A STUDY ON THE SYSTEM FOR REAL-TIME GENERATION OF GEOSTATIONARY OCEAN COLOR IMAGER LEVEL 1B PRODUCT

Abstract

Geostationary Ocean Colour Imager (GOCI) is the world’s first geostationary satellite image sensor for the observation of ocean colour and has been employed for various ocean remote sensing applications so far. GOCI is distributed to users in the form of a Level 1B (L1B) product that has precise positional information and is the basis for generating various ocean analysis data. In this study, in response to the demand for real-time distribution of the GOCI L1B product, the GOCI pre-processing system has been designed and developed to generate GOCI L1B in real-time. First, simplification of the process is implemented by applying the updated correction angle in real time, instead of using the landmark determination process. Secondly, an optimum number of control points has been selected to reduce the processing time of least squares estimation to find the Rational Polynomial Coefficient (RPC) that provides the relationship between the image coordinates and their corresponding ground coordinates. Third, parallel programming is applied to both the generation of RPC and to the image re-sampling process, which together take the most time to generate the L1B. Through this process, it was confirmed that the processing time decreased by 5-10 times using the new process, compared with the existing processing time for example, it took less than 20 seconds for L1B production compared to 200 seconds using the current process in