This paper addresses one of the most pressing challenges of Ukraine today, namely the establishment of a new cutting-edge automatized system for operational forecasting of oceanographic parameters in the Sea of Azov - the Black Sea basin. To reestablish a national maritime prediction system of Ukraine, lost after the Russian Federation had annexed the Crimea in 2014, the suite of dynamically coupled numerical models Delft3D-FLOW + Delft3D-WAVE (SWAN) is considered to be applied. This set of coupled numerical models was previously adapted to the conditions of the Black Sea area with employment of meteorological forcing fields from the Global Forecast System (GFS) model. Results of the model trial runs, which were used to evaluate and predict the marine oceanographic conditions in the North-Western part of the Black Sea near the Odessa Region are presented. The current version of the automatized modelling complex allows to obtain the following predictive oceanographic data: wind conditions, sea level deviations from the undisturbed state, spatio-temporal variability of the wind waves parameters, water circulation (currents) in the coastal zones with waves taken into consideration. Embedding the automatized modelling complex 'Delft3D-FLOW + SWAN' into the structure of the intelligent information system for revealing a hydrographic situation in the Black Sea can meet the challenge to operationally forecast the oceanographic conditions in the present (with a hindcast up to 5 days) and in the future (up to 4 days) for the entire Black Sea basin, focusing on its northwestern part and selected coastal areas with the required spatial resolution.
As a result of the occupation of the Crimean
Peninsula by the Russian Federation in 2014,
Ukraine lost the national automated maritime
forecasting system for the Black and Azov Seas,
which was established and operated on the basis
of the Marine Hydro-Physical Institute of the
National Academy of Sciences of Ukraine
(Sevastopol, Crimea) under financial and
technical support of the European Union [
Consequently, there is a demanding need for re-establishing the modern national system of operational forecasting of oceanographic parameters in the Ukrainian Azov-Black Sea basin to meet the needs of the maritime complex, maritime transport infrastructure, and the Naval Forces of Ukraine.
To accomplish this task, an automated
software complex, employing modern numerical
models, was developed at the Odessa State
Ecological University [
This paper presents the description of the structure of automated modelling complex for operational short-term forecasting of the oceanographic conditions in the Black Sea waters, the results of verification and validation of modules, comprising this complex, and discussion of the prospects for future improvements.
The automated modelling complex for
predicting the variability of oceanographic
characteristics in the Azov-Black Sea basin is
built around newer generation numerical models,
as compared against the ones [
The complex is based on the usage of two
software modules Delft3D-FLOW and
Delft3DWAVE of the suite of integrated environmental
models Delft3D [
Delft3D-FLOW is a multi-dimensional (2D or 3D) hydrodynamic (and transport) simulation program which calculates non-steady flow and transport phenomena that result from tidal and meteorological forcing on a rectilinear or a curvilinear, boundary fitted grid. It simulates thermal stratification in lakes, seas and reservoirs; stratified and density driven flows; tide and winddriven currents (i.e. storm surges); fresh-water river discharges in bays; non-hydrostatic flows; transport of dissolved material and pollutants etc.
Delft3D-FLOW solves the Navier-Stokes
equations for an incompressible fluid, under the
shallow water and the Boussinesq assumptions.
The system of equations consists of the horizontal
momentum equations, the continuity equation, the
transport equation, and a turbulence closure
model [
Delft3D-WAVE is based on the spectral model
SWAN (Simulating Waves Nearshore Model)
[
Both modules employ curvilinear computational grids in the horizontal plane and use the ‘telescoping’ technique for the results of calculations.
The modules are coupled by means of a shared interface and interact with each other. The influence of currents on the parameters of wind waves and wave propagation is taken into account in the coupled model. The computation of coastal currents and the intensity of turbulent mixing of waters incorporates wave processes as well.
A correct accounting for the effects of sea waves and currents interaction makes it possible to enhance the quality of calculation of the sea currents, water temperature and salinity in the upper layer of the water column.
The program codes of the Delft3D-FLOW and SWAN are compiled into executable files using the Visual Fortran and C ++ compilers. Both modules use the same set of computational grids and utilize all cores of workstation (or cluster nodes). The Delft3D-FLOW model splits a task for its parallel execution on processor cores (nodes) using the Message Passing Interface (MPI). The SWAN model (WAVE module), by default, uses parallel computations on all processor cores in accordance with the OpenMP (Open Multi-Processing) standard.
The basis of the oceanographic forecast is the
data of 10-days meteorological forecast from the
global weather forecast numerical model GFS
(Global Forecast System). A GFS web-service
(National Operational Model Archive and
Distribution System – NOMADS) is situated in
the United States [
The US National Weather Service provides
free access to the GFS forecast data. Ongoing
operational forecasts of meteorological
parameters are being read from the NOMADS
web resource (Data Transfer: NCEP GFS
Forecasts (0.25-degree grid) [
The modelling complex Delft3D-FLOW + SWAN is equipped with a service shell, which includes a graphical interface for use by end users. This shell automates the procedure of reading meteorological information from the NOMADS web service, filters these data and prepares it for use in the models, facilitates the procedure of setting up the Delft3D-FLOW and Delft3DWAVE (SWAN) software modules, performs model calculations on nested grids (NESTING procedure), provides visualization technique for input meteorological data and results of operational forecasting of oceanographic characteristics (using the QUICKPLOT software module).
Current version of the automated software complex initially performs the calculations on a generalized grid for the entire Azov-Black Sea basin with a spatial resolution of Δxy = 2.5-5 km (1 in Fig. 1A). Inside the basic computational grid, the following nested computational grids with higher spatial resolution were generated: 1. Grid for the northwestern part of the Black Sea with Δxy = 0.8-1.5 km (2 in Fig.1A). 2. Grid for the water area of the Odessa region at the North-Western parts of the Black Sea, where the seaports of Chernomorsk, Odessa, Yuzhny are located (Δxy = 90-250 m) (Fig. 1B).
Fig. 2 presents a schematic overview of the forecasting procedure, including data processing and interconnections between modules.
The task of ensuring the reliable oceanographic forecasts with the use of numerical models involves the implementation of procedures for models’ adaptation to the conditions of the studied water areas, their verification and validation.
Verification of the modelling complex was
performed by means of comparing the results of
simulated water level with observational data
from the marine hydrometeorological stations of
Hydrometeorological Center of the Black and
Azov Seas, located at the Chernomorsk, Odessa
and Yuzhny ports. Furthermore, the modelled sea
drifting currents and wind wave parameters were
compared against in-situ data logged at
hydrometeorological buoy SW Midi-185 (Fugro
OCEANOR, Norway) stationed in the Odessa
Bay (46.484N, 30.785E) [
Fig. 3, 4 present several results of model verification runs for the time periods of 08.10.18.10.2016 and 16.04.-25.04.2017, under stormy wind conditions. The assimilated meteodata from the GFS global atmospheric model was used as an input to the models.
The verification showed promising potential for employing the software complex of integrated numerical models ‘Delft3D-FLOW + SWAN’ as a part of the operational forecasting system for predicting the oceanographic parameters of the Ukrainian marine environment.
The validation of the model complex was performed by means of making the forecasts with different warning times. The produced 10-days forecasts of storm surges and wind waves in the water area of the Odessa region at the northwestern part of the Black Sea were compared against the observed values.
Validation results show that the forecast of wind surges and wave heights with warning time up to 5 days are in good agreement with the observation data, provided that there is no significant uncertainty of the meteorological forecast, in particular, wind conditions predicted by the GFS model.
Selected results of approbation of the modelling complex in the forecasting mode using the GFS synoptic forecasts of wind conditions over the Black Sea, are presented at fig. 5.
The results of the verification and validation of the complex of integrated numerical models ‘Delft3D-FLOW + SWAN’ demonstrate good prospects of using this complex as a part of the system of operational forecast of the variability of oceanographic parameters in the Ukrainian part of the Azov-Black Sea basin with assimilation of predictive meteorological information from the GFS global atmospheric model.
Operational oceanographic information, which can be obtained as a result of the application of the automated software complex ‘Delft3D-FLOW + SWAN’, contributes to the improvement of navigation safety, especially in shallow coastal and estuarine areas of the sea, on the approaches to the sea-ports and other areas of the Azov-Black Sea basin. The application of obtained prognostic information will result in increasing efficiency of the search and rescue operations due to recording the current (operational) and expected hydrometeorological conditions, and, especially, the pattern of the distribution of currents which determine the movement of objects with different buoyancy, including wind drift.
Integration of the automated modelling complex ‘Delft3D-FLOW + SWAN’ into the structure of an intelligent information system for revealing the hydrographic situation in the Black Sea will allow to solve the problems of the operational (fast) assessment of the state of the surrounding (marine) environment by providing the information about the oceanographic situation in the present (with a hindcast up to 5 days) and future (up to 4 days) time for the entire Black Sea, its northwestern part and selected areas of the coastal zones with the necessary spatial discretization of data.
The set of oceanographic data that can be obtained with the current version of the automated modelling complex ‘Delft3D-FLOW + SWAN’ includes: wind conditions, sea level deviations from the undisturbed state under the influence of wind in the coastal zones (which determine the actual depths), spatio-temporal variability of the parameters of wind waves, waters circulation (currents) in the coastal areas of the sea, considering the influence of wind waves.