Gu, B.-H.; Woo, S.-B., and Kim, S.I., 2019. Improved estuaries salinity stratification at Gyeonggi Bay using data assimilation with Finite Volume Coastal Ocean Model (FVCOM). In: Lee, J.L.; Yoon, J.-S.; Cho, W.C.; Muin, M., and Lee, J. (eds.), The 3rd International Water Safety Symposium. Journal of Coastal Research, Special Issue No. 91, pp. 416-420. Coconut Creek (Florida), ISSN 0749-0208.

Assimilating salinity data into coastal models is very important and salinity vertical change is an important component to coastal forecasting. The goal of this study is to represent the salinity stratification at Gyeonggi Bay (GGB) in South Korea and to improve the accuracy of the prediction in hydrodynamics via data assimilation. The GGB area is located in the Yellow Sea between Korea and China and is a semi-closed estuary that has a tidal range above 7 m with a tidal flat where salt water and fresh water is mixed. Salinity stratification appears at Yeomha channel in GGB because it has characteristics of high tide and strong river flow. To represent the tidal flat and creek in the GGB in this study, the unstructured grid numerical model called Finite Volume Coastal Ocean Model (FVCOM) is employed. The performance of the model results is validated against the measured tide, current, and salinity at the Yeomha channel. In order to improve the numerical results, the salinity observations from the surface and bottom Conductivity Temperature Depth (CTD) data in the Yeomha channel are assimilated into the model only for the 4-day middle period out of total simulation days of 15. The direct comparison results show the clear effect of data assimilation for salinity stratification along vertical salinity, and the improvement of salinity profile lasts for 1 day after the end of data assimilation. The salinity stratification in the coastal are also improved as the number of observations and/or data assimilation periods is increased. These results are expected to improve the coastal forecasting system over the GGB area when applied the data assimilation scheme in to the forecasting system.