摘要
利用2012年2月3日和3月6日在北京市组织的星地同步观测实验,开展了天宫一号(TG-1)高光谱成像仪各通道在轨辐射定标与检验研究。首先,利用3月6日地面定标场的定标实验数据,采用MODTRAN辐射传输模型,根据反射率基法获取天宫一号高光谱成像仪各通道辐射定标修正系数,修正后的辐射定标参数与实验室和星上参数存在一定程度的差异,可见近红外和短波红外各通道相对差异均超过了10%。然后,利用2月3日的星地同步观测数据,对辐射定标结果进行验证。结果表明:定标后的图像精度大大提高,并且定标结果与验证数据具有很好的一致性。在绝大多数通道,两者的相对误差小于10%,相对误差较大的通道主要分布在臭氧、水汽等气体吸收波段。因此,天宫一号在轨定标的星地实验校正了高光谱成像仪的辐射定标系数,提高了改进辐射校正产品的定量化精度。
TG-1, the first Chinese space laboratory mod large numbers of high-resolution image data by the can't be applied to quantitative analysis because ule launched on 29 September 2011 ,has accumulated hyper-spectral imager. However, these images still the huge spectral and radiance difference between ground and onboard conditions. So a radiometric calibration is very necessary to correct these laboratory and on-board radiance calibration parameters for better quantitative application. This paper aims to cali- brate TG-1 hyper-spectral imager using reflectance-based calibration method through performing ground calibration experiments on 3 February and 6 March 2011. Firstly,this study used the ground experiment da- ta on 6 March to obtain each channel's calibration correction coefficients, which were different a lot from laboratory and on-board ones,relative error in all channels are bigger than 10 percent. Then,the calibration experiment data on 3 February 2011 were used to validate the result. Results show that calibration correc- tion coefficients improved the image accuracy, the calibration and validation experiments' results are in good agreement. In most channels,the relative errors are less than 10 percent, except that bigger error ap- pears in absorption channels. Therefore, this calibration experiment renewed the calibration coefficients, and improved the quantitative level of the TG radiance products.
出处
《遥感技术与应用》
CSCD
北大核心
2013年第5期850-857,共8页
Remote Sensing Technology and Application
基金
国家973计划项目(2010CB951701)
载人航天工程天宫一号民用试应用项目及国家自然科学基金项目(41001206)资助
关键词
辐射定标
修正系数
反射率基法
MODTRAN
天宫一号
高光谱成像仪
验证
Radiometric Calibration
Correction coefficients
Reflectance-based method
MODTRAN
TG-1
Hyper-spectral imager
Validation