Assimilation of SST in the Yellow and East China Sea using the ensemble Kalman filter
In order to increase accuracy of ocean circulation prediction,sea surface temperature (SST) data have been assimilated to the operational Yellow and East China Sea circulation model using the ensemble Kalman filter.As observed data were continuously assimilated into the model in time,ensemble spread decreased and ensemble members converged to a certain state making no use of the observed data after 20 days.In this study,to prevent this ensemble divergence we made the ensemble with multi-models by perturbing model parameters such as bottom drag coefficient,light attenuation depth and wind data.Assimilation result from the multi-model ensemble was compared with that from the single model ensemble.Ocean circulation was forced with a regional atmospheric model product from September to December 2011.SST from a daily composited NOAA AVHRR dataset was assimilated every day.Simulated SST was compared to the daily mean sea surface temperature observed at four buoys in the coastal region and hydrographic observation data.RMSD (root-mean-square deviation) between simulated SST by non-assimilative model and observed SST at the buoys was 1.20-1.50oC.Assimilation of the satellite-borne SST data using the multi-model ensemble reduced RMSD to 0.70-1.09oC at the buoy stations.Assimilation of SST also made subsurface temperature profile closer to the observation.The multi-model ensemble has larger spread and smaller RMSD in temperature than a single model ensemble and a non-assimilative model.
SST assimilation ensemble Kalman filter the Yellow Sea
Byoung-Ju Choi Kyung Man Kwon Sang-Ho Lee Gwang-Ho Seo Yang-Ki Cho
Department of Ocean Science and Engineering, Kunsan National University, Gunsan, Korea School of Earth and Environmental Sciences, Seoul National University, Seoul, Korea, 151-742
国际会议
The 17th Pacific -Asian Marginal Seas Meeting(第十七届太平洋与亚洲边缘海国际会议)
杭州
英文
622-631
2013-04-23（万方平台首次上网日期，不代表论文的发表时间）