Taking villages in the northern plain of Shaoxing County Zhejiang Province as the research object and by using heat transfer model and remote sensing image analysis method and taking advantage of surface temperature information varying in different areas recorded by ETM thermal infrared band and through selection of special endmember, we realize the separation of background and ambient superposed brightness temperature and establish statistical model on change of superimposed environmental brightness temperature based on distance and analyze characteristics of rural environment temperature field according to the features of heat exchange type. Study shows that endmember brightness temperatures of different surface features in the study area are respectively as follows: hard surface is 304.663K, water body is 297.851K, grassland is 298.966K, woodland is 298.827K; superimposed environmental temperature in village area is about 1.737K. Environment superposed brightness temperature and distance function are tools to describe the temperature field, predicting pixel brightness temperature by using the heat transfer model is more accurate than using linear spectrum mixed model.