例较高,这种负订正量与人工转自动观测后海表温度观测值偏低有密切关系.这也使得订正后中国沿海平均海表温度趋势与订正前存在明显差异,订正后中国沿海海表温度呈明显的加速上升趋势.“,”The monthly sea surface temperature (SST)data of 27 marine observation stations have been detected and adj usted by the Penalized Maximal T test(PMT)and the detailed metadata archive.In this study,the homog-enous surface air temperature (SAT)data from neighboring meteorological observation stations are used to con-struct the reference series by correlation coefficient weighted averaged method.Results show that almost all of the stations along coast of the China Seas,the inhomogenities or change points can be detected.Combined with the de-tailed metadata of each station,the main reasons causing the inhomogenization problems are instrument change (including the change from artificial observation to automations)and relocation,accounting for about 52.4% and 33.3% of the total change points.It is also found that the annual mean SST trend has changed obviously and hom-ogeneity is improved well after adj ustment.The warming rate of SST series along the coast of the China Seas after adj ustment is much larger than that before adj ustment.This phenomenon links tightly to the SST decreasing after automation which leads to the large proportion of negative correction values.