GOCI-II 자료처리의 구현 간소화를 지원하기 위한 해색 알고리즘 개발 환경

Abstract

In 2019, Geostationary Ocean Color Imager&#8211 II (GOCI-II) will be launched as the follow-up satellite sensor of GOCI. GOCI has observed ocean color images 8 times a day for 8 spectral bands. In terms of GOCI-II observations, on the other hand, 5 more spectral bands will be added and 2 more ocean color images a day will be obtained in order to widen the spectra range and increase the number of observations, respectively. Also, the spatial resolution of GOCI-II will be more precise as 250 m (500 m for GOCI) and its number of products will be increased to 26 (13 for GOCI). We expect that the remote-sensing capability of GOCI-II will be greatly improved than that of GOCI. However, the existing ocean color algorithm development methodology with no regard for the high-speed data processing will not be accepted because significantly large GOCI-II data would be processed and distributed in real time. In this study, therefore, we designed a software development environment for helping the implementation of GOCI-II ocean color algorithm using the high-performance computing schemes. We expect that GOCI-II ocean color algorithm developers will be able to easily implement their source codes using the various parallelism techniques such as open multi-processing (OpenMP), open computing language (OpenCL), and message passing interface (MPI) under the proposed software development environment. In addition, it will support that th hand, 5 more spectral bands will be added and 2 more ocean color images a day will be obtained in order to widen the spectra range and increase the number of observations, respectively. Also, the spatial resolution of GOCI-II will be more precise as 250 m (500 m for GOCI) and its number of products will be increased to 26 (13 for GOCI). We expect that the remote-sensing capability of GOCI-II will be greatly improved than that of GOCI. However, the existing ocean color algorithm development methodology with no regard for the high-speed data processing will not be accepted because significantly large GOCI-II data would be processed and distributed in real time. In this study, therefore, we designed a software development environment for helping the implementation of GOCI-II ocean color algorithm using the high-performance computing schemes. We expect that GOCI-II ocean color algorithm developers will be able to easily implement their source codes using the various parallelism techniques such as open multi-processing (OpenMP), open computing language (OpenCL), and message passing interface (MPI) under the proposed software development environment. In addition, it will support that th