Methods for the application of topography and NDVI in re-identification of remote sensing-based monitoring of forest fires
CHEN Yanying1(), MA Xincheng2(), XU Yanping3, WANG Ying4, WANG Yanbo5
1. Chongqing Institute of Meteorological Sciences, Chongqing 401147, China 2. CCCC Highway Consultants Co., Ltd., Beijing 100010, China 3. Kaizhou District Release Center of Emergency and Early Warning Information in Chongqing, Chongqing 405400, China 4. Jiangjin Modern Agricultural Meteorological Trial Station in Chongqing City, Chongqing 402260, China 5. Chaoyang Teachers College, Chaoyang 122000, China
The indicative significance of normalized vegetation index (NDVI) and terrain factors in land classification can be applied to specific scenarios. This study extracted the land classification information of Chongqing using the AQUA/MODIS NDVI and terrain indices (height and slope) of 2002—2020 and accordingly divided the land in Chongqing into seven types, i.e., forest land, grassland, orchard, dry fields, paddy fields, waters, and residential and building land, with the former three types being economic forest land. Based on the characteristics of broken terrain caused by the staggered distribution of agricultural, forest, and grassland, as well as the need for fire prevention in Chongqing, this study categorized the economic forest land and dry fields as concern areas of forest fires and categorized paddy fields, waters, and residential and building land as unconcerned areas of forest fires. The hotspots monitored using AQUA/MODIS in 2002—2020, FY3-C/VIRR in 2014—2020, and FY3-D/MERSI in 2019—2020 individually were re-identified based on the classification results of the concern areas of forest fires. The results are as follows. The extraction accuracy of individual land types (except for orchard and dry fields) was over 64%, and that of the concern areas of forest fires was over 86%. Based on the classification results of concern areas of forest fires, the forest fire points monitored using the remote sensing techniques were re-identified. The re-identification results showed that the 46.27%, 26.47%, and 11.76% of forest fire points monitored using AQUA/MODIS, FY3-C/VIRR, and FY3-D/MERSI, respectively were in unconcerned areas of forest fires. The forest fires monitored using remote sensing techniques on May 1-2, 2021 were re-identified, and 71.4%and 81.08% of forest fire points monitored using FY3-C/VIRR and both AQUA/MODIS and TERRA/MODIS, respectively were in unconcerned areas of forest fires. Therefore, extracting land classification information in complex terrain areas using NDVI and terrain indices and applying the extraction results to the re-identification of forest fires monitored using remote sensing techniques can effectively reduce the interference to forest fire monitoring in complex terrain areas, thereby minimizing the input of manpower and properties for the verification of hotspots.
陈艳英, 马鑫程, 徐彦平, 王颖, 汪艳波. 地形及NDVI在林火遥感监测二次识别中应用的方法探讨[J]. 自然资源遥感, 2022, 34(3): 88-96.
CHEN Yanying, MA Xincheng, XU Yanping, WANG Ying, WANG Yanbo. Methods for the application of topography and NDVI in re-identification of remote sensing-based monitoring of forest fires. Remote Sensing for Natural Resources, 2022, 34(3): 88-96.
Qin X L, Li X T, Liu S C, et al. Forest fire early warning and monitoring techniques using satellite remote sensing in China[J]. Journal of Remote Sensing, 2020, 24(5):511-520.
Zhang J, Zhang W Y, Feng J D, et al. An improved algorithm for forest fire detection:A study based on brightness temperature, vegetation index and AOD[J]. Remote Sensing Technology and Application, 2016, 31(5):886-892.
Rao Y M, Wang H, Huang H G. Forest fire monitoring based on multisensor remote sensing techniques in Muli County,Sichuan Province[J]. Journal of Remote Sensing, 2020, 24(5):559-570.
[8]
Kaufman Y J, Tucker Q, Fung L. Remote sensing of biomass burning in the tropics[J]. Journal of Geophysical Research, 1990, 95:9927-9939.
doi: 10.1029/JD095iD07p09927
[9]
Giglio L, Descloites J, Justice C, et al. An enhanced contextual fire detection algorithm for MODIS[J]. Remote Sensing of Environment, 2003, 87(2/3):273-282.
doi: 10.1016/S0034-4257(03)00184-6
Chen X F, Liu L, Li J G, et al. Application and research progress of fire monitoring using satellite remote sensing[J]. Journal of Remote Sensing, 2020, 24(5):531-542.
Hu M, Qi S H, Shu X B, et al. Monitoring fire from crop residues burning with MODIS data in north China plain[J]. Geo-Information Science, 2008, 10(6):802-807.
Chen Y Y, You Y S, He Z M, et al. Analysis on temporal-spatial feature of forest fires in Chongqing[J]. Chinese Agricultural Science Bulletin, 2014, 30(13):51-58.
Chen Y Y, You Y S, Tang Y H. Research of forest fire zoning based on local topography and fire numbers[J]. Journal of Natural Disaters, 2015, 24(2):228-234.
Li L L, Li J W, Zhong S Q, et al. Suitability evaluation of suitable-for-mechanization transformation of cultivated land based on topographic complexity in Chongqing,China[J]. Chinese Journal of Eco-Agriculture, 2022, 30(2):302-313.
[15]
毛玉龙. 第三次国土调查对耕地的调查和认定[J]. 西部资源, 2021(6):166-168.
Mao Y L. Investigation and identification of cultivated land in the third national land survey[J]. Western Resources, 2021(6):166-168.