北京1841年以来均一化最高和最低气温日值序列的构建

New series of daily maximum and minimum temperature observations for Beijing, China since 1841

  • 摘要: 长期连续的日值观测资料是研究百年来极端气候事件及其变化特征的重要基础支撑。目前中外由于缺乏可靠的逐日百年尺度气候资料,使得20世纪50年代以前的极端气候变化规律仍然没有得到很好的认识。基于国家气象信息中心收集整理的日最高和最低气温观测资料,构建北京1841—2019年气温日值序列。首先,通过数据质量控制剔除原始基础资料中因人工观测、仪器故障及数字化过程造成的错误数据;其次,采用标准化序列法对北京气象观测站1951年以前的日最高和最低气温序列进行了插补;最后,利用惩罚最大t检验(PMT)和惩罚最大F检验(PMFT)对插补后的日值序列进行均一性检验,并采用分位数匹配法(QM)对检验得到的因迁站和仪器变更导致的统计显著断点进行订正,由此建立了北京百年均一化气温日值序列。分析表明,构建的北京百年逐日气温序列的年代变化特点与美国伯克利地球研发中心的月平均地表温度数据(Berkeley Earth-Monthly)、英国东英格利亚大学气候研究中心的全球月平均地表温度数据(CRUTS4.03)以及美国国家气候资料中心的全球历史气候数据(GHCNV3)具有一致性。从趋势变化来看,基于构建的逐日最高和最低气温序列统计得到1911—2019年北京地区年平均气温升高趋势为0.154±0.018℃/(10 a)(95%不确定性水平估计),与中国区域气温趋势变化速度基本一致。从而在一定程度上说明构建的北京百年尺度日最高和最低气温序列能够为极端气候变化领域提供新的基础数据。

     

    Abstract: Continuous daily observations over as long a period as possible are an important basis for the study of extreme climate events and their characteristics over the past hundred years. Due to the scarcity of reliable and continuous daily observations over century-long scale, the pattern of extreme climate change before the 1950s still can not be well interpreted both in China and abroad. New series of daily maximum and minimum temperature observations for Beijing from 1 January 1841 (for minimum) and 1 January 1880 (for maximum) to 31 December 2019 have been constructed. Two daily observational datasets are obtained from China Meteorological Administration (CMA) as the basic sources: (1) The digital daily observations covering the period of 1 January 1841—31 December 1950 from 5 in-situ sites and (2) the daily data observed by new weather and climate systems over the period of 1 January 1951—31 December 2019. Errors caused by manual observations, instrument malfunctions and digital inputs are removed by quality control. Those missing daily maximum and minimum temperature data before 1951 at Beijing meteorological observation station are filled by interpolations using standardized series method. Several significant breakpoints caused by changes in station locations and instrument manufacturers are detected by penalized maximal t test (PMT) and penalized maximal F test (PMFT). The homogenized daily maximum and minimum temperature series for Beijing are obtained by Quantile Matching (QM) adjustment with reference series from Berkeley Earth-daily data. It is found that the decadal variabilities of the new daily temperature series over century-long scale are consistent with those of Berkeley Earth-monthly, CRUTS4.03 and GHCNV3 data in Beijing. The warming trend estimated based on the annual mean temperature series averaged from the new daily maximum and minimum temperatures is 0.154±0.018°C/(10 a) (at the 95% confidence level for statistical uncertainties) during the period of 1911—2019 in Beijing, which is similar to that over China. Therefore, the time series of the homogenized daily maximum and minimum temperatures over century-long scale in Beijing city can to a certain extent provide a new data source for the study of climate extremes.

     

/

返回文章
返回