Premium
Estimating the volume and salt transports during a major inflow event in the Baltic Sea with the reanalysis of the hydrography based on 3DVAR
Author(s) -
Fu Weiwei
Publication year - 2013
Publication title -
journal of geophysical research: oceans
Language(s) - English
Resource type - Journals
eISSN - 2169-9291
pISSN - 2169-9275
DOI - 10.1002/jgrc.20238
Subject(s) - inflow , hydrography , sill , environmental science , data assimilation , barotropic fluid , salinity , advection , oceanography , volume (thermodynamics) , baltic sea , climatology , geology , meteorology , geography , physics , quantum mechanics , geochemistry , thermodynamics
Volume and salt transports are estimated from three dimensional variational analysis (3DVAR) reanalysis in the Baltic Sea. With the assimilation of temperature and salinity profiles, we find that 3DVAR assimilation causes larger volume and salt transports into the Baltic Sea than directly estimated in the Skagerrak for the 2003 major inflow. In addition, the daily transport obtained as the derivatives of total salt and volume reveals pronounced difference from what is directly estimated. The reason is found to be related to the unbalanced part in the reanalysis obtained from weak constraint 3DVAR. The reanalysis of hydrography is further used to estimate the transport in the Danish strait. For the 2003 major inflow, the total net volume/salt transports across the Darss Sill and Drogden Sill are comparable to observations and other model studies. However, the effect of the 3DVAR assimilation amounts to a volume decrease of 7 km 3 and a salt increase of 0.23 Gt in the Arkona region. The reason is related to the barotropic flow through the Danish strait that was supposed to be enhanced in the reanalysis. It suggests that the solution of 3DVAR is statistically optimal but not physically balanced. In coastal regions, reanalysis data must be used with special care when transport is estimated. Meanwhile, misfit of the transports acquired in different ways helps to identify problems in the application of 3DVAR and the model quality.