Mapping coal fires based on OMIS1 thermal infrared band image

Abstract

OMIS1 (Operative Modular Imaging Spectrometer) is a new imaging spectrometer designed by Shanghai Institute of Technical Physics, and has eight thermal infrared bands (nos. 105–112). This paper presents the use of images obtained from these thermal infrared bands to detect and map coal fires in north-western China. According to experiments in the Rujigou area of Ningxia municipality, each thermal infrared band has high correlation (>0.939). Regression analysis of land surface temperature (LST) with the pixel value in each thermal infrared band indicates: (1) for images acquired at daytime (from 11:00 to 13:00 local time), the first four thermal infrared bands have a linear relationship, and the final four bands have an exponential relationship; (2) for images acquired in the early morning (from 06:00 to 07:30), each band has a linear relationship. In conclusion, the ground temperature map is based on the 107th band of morning time. The mapping error is greatly determined by the quantity and precision of ground data measured synchronously.