This paper studies the effects of relative humidity, temperature, sulphur dioxide, and chlorides on the short-term corrosion behavior in the dynamic environment. A multi-parameter method is developed to characterize the statistical distributions of the environmental factors with high accuracy. The results suggest that TOW (time-of-wetness) should be replaced by temperature and relative humidity distributions. A corrosion model which is combined with physical and empirical knowledge of corrosion is presented and gives more accurate corrosion estimation than using the mean values of the environmental factors and fitting them independently. It is also demonstrated that relative humidity is the most influential factor on corrosion and temperature is secondary. The nonlinearity of their accelerating effects on corrosion rate are remarkable and should be considered in the daily dynamic environment. Sulphur dioxide and chlorides are important accelerating variables and their nonlinear accelerating effects are less significant.