Abstract: Based on the 1:250000 Digital Elevation Model (DEM) data and statistical data of the air temperature of 67 weather stations, considering the distance to coastline in Fujian Province considering the feedback effect of ocean to continent, the geographic mathematical model is established depending on the connections between factors such as the lowest temperature, latitude, longitude, altitude, and is used to simulate the fine distribution of lowest temperature in cold air processes of winter from 2008 to 2010.On the basis of the ascertainment of coastline and proven distributions, the appropriating calculation formula of the distance to coastline is ensured and the three monitoring models of low temperature processes are founded based on the choice of different distance to coastline or no distance to coastline. Moreover, the models are analyzed comparably and the best model is applied to simulate the low temperature. Contemporarily, the method of selecting appropriate the distance to coastline is approved by regression models and integrative residual sum of squares, and the transacting process of simulated errors in the joint of inter and outer coastline is introduced.The results show that the lowest temperature is well simulated by introducing the appropriate distance to coastline to the low temperature monitoring model during the cold air processes. With the increase of average cooling range of cold air, the efficiency of distance to the coastline factor to the value of the minimum temperature simulation decreases. Moreover, the distances to the coastline are changed with the different cold air processes and are not more than 50 kilometers. Furthermore, the method of how to select appropriate distance to the coastline is confirmed based on the value of different square sum. Although there is adjusting effect of ocean temperature to land temperature, with the increase of distance to coastline, the feedback effect on temperature of ocean to land decreases, and the mathematical model made up of factors such as longitude, latitude, altitude and the distance to coastline is suitable for low temperature monitoring simulation in regions where the distance to the coastline are more than 50 kilometers. Similarly, the mathematical model made up of factors as longitude, latitude, altitude and the distance to coastline is suitable for low temperature monitoring simulation in regions of the distance less more than 50 kilometers, which could increase the precision of low temperature monitoring simulation and embody the adjustment function of sea to land temperature. Finally, 9 destined samples (each sample is selected optionally from one city of Fujian) are verified in the model and the simulated results of low temperature are proved to match with actual situation substantially.