An active fire detection algorithm based on multi-temporal FengYun-3C VIRR data

REMOTE SENSING OF ENVIRONMENT

卷:211

Article

Lin, Zhengyang^1,2  Chen, Fang^1,2,3,4  Niu, Zheng⁴  Li, Bin¹  Yu, Bo¹  Jia, Huicong¹  Zhang, Meimei¹ 

[1] Chinese Acad Sci, Inst Remote Sensing & Digital Earth, Key Lab Digital Earth Sci, 9 Dengzhuang South Rd, Beijing 100094, Peoples R China

[2] Univ Chinese Acad Sci, Beijing 100049, Peoples R China

[3] Chinese Acad Sci, Inst Remote Sensing & Digital Earth, Hainan Key Lab Earth Observat, Sanya 572029, Peoples R China

[4] Chinese Acad Sci, Inst Remote Sensing & Digital Earth, State Key Lab Remote Sensing Sci, Beijing 100101, Peoples R China

关键词: FengYun-3C;    VIRR;    Active fire detection;    Multi-temporal;   

DOI  :  10.1016/j.rse.2018.04.027

来源: Elsevier

【 摘 要 】

The Visible and Infra-Red Radiometer (VIRR) is an improved third-generation sensor used for Earth observation with channels ranging from visible to thermal bands and carried on board the Chinese FengYun-3C satellite. An active fire detection algorithm based on VIRR data has already been designed and tested in different study areas worldwide. Most of the previously related algorithms were developed merely focusing on the spatial and spectral features of pixels while the temporal attributes of these observed active fires were ignored. In this research, multi-temporal VIRR data were used to construct time series of pixels. The core content of the algorithm consists of the changes in the time-series profiles together with the observed data. By calculating the predicted mid infrared (MIR) value and the stable MIR value of the target area, fire pixels can be easily distinguished. To assess the performance of this algorithm, a total of eight target areas distributed across the world were used for testing. Two stages of validation were carried out with data of different spatial resolutions. A rough comparison was carried out first. During this step, results from Collection 6 of MODIS Fire and Thermal Anomalies products (MOD14A1) and results generated from the previously used algorithm were used for comparison. The detailed validation work was conducted with the support of Landsat series (including ETM + and OLI sensors) data even though the different imaging time may affect the actual validation results.

【 授权许可】

Free