Research On Apparent Reflectance Characteristics Over Bright Objects And Urban Optical Thickness Retrieval Based On Low Light Data

The application of lowlight detection technology is one of the hotspots in recent years,which is of great significance to nighttime weather monitoring,model forecasting and human activity research.This paper first introduced the background and status of the research,compares the differences between DMSP/OLS and VIIRS/DNB in detail,summarized the research progress of radiometric calibration methods and application research of the satellite low light data.And the temporal and spatial changes of the apparent reflectance over three bright targets are analyzed.Based on this,the experiment was carried out to combine the low light and the infrared channel to detect the ice and snow region of the Qinghai-Tibet Plateau.Finally,the aerosol optical depth of the Beijing was retrieved by using the city light data.The current domestic satellite remote sensing calibration field is mainly composed of the lower reflectance desert and lake,which can't meet the requirement of low light sensor radiometric calibration.In this paper,we selected Guliya ice cap,Puruogangri ice field and Caka salt lake,which appear bright at night,extracted the information of radiance,observation geometry and moon phase from the homogeneous area,and calculated the apparent reflectance changes with time and moon phase angle in the near-full moon in three regions using apparent reflectance model at night and the lunar irradiance model MT2009.The apparent reflectance over the Guliya ice cap is the largest and stable for a long time,the apparent reflectance over the Puruogangri ice field is slightly smaller and susceptible to snowfall.The apparent reflectance of the Caka salt lake may closely related to the season and precipitation,but was the most stable in a specified time duration.It also shows that the larger the moon phase angle,the smaller and scattered the apparent reflectance,the maximum error occurs when the full moon.In general,the Guliya ice cap,the Puruogangri ice field and the Caka salt lake meet the basic requirements as a low-light satellite remote sensing radiometric calibration field.The snow and ice detection in the Qinghai-Tibet Plateau is of great significance to the snow disaster rescue and enrich the snow products.Based on the difference between the reflectance and emissivity between snow/land/cloud,the DNB apparent reflectance in the Qinghai-Tibet Plateau was calculated and the threshold was determined to separate the snow and cloud mixed areas from the land,and combined with pseudo-emissivity calculated from the M12/M15 IR Channel,which can effectively distinguish between land and cloud,and finally obtained the Qinghai-Tibet Plateau night snow area.And MODIS daily snow productMOD10C1 was compared,the consistency of snow/land detection results was 0.83,the results were affected by the low resolution and verify product precision.Night aerosol changes can help understand the role of aerosols in climate change and human production activities.According to the comparison of the standard deviation of the light between aerosol exists or not,the"deviation method"is used to retrieve the short time variation and regional distribution characteristics of aerosol optical thickness in March 2015 in Beijing.Compared with the average of MODIS aerosol products during the day in an aerosol northward case,found a good consistency while aerosol changed slightly at night,and based on four AERONET sites data to verify the authenticity of the results,the determination coefficient(R~2)was 0.69,RMSE was 0.24.Differences in observations were mainly due to different observation time of methods and the diurnal variations of aerosol.