Journal of Earth Science & Climatic Change
Open Access

Abstract

Gas and particulate matter which result from straw burning can seriously pollute the atmospheric environment, threatening human health and traffic safety. Straw field distribution provides a basis for Remote Sensing Monitoring of straw burning and estimation of false fire points. In this paper, HJ-1B CCD data and IRS data was used to study straw field identification and burning in Jiangsu Province, China. From the spectral reflectances of crops, straw, and turned farmland (dirt) collected from the HJ-1B CCD image, we found that straw and crops can be distinguished using the red band (CCD channel 3), while straw and turned farmland can be distinguished in the near-infrared band (CCD channel 4). Accordingly, straw spectral diagnosis indexes with three different forms (SMI1, SMI2 and SMI3) were established. The straw extraction results for the study area showed that SMI3 performs the best in differentiating straw from the mixed farmland, hence the accurate distribution of straw can be gained using SMI3. Based on HJ-B IRS Channel 3 and Channel 4 measurements and the fire detection algorithm proposed in this paper, fire pixels of all types on the ground can be easily identified. However, in addition to the straw-burning fires, the fire pixels initially identified include fires of other types, such as industrial hotspots, forest fires, and grassland fires, combined with the straw distribution map, straw-burning fires can successfully distinguished from other fires. Our study suggests that the straw distribution obtained in advance using HJ-1B CCD data can not only assist in identifying straw fire locations, but also can help to simplify the heat detection algorithm, omitting a series of complicated tests while improving the efficiency of the fire detection algorithm. A comparison of straw-burning fires detected by HJ-1B satellite with that detected by MODIS shows that the two results have similarities in spatial distribution patterns, and it also indicates that HJ-1B has more advantages in monitoring straw-burning fires which are usually small and dispersed.

Biography

Qingjiu TIAN is a professor of the Nanjing University, China. He received his BS in Optics from the Shandong University in 1987, his MS in Remote Sensing from the Institute of Remote Sensing Applications, CAS, in 1993, and PhD degrees in GIS & Remote Sensing from the Nanjing University in 2003. He is the author of more than 90 journal papers. His current research interests are hyperspectral and multispectral remote sensing. He is a vice Editor-in Chief of the Journal of Remote Sensing, China.