Inter-annual comparison of the East Sea with ARGO data assimilation model results

Abstract

The Korea Ocean Prediction System (KOPS) consists of a primitive equation, three-dimensional circulation model commonly known as the Princeton Ocean Model used for the operational forecasting in the East Sea region. ARGO data assimilation was performed using nudging technique. The ARGO observations used for a given day are supplied through a 10 days time window. If an observation falls with in the recent 10 days time window with respect to the model assimilation time (t-10 to t) then only the observation is used for the assimilation after giving proper weight (the oldest and the farthest observation gets the less the weight as compared to the recent and nearby observation). When we are limited with only SSH data assimilation, model doesn’t resolve the vertical structure quite well. While addingARGO data assimilation the vertical structure improved and the results look more realistic. Comparison of these results (only SSH and SSH + ARGO) with ARGO profiles reveal the difference more clearly. As the number of observations increased (very sparse in 2001, but increased year by year), significant changes are observed in the vertical structure of temperature. Selected vertical sections of Temperature are studied which show upper layer warming during summer months of 2006 and 2007 as compared to other years (2001 to 2005).