高级检索

中国百年(1900~2022年)均一化降水量日值和月值数据集研制

Development of An In-Situ Homogenized Daily and Monthly Precipitation Dataset for China over 120 Years (1900–2022)

  • 摘要: 本项工作以中国气象局发布的我国60个重点城市早期(1950年前)降水观测记录以及1950年后我国2419个国家级气象观测台站逐日和逐月降水观测数据为数据源,应用PMF(Penalized Maximal F test)和PMT(Penalized Maximal T test)方法检测历史降水中的非均一断点以及QM(Quantile Match)方法开展偏差订正,消除了数据衔接、观测站迁址、设备更换等引起的降水历史数据系统偏差,研制完成一套时间长度达到百年(1900~2022年)的中国均一化降水量日值和月值数据集。数据集收录了共计2446条中国降水序列,起始时间早于1945年、1930年、1915年的序列数量分别为321条、126条、60条。通过非均一性检验,在107个观测站的降水数据中发现了113个非均一断点,它们不仅会导致降水数据均值偏离区域降水事实,也会引起降水数据标准差突变,干扰降水变率的研判。均一化校正能够较好地消除这些非气候变化引起的偏差,不再出现显著偏离局地降水均值和标准差的情况。最终建立起一套能够客观地描述我国过去百年降水气候变化事实,为气候变化研究提供支撑的中国百年降水数据产品。基于均一化的降水数据评估中国降水百年变化趋势,结果表明在过去近120年中国的降水并未呈现显著的变化趋势,1900~2022年中国降水距平百分率的变化速率为0.35%±0.21% (10 a)−1

     

    Abstract: This study establishes a precipitation dataset for China over the last 120 years. Pre-1950 precipitation records from 60 sites rescued by the China Meteorological Administration (CMA) and post-1950 in-situ measurements from 2419 national observatories are merged to obtain long time series since the 1900s. A total of 2446 daily and monthly precipitation time series are included in this dataset, including 321, 126, and 60 sequences with records before 1945, 1930, and 1915, respectively. The breaks caused by inhomogeneity issues in historical data are checked and removed via the PMF (Penalized Maximal F test)/PMT (Penalized Maximal T test), and QM (Quantile Match) methods, respectively. A total of 113 breakpoints from 107 stations are identified, which notably deviate the precipitation and seasonal variability from the regional climate state. Homogenizing adjustments eliminate these surges and slumps in historical data induced by means of nonclimatic factors. The homogenized data exhibit considerably higher spatial and temporal consistencies with neighboring measurements. Based on the homogenized data, the precipitation amounts over China exhibit slight changes over the last 120 years and the rate of change of the precipitation anomalies (in percentage) from 1900 to 2022 is approximately (0.35%±0.21)% (10 a)−1.

     

/

返回文章
返回