Detecting coal fires with a neural network to reduce the effect of solar radiation on Landsat Thematic Mapper thermal infrared images

Coal fires in the north of China have already resulted in serious problems, including huge losses in coal resources, air pollution and so on. Thermal infrared images by Landsat Thematic Mapper (TM) can be used to detect some thermal anomalies. However, an initial necessity is to reduce the effect of solar radiation on TM thermal infrared images. In this paper, a neural net work is used to set up a mathematical model of ground temperature for the f irst time. After the neural network completes training, we can use it to ca lculate the ground temperature caused by solar radiation. Thus, the result can be used to reduce the effect of solar radiation on TM thermal infrared images, and extract the thermal anomalies caused by coal fires.