Coal fire is a commonly existed disaster in coal fields of northen China

it burnt huge amount of coal resource. Firefighting is a long-term and tough job.Thermal infrared images of TM band 6 show us temperature information

which depends on solar radiance and ground heat. Based on remote sensing principle

this paper focuses on using DTM to improve the accuracy of detecting coal fires with thermal infrared images. On the resultant image thermal anomaly caused by coal fire is much clear

this is helpful for firefighting and dynamic monitoring. Junnan coal field was selected as a test area. Several pictures were displayed for the result.