干旱气象

• 论文 • 上一篇    下一篇

MODIS气溶胶光学厚度产品在西南地区林火检测中的应用

王研峰1,张  婕2张弘毅3,谢  巍2阿如哈斯2吴妮晏2   

  1. 1.甘肃省人工影响天气办公室,甘肃 兰州 730020;
    2.成都信息工程大学大气科学学院,高原大气与环境四川省重点实验室,四川 成都 610225;
    3.四川省珙县气象局,四川 珙县 644500
  • 出版日期:2018-10-31 发布日期:2018-10-31
  • 通讯作者: 张婕(1975— ),女,云南个旧人,讲师,研究方向为大气辐射与大气遥感. E-mail:zj3130@cuit.edu.cn。
  • 作者简介:王研峰(1986— ),男,甘肃会宁人,工程师,主要从事大气遥感和云物理降水研究. E-mail:wangyanfeng_1986@163.com。
  • 基金资助:

    国家自然科学基金重大研究计划重点支持项目(91644226)、国家自然科学基金项目(41775147)和甘肃省气象局气象科研面上项目(GSMAM2017-02)共同资助

WANG Yanfeng1, ZHANG Jie2, ZHANG Hongyi3,XIE Wei2, Aru Hasi2, WU Niyan2   

  1. 1. Gansu Weather Modification Office, Lanzhou 730020, China;
    2.College of Atmospheric Sciences, Chengdu University of Information Technology, Plateau Atmosphere and Environment Key Laboratory
     of Sichuan Province, Chengdu 610225, China;
    3. Hong County Meteorological Bureau of Sichuan Province, Hong County 644500, Sichuan, China
  • Online:2018-10-31 Published:2018-10-31

摘要:

考虑到现有MODIS火灾检测算法易对低温火点产生漏判,通过研究MODIS C6-3 km、C5.1-10 km AOD产品在西南地区不同区域森林火灾伴生烟羽扩散和分布的特性表征,对比火点中心32个方位、不同扩散范围下烟羽正、反方向AOD累积效应,以提高卫星遥感数据对火点的检测精度。结果表明:在西南中东部林区,MODIS C6-3 km AOD产品能较准确地表征该地区林火伴生烟羽的分布特征及扩散方向,且以火点为圆心,在3、6、9、15、24 km烟羽扩散半径内,当MODIS C6-3 km AOD累积AOD比值大于1.55的特征半径存在时,该产品对烟羽检测有一定指示意义。在西南高山林区,基于暗像元法反演的MODIS C6-3 km AOD产品,由于算法中地表反射率的假定误差较大,且烟羽易识别为薄云,可能无法反演出有效数值,对林火伴生烟羽的检测能力较差,而深蓝算法反演的MODIS C5.1-10 km AOD产品对伴生烟羽的检测具有一定优势。

关键词: 西南地区, 卫星遥感, AOD, 火灾, 检测

Abstract:

Considering the existing shortcomings of MODIS fire detection algorithm to low temperature fire point, the diffusion and distribution characteristics of smoke plume accompaning with forest fires from MODIS C6-3 km AOD products and C5.1-10 km AOD products were analyzed in different areas of southwestern China, and the accumulative effects of AOD of positive and negative direction of fire point center in 32 azimuth angles with different diffusion ranges were compared to improve detection accuracy of fire point. The results show that MODIS C6-3 km AOD products could accurately represent the distribution characteristics and diffusion direction of smoke plume, and when the ratio of accumulative AOD was greater than 1.55 in the radius range of 3, 6, 9, 15 and 24 km from fire point in central and eastern forest part of southwestern China, it had a certain indicative significances for the detection of smoke plume, which would help to identify fire point. MODIS C6-3 km AOD products were retrieved by using the dark target algorithm. However, the error of surface reflectivity assumption in dark target algorithms was bigger in mountainous area of southwestern China, and the smoke plume was easily identified as thin clouds, and the valid data might not be retrieved, so the detection of MODIS C6-3 km AOD products to accompanying smoke plume of fire was poorer. MODIS C5.1 -10 km AOD products generated by using the deep blue algorithm had a certain advantages to detect smoke plume in alpine area of southwestern China.

Key words: southwestern China, satellite remote sensing, AOD, fire, detection

中图分类号: