Abstract and keywords
Abstract (English):
This study is aimed at developing of an operational oceanographic system for the Russian sector in the Gulf of Finland and South-Eastern part of the Baltic Sea for operational forecast of hydrodynamic and ecosystem parameters on the basis of high and ultra-high spatial resolution models. The system is presented as a complex of regional and local models; for which a coupled modeling integration at boundary conditions exchange is fulfilled. The models share common mathematical formulation of general motion equations and a unified realization on the basis of programme code modeling modules designed for the ocean modeling – NEMO. The regional model of the Baltic Sea circulation is complemented by a module for the inert matter transport simulation. The latter is set up on the basis of a matter turbulent diffusion model with the use of two consistent equation systems: deterministic and stochastic. The designed operational system consists of two subsystems: operational oceanographic system of the coastal areas of the Baltic Sea and an expert-analytical system of operational monitoring of the aquatic environment and effective response to accidents at sea.

Baltic Sea, coastal zones, hydrodynamic model, operational forecast, matter transport, expert-analytical system, operational monitoring



Today the operational oceanography in the region of the Baltic and Northern Seas is represented by the operational oceanographic system developed in Danish Meteorological Institute (DMI) and the system developed in frames of the Baltic Operational Oceanographic System (BOOS) project under the auspices of the European module of "Global Ocean Observing System (GOOS)". The basis for DMI operational oceanographic system is the hydrodynamic model HBM. The model has four built-in levels with the resolution of 9, 3, 1 and 1/3 nautical miles, with the highest resolution in the Danish waters. Since 2001 the participants of the BOOS project have been Sweden, Finland, Russia, Estonia, Latvia and Lithuania. The basis for BOOS system is the three-dimensional hydrodynamic model HIROMB (Swedish Meteorological and Hydrological Institute, SMHI). The forecasting area extends out to the English Channel in the west and to the northern North Sea in the northwest, where the horizontal resolution is 3 nm (nautical miles). The grid is nested to a higher-resolution grid (1 nm resolution) which covers the whole Baltic Sea, the Danish Straits, Kattegat and Skagerrak.

Spatial resolution of the models used for operational forecasting of the Baltic Sea state is not sufficient for solution of vortex dynamics in Russian marine waters, and that negatively affects the possibility of surface circulation prediction. This operational information is required for reduction of negative anthropogenic impacts including oil spills, and therefore for ensuring the ecological safety of the marine environment.

In RSHU the operational oceanographic system has been developed that is aimed at producing the 3-D hydrophysical fields (current velocities, sea level, water temperature and salinity, concentration and thickness of ice cover) in the Russian water area: the Eastern part of the Gulf of Finland and the Kaliningrad Shelf. The operational system consists of two local hydrodynamic models of the Gulf of Finland and Southeastern Baltic Sea (the Kaliningrad Shelf). Shelf sea modelling is characterized by a demand for many different configurations to meet multiple science and user needs. NEMO (Nucleus for European Modelling of the Ocean) gives the capability to rapidly configure shelf sea models using appropriate high resolutions and parameterizations on the representation of coastal dynamics, thus for realization of the 3-D thermo-hydrodynamic model the NEMO code was selected.



NEMO is a 3-D hydrostatic, baroclinic primitive equation model written in nonlinear orthogonal coordinate system implying the existence of confluence areas of the computational grid. Equation approximation is produced on the Arakawa C-grid (Madec et al., 1998; Madec, 2012). Vertical coordinate system is realized in two versions: z* and hybrid coordinate. NEMO is being developed in a framework of a community of European institutes and thus benefits of the recent scientific and technical developments implemented in most ocean modelling platforms. The NEMO implementation for the Baltic Sea uses the TVD advection scheme in the horizontal direction, the piecewise parabolic method (PPM) in the vertical direction [Liu and Holt, 2010], the non-linear variable volume (VVL) scheme for the free surface. In the horizontal plane, the model uses the standard Jacobean formulation for the pressure gradient, the viscosity and diffusivity formulation with different parameterizations of coefficient for momentum and tracer diffusion, including the Smagorinsky model of subgrid turbulence. The horizontal viscosity and diffusivity operators are rotated to be aligned with the density iso-surfaces to accurately reproduce density flows. In vertical direction the TKE and GLS schemes of turbulent closure can be used with additional inclusion of parameterization of breaking surface waves.

For the Gulf of Finland the local 3-D hydrodynamic model of ultrahigh resolution is realized with the possibility of reproduction of coastal floods phenomena during storm surge events, taking into account the influence of hydraulic engineering protective constructions (Flood Protection Barrier, FPB). The resolution required for description of submesoscale eddies is defined by Rossby baroclinic radius of deformation, approximately equal to 2-4 km in the Gulf of Finland (Alenius P. et al, 2003, Soomere T. & Quak E., 2013), and makes maximum 0.5 km.

The calculation domain covers the entire Gulf of Finland including the Neva Bay. Its open boundary lies along the meridian 23.5 ° W. Horizontal resolution varies from 200 m near the Neva Bay to 2 km in the western part of the gulf, with parameterization of water flow-through and ship-through constructions of FPB. The scheme of calculation domain for the Gulf of Finland is presented at figure 1. Water column is vertically divided in 80 layers. Average vertical resolution is about 1 m.


Figure 1 – Scheme of the calculation domain (bathymetry and grid points). One cell corresponds to 10 grid cells.


When the local hydrothermodynamic model of the Kaliningrad Shelf has been developed the following key geophysical features of the modeled object has been considered: bathymetry, drift circulation, coastal upwellings and downwellings, vortex dynamics. The baroclinic radius of deformation according to the estimates presented in Golenko M. N. & Golenko H.H 2011 is 4 km. The model computational domain boundaries are: from west to east – the meridians 18 and 21 º W, in the south – coastline, in the north – parallel 56 ºN. The Vistula, Kaliningrad and Kuronian Bays are not included in the domain, i.e. only the offshore area of Vistula (Baltic) and Kuronian Spits is modelled. Horizontal resolution in confluence areas reaches 200 m. Average vertical resolution is about 1 m.

Boundary conditions for the two local models are delivered by the NEMO-Nordic model (Hordoir et al., 2015), jointly developed by SMHI, FMI and RSHU and realized for the Baltic Sea and part of the North Sea with the horizontal resolution of 2 nm.

In this project the daily computational fields of the HIRLAM model (High Resolution Limited Area Model) with spatial resolution of 7.5 km will be used as a basic meteoforcing for submesoscale processes study (horizontal scale for the Baltic Sea about 2-4 km). Besides, in the test mode the predictive fields of another FMI forecasting product – the mesoscale HARMONIE model with a spatial resolution of 2.5 km are available. These model products actually represent the realization of a 2nd type dynamic downscaling at first of global predictive model of the European Center for Mediumterm Weather Forecasts (ECMWF) – HIRLAM, and then in a link HIRLAM – HARMONIE.

An integral part of any operational oceanographic system is the procedure of observational data assimilation. In the considered OOSB the three-dimensional variation assimilation scheme (3D-VAR) stated by Dobricic S. and Pinardi N. (2008) is used for this purpose.

Here both assimilation of remote sensing (satellite SST and altimetry), and of contact measurement data is possible (data on water temperature and salinity from buoy stations, data from drifters).

Data assimilation is a cyclic procedure according to the scheme "analysis-forecast". At the analysis step in each cycle all observations in a certain (assimilation) time window are acquired. The assimilation procedure allows to produce predictive cycles of the operational system presented at figure 2.2. A predictive cycle is 48 hours, with the forecast updating every 12 hours by starting the forecast from a new initial field. So the assimilation window is 24 hours, thus, every second forecast starts with an analysis field corrected by the assimilation procedure. Data of observations during the assimilation window (24 hours) are considered as instantaneous.

Unlike the basic assimilation cycle, there is also an additional periodic assimilation cycle covering 10 "past" time assimilation windows. Use of the additional cycle is due to the fact that part of observational data comes with a delay or, as in a case with satellite altimetry, is periodically updated as presented at figure 2. In this case the cycle forecast-assimilation-analysis is done all over again (every 12 hours with the moving assimilation window of 24 hours within 10 days) to obtain a new analysis field taking into account observational data for the past 10 days.



Figure 2. – Operational system predictive cycles.


Also as a part of the operational system there is an expert and analytical system of operational control of the state of the study water area and of fast emergency response to marine accidents. The main aim of the expert and analytical system is to organize remote automated workplaces for an observer – hydrometeorologist. The automated workplace of observer - hydrometeorologist can be used for operational analysis of various options of environmental impact on the basis of a numerical model of admixture distribution remotely started on the main server and adapted for various options of emergency actions.

The basis of the expert and analytical system is the admixture transfer model formulated in the Lagrangian coordinates. Initial data for it are the predictive fields automatically delivered by the operational oceanographic system (surface current velocities, diffusion coefficients), and location and volume of emission set by the user. On the basis of data on current velocities the ensemble of possible trajectories of passive admixture is calculated.



The study considers the operational oceanographic system developed for the Russian sector of the Baltic Sea: the Gulf of Finland and the Kaliningrad Shelf. It includes local hydrodynamic models of these water areas of ultrahigh resolution up to 200 m based on the NEMO programme code. Boundary conditions at open boundaries of computational domains are delivered by the NEMO-Nordic model developed jointly by SMHI, FMI and RSHU and realized for the Baltic Sea and part of the North Sea with the horizontal resolution of 2 nm. The meteorological forcing is provided by the HIRLAM model. The operational system includes the block of data assimilation based on variation assimilation 3D-VAR. There is also a separate component: an expert and analytical system designed for calculation of an ensemble of possible trajectories of passive admixture (including oil spills) using predictive fields of surface current velocities.


IV. Acknowledgments

This work was supported by the Federal Target Programme for Research and Development in Priority Areas of Development of the Russian Scientific and Technological Complex for 2014–2020 (Identification No.: RFMEFI57414X0091).



1. Alenius, P., Nekrasov, A., and Myrberg, K.: The baroclinic Rossby-radius in the Gulf of Finland, Cont. Shelf Res., 23, 563–573, 2003.

2. Golenko M.N., Golenko N.N. Effects of entrainment and vorticity at wind-driven coastal upwelling and downwelling for the South-Eastern Baltic // Aerokosmicheskie issledovaniya, prikladnaya mekhanika, TRUDY MFTI, 2011, 3(3), 56-63

3. Dobricic S, Pinardi N. An oceanographic three-dimensional variational data assimilation scheme // Ocean Modelling. 2008. V.22. P.89–105.

4. Hordoir, R., Schimanke, S., Axell, L., et al.\ 2015, EGU General Assembly Conference Abstracts, 17, 12657

5. Madec, G., Delecluse, P., Imbard, M., and Levy, C.: OPA 8.1 Ocean General Circulation Model reference manual. Note du Pole de modelisation, Institut Pierre-Simon Laplace (IPSL), Paris, France, No 11, 91 p., 1998.

6. Madec, G.: NEMO ocean engine. Note du Pôle de modélisation, Institut Pierre-Simon Laplace (IPSL), Paris, France, No 27 ISSN No 1288–1619, 2012.

7. Soomere T., Quak E. Preventive methods for coastal protection: Towards the use of ocean dynamics for pollution control / Eds. Springer. 2013. 442 p

Login or Create
* Forgot password?