In this paper, we introduce a cloud top-height (CTH) retrieval algorithm using simultaneous observations from the Himawari-8 and FengYun (FY)-2E geostationary (GEO) satellites (hereafter, dual-GEO CTH algorithm). The dual-GEO CTH algorithm estimates CTH based on the parallax, which is the difference in the apparent position of clouds observed from two GEO satellites simultaneously. The dual-GEO CTH algorithm consists of four major procedures: (1) image remapping, (2) image matching, (3) CTH calculation, and (4) quality control. The retrieved CTHs were compared with other satellite CTHs from the Cloud-Aerosol Lidar with Orthogonal Polarization (CALIOP) and the Cloud-Profiling Radar (CPR), on three occasions. Considering the geometric configuration and footprint sizes of the two GEO satellites, the theoretical accuracy of the dual-GEO CTH algorithm is estimated as ±0.93 km. The comparisons show that the retrieval accuracy generally tends to fall within the theoretical accuracy range. As the dual-GEO CTH algorithm is based on parallax, it could be easily applied for the estimation of the height of any elevated feature in various fields.

Stereoscopic cloud-top height (CTH) retrieval from two geostationary (GEO) satellites is usually realized through a visible (VIS) band with a high horizontal resolution. A stereoscopic-based CTH retrieval algorithm (prototype dual-GEO CTH algorithm) proposed in our previous study also adopts this approach. Although this approach can retrieve accurate stereoscopic CTHs, the heights of optically thin upper clouds overlying the lower clouds are challenging to retrieve because the parallax difference between two GEOs is determined by the lower clouds owing to the low reflectance from the upper clouds. To address this problem, this paper proposes an improved stereoscopic CTH retrieval algorithm, named the improved dual-GEO CTH algorithm, for Himawari-8 and FengYun (FY)-4A GEOs. The proposed algorithm employs an infrared (IR) band in addition to a VIS band. A seamless image cloning technique is adopted to blend the VIS and IR images, which are then used to retrieve the stereoscopic CTHs. The retrieved CTHs are compared with the Cloud-Aerosol Lidar with Orthogonal Polarization (CALIOP) and Cloud Profiling Radar (CPR) CTHs for three occasions involving upper clouds overlying lower clouds. Results show that the proposed algorithm outperforms the prototype dual-GEO CTH algorithm in the case of upper clouds overlying lower clouds. Notably, although the proposed algorithm is intended for Himawari-8 and FY-4A GEOs, it can be easily extended to any combination of two GEOs.

No abstract