한국 주변해 중기 해양환경 예측모형 구축: 시범 예측

Abstract

Despite its socio-economic implications for the effective management of fisheries resources, a prediction of marine environment has been rarely performed on the medium range timescale of several months to a year. In this study, we have developed an ocean mid-range prediction system (OMIDAS) based on the Regional Ocean Model System (ROMS) for the Northwest Pacific Ocean focusing on the seas around Korea. To assess the skill of mid-range prediction by the OMIDAS system, a three-month reforecast was performed for the case of March 2016 by using dynamic downscaling of Climate Forecast System version 2 operational analysis data. Although the reforcasted SST shows enhanced amplitudes relative to the Optimum Interpolation SST version 2 (OISSTv2) reanalysis data, the spatial pattern is generally similar to the OISSTv2, suggesting a relatively good performance of the OMIDAS. A reforecast experiment with two different initial conditions suggests that a higher resolution of data used for model initial condition contributes to the performance of the mid-range reforecast in the seas around Korea. Based on the preliminary analysis for the reforecast experiment, we can conclude that the OMIDAS system has measurable skill on the mid-range timescale for the seas around Korea.