Abstract: Based on datasets of Zhoushan Bridge over the sea,including air temperature and road-surface temperature,the features of maximum air temperature and maximum road-surface temperature in summer were discussed after the data undergone the quality control.According to different types of underlying surface,the statistic models of road-surface temperature were established using sunshine duration dataset of Zhoushan and a maximum air temperature of Zhoushan Bridge over the sea.The results show that the maximum air temperature decreases and then increases weakly from west to east,the transition zone appears on the Cezi island.Most of the stations' maximum road-surface temperature are directly related to the underlying surface type.The stations can be divided into two categories,i.e. over the sea and over the land.The maximum road-surface temperature at stations over the sea category is much higher than that at stations over the land category.The statistic models of road-surface temperature reflect that sunshine duration item contributes a larger portion at stations over the sea category than that at stations over the land.According to the evaluation results of the statistic models,it shows that models perform poorly,partly due to the flaw of the linear regression method.Meanwhile,the quality control approach adopted in this paper can sensitively reflect the wrong station data.At a certain extent,it avoids the disturbance of the wrong data on the results.

Key words: Zhoushan Bridge over the sea, Data quality control, Road-surface temperature, Statistic models