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).