基于不足气象要素时间序列的小时界限值提取

气象要素质量控制中，往往是以全球或全国历史极值、气象站历史极值或季节性极值作为界限值检查标准，该方法在时间尺度上具有广泛性和粗糙性。为了使气象站气象要素界限值检查标准具有唯一性和精细性，将时间尺度缩至以小时为单位，利用全国1942个气象站2019年和2020年每小时气象数据，基于Python语言实现了一套确定每小时界限值的概率统计算法。由于时间尺度小，所以在相邻时刻，该界限值确定方法适用于具有规律性的气象要素（气温、气压和相对湿度）。经过样本统计，气温、气压和相对湿度的每小时界限值与其实际变化规律一致，证明了该算法的有效性，为后续质量控制界限值检查步骤的实施打下坚实基础。

Extraction of Hourly Extreme Values Based on Insufficient Time Series in Meteorological Elements

Global or national extreme values, historical extreme values of weather stations or seasonal extreme values are usually used as the extreme value check standards for the quality control of meteorological data is extensive and rough on time scale. In order to make the extreme value check standards of weather stations unique and fine, based on the Python language, a set of probabilities statistical algorithms to evaluate hourly extreme values is implemented by reducing the time scale to hours and using hourly meteorological data from 1942 weather stations around the China in 2019 and 2020. Because of small time scale, the method to evaluate hourly extreme values is suitable for regular meteorological elements (air temperature, pressure and relative humidity). After sampling statistics, variation of hourly extreme values in air temperature, air pressure and relative humidity are consistent with their actual variation. The algorithm has been shown to be efficient, which provides a solid basis for the subsequent check of extreme values in quality control.