Multi-layer cloud detection and distributions over the Asia–Pacific region based on geostationary satellite imagers

This study improved a threshold algorithm for detecting ice-over-water multi-layer (ML) clouds using geostationary satellites. The optimal thresholds were established by the Advanced Himawari Imager (AHI) and the Advanced Geostationary Radiation Imager (AGRI) spectral characteristics, respectively. Validation with collocated space radar (CloudSat Cloud Profiling Radar, CPR) and lidar (Cloud-Aerosol Lidar with Orthogonal Polarization, CALIOP) measurements. The ML clouds inferred by AHI and AGRI exhibited similar annual distributions. Furthermore, hourly AHI observations over six years detected obvious monthly and daily variations in ice-over-water clouds over the Asia–Pacific region. The specific descriptions of the data are as follows: Datasets(“fig1”): AHI and AGRI cloud classification results and the CPR- CALIOP cloud vertical distribution observations. It is a case study of ML cloud detection to verify the accuracy of the detection results. Datasets(“fig2”): statistical percentages of ice, Pro ML, and ML clouds identified by AHI and AGRI compared with CPR- CALIOP products. Datasets(“fig3”): the percentages of AHI and AGRI cloud occurrence frequencies, including ice-over-water cloud, ice cloud, water cloud, total cloud occurrence frequency. Datasets(“fig4”): Zonal and meridional occurrence frequency distributions of ice-over-water, ice, water, and total clouds. Datasets(“fig5-6”): ice-over-water cloud occurrence frequency distributions and anomalies for each month derived from the AHI for the 2017–2022 period. Datasets(“fig7”): mean percentage of ice-over-water cloud occurrence frequency for each month for 2017–2022. Datasets(“fig8”): average occurrence frequency of six time zones (UTC+06 to UTC+11) in the Northern and Southern Hemispheres across the four seasons. Datasets(“fig9-10”): anomalies of ice-over-water cloud occurrence frequency from UTC+06 to UTC+11 LST in the Northern and Southern Hemisphere from 2017 to 2022.本研究针对利用静止卫星检测冰覆水多层（Multi-Layer, ML）云的阈值算法开展了改进工作。研究分别基于先进葵花成像仪（Advanced Himawari Imager, AHI）与先进静止轨道辐射成像仪（Advanced Geostationary Radiation Imager, AGRI）的光谱特性确定了最优阈值。随后利用协同观测的星载雷达——CloudSat云廓线雷达（CloudSat Cloud Profiling Radar, CPR）与正交偏振云-气溶胶激光雷达（Cloud-Aerosol Lidar with Orthogonal Polarization, CALIOP）的实测数据开展验证。由AHI与AGRI反演得到的多层云呈现出相似的年度分布特征。进一步地，基于六年逐小时AHI观测数据，研究发现亚太区域冰覆水云存在显著的月际与日际变化特征。 数据的具体说明如下： 数据集"fig1"：包含AHI与AGRI的云分类结果，以及CPR-CALIOP协同观测获取的云垂直分布实测数据，用于开展多层云检测的案例研究以验证检测结果的准确性。 数据集"fig2"：AHI与AGRI识别得到的冰云、Pro ML云及多层云的统计占比，并与CPR-CALIOP产品进行对比分析。 数据集"fig3"：AHI与AGRI的云发生频率占比，涵盖冰覆水云、冰云、水云及总云发生频率。 数据集"fig4"：冰覆水云、冰云、水云与总云的纬向与经向发生频率分布。 数据集"fig5-6"：基于2017-2022年AHI观测数据得到的各月冰覆水云发生频率分布及其距平值。 数据集"fig7"：2017-2022年各月冰覆水云发生频率的平均占比。 数据集"fig8"：南北半球四个季节下6个时区（UTC+06至UTC+11）的平均云发生频率。 数据集"fig9-10"：2017-2022年南北半球UTC+06至UTC+11地方时的冰覆水云发生频率距平值。