精细化要素客观预报中的站点差异性研究

以山东省乡镇级精细化要素预报为例，预报站点由县级123个增至1 561个。在精细化要素客观预报方法开发中，如何处理多量预报站（格）点问题？利用山东省稠密的区域自动站观测资料，通过分析不同站点空间分辨和地形属性的观测要素的统计学差异，研究各种气象要素的站点差异性。结果表明：温度、降水等要素具有明显的局地小气候特征，需要更高空间分辨率的预报站点来捕捉；市区、郊区以及不同高度山区等站点间最低气温各分位数统计值差异均较大，但相关系数较高，且频率分布形态相似；风对地形更敏感，不同测站间相关系数较低，分位数统计值和频率分布形态差异均较大。上述结果可为精细化要素客观预报业务中的站（格）点处理提供理论和技术支撑。

Study on differences between stations in fined objective meteorological element forecast

Taking fined meteorological element forecast at township level in Shandong Province as an example, the number of forecasting stations at county level has increased from 123 to 1 561. In the development of fined objective meteorological element forecast, it remains a problem how to deal with the numerous forecasting stations. Based on the observation data of dense regional automatic stations in Shandong, the differences between stations in meteorological element forecast are studied by analyzing the statistical differences of observation elements by different spatial resolutions and different topographic attributes at varied stations. The results show that elements such as temperature and precipitation have obvious local microclimate characteristics, which need forecasting stations with higher spatial resolution to capture; the quantile statistics of the minimum air temperature vary greatly among urban stations, suburban stations, and mountain stations at different altitudes, but the correlation coefficients are relatively high with similar frequency distribution patterns; the wind is more sensitive to the topography, the correlation coefficients between different stations are lower, and the quantile statistics and frequency distribution patterns are quite different. The above results can provide theoretical and technical support in fined objective meteorological element forecast between stations (grids).