Abstract:By analyzing the synchronously observed temperature data (January, April, July 2011)from the new and old meteorological observation stations in Shifang, it is found out that temperature in the old station was higher than that in the new station because of urban heat island effect, which was more obvious in the daytime than at night, more obvious in April of spring, and in January of winter than in July of summer. The urban heat island effect increased maximum temperature and the diurnal temperature range, made the old station temperature higher than the new station. The analysis shows that temperature in the old station had a positive correlation with that in the new station; meanwhile there was obvious difference between temperatures in the old and new stations. Though a linear regression equation correction make the data with representative, continuity and comparability.