Analysis on spectral reflectivity of deep convective clouds towards vicarious calibration of UV/VIS hyperspectral instruments onboard geostationary satellites

To meet the increasing demand for obtaining reliable information on the atmospheric distribution of trace gases and aerosols, GEO-constellation consisting of Geostationary Korean Multi-Purpose Satellite-2B (GK-2B), Tropospheric Emissions: Monitoring Pollution and Sentinel-4 are planned to be operated in this decade. As one of the environmental instruments, Geostationary Environment Monitoring Spectrometer (GEMS) onboard GK-2B planned to launch in February 2020 is designed to provide spectral radiance in the wavelength range of 300-500 nm as observing the tropical western Pacific region. To prepare a means of monitoring the calibration accuracy of GEMS, we aim to evaluate the feasibility of deep convective clouds (DCCs) as a possible target for vicarious calibration of GEMS. While the DCC calibration technique has been continuously verified from various meteorological satellite programs, it has been rarely researched in the ultraviolet and visible spectral region especially for the hyperspectral data of the environmental sensor. To finely detect DCCs reflecting stable signal throughout the spectral range of GEMS, we update the DCC detection thresholds based on the conventional detection method by applying both visible and infrared detection thresholds. To examine the effectiveness of the detection, Tropospheric Monitoring Instrument (TROPOMI) onboard Sentinel-5 Precursor is used as a proxy of GEMS. Advanced Himawari Imager onboard Himawari-8 is also used to construct the collocated data with TROPOMI since the environmental sensor only provides spectral radiance at shorter wavelengths. The DCCs detected by the updated thresholds show higher reflectivity over 0.9 as presenting homogeneous spectral features even at the Fraunhofer lines in which the atmospheric effects are prominent. Cloud properties such as the cloud optical thickness and cloud top height also become relatively homogeneous when both visible and infrared thresholds are used for the DCC detection since both radiation thresholds can be complement to limit the cloud properties of the detected clouds. With the detailed results, bidirectional reflectance distribution function (BRDF) is also to be estimated by applying the updated DCC detection method hereafter in the study.