摘要
基于环境一号卫星（HJ-1）遥感观测数据,采用改进的暗目标算法和深蓝算法分别对暗像元、亮像元地表的气溶胶光学厚度（AOD）进行反演,并将其结果与MOD04产品对比。结果表明：二者反演的AOD空间变化特征较一致,相关系数达0.874,HJ-1星反演结果有更高的空间分辨率。当AOD值〈0.2时,HJ-1星反演结果可靠性较低;当AOD值在0.2-0.8之间时,HJ-1星CCD相机反演气溶胶的结果与MOD04产品最相近,二者相对误差〈10%;当AOD值在0.8-1.5之间时,HJ-1星反演结果相对MOD04产品显著偏高;当AOD值〉1.5时,HJ-1星反演结果相对MOD04产品偏低。
Based on the data of HJ-1 satellite,aerosol optical depth（ AOD） over dark pixel and light pixel land surfaces was inversed by dark dense vegetation（ DDV） method and deep blue algorithm（ DBA）,and was compared against MOD04. The results showed that the spatial variation of AOD from HJ-1 inversion and MOD04 was more consistent,the correlation coefficient reached 0. 874,the results of HJ-1 inversion has higher spatial resolution. When the AOD value was less than 0. 2,the reliability of HJ-1 inversion was low. When the AOD value was between 0. 2 and 0. 8,HJ-1 CCD camera aerosol retrieval results showed the most similar to that with MOD04,the relative error of the two was less than 10%. While the AOD value was between 0. 8 and 1. 5,HJ-1inversion results showed significantly higher than MOD04. When the AOD value was higher than 1. 5,HJ-1 inversion results were lower than that of MOD04.
出处
《环境监测管理与技术》
CSCD
2015年第4期30-34,共5页
The Administration and Technique of Environmental Monitoring
基金
"十二五"国家科技支撑计划重大专项基金资助项目(2011BAK21B03)
江苏省科技支撑计划"绿色青奥南京区域大气污染监控预警及应急关键技术应用性研究"基金资助项目(BE2012771)
关键词
卫星遥感
气溶胶光学厚度
暗像元算法
深蓝算法
Satellite remote sensing
Aerosol optical depth（AOD）
Dark dense vegetation
Deep blue algorithm
