2 Topography, Hydrography, Circulation and Modelling of the Baltic Sea
59
resolution simulations the output of local atmospheric models such as different versions of the High Resolution Limited Area Model (HIRLAM) (www.hirlam.org)
or the DWD model (www.dwd.de), ERA-40 reanalysis (Uppala et al. 2005) or its
downscalings (Höglund et al. 2009; Samuelsson et al. 2011) are typically used as
meteorological forcing. As the resolution of such analyses is often too coarse for
local hindcast studies and several systematic features of air flow in some basins
are not captured (Keevallik and Soomere 2010), new, high-resolution reanalyses are
urgently needed.
One of the most important inputs in the Baltic Sea modelling is the voluminous river discharge. The relevant data are available, e.g., from the BALTEX Hydrological Data Centre in terms of monthly mean values (Bergström and Carlsson
1994). The initial conditions for hydrographic parameters are available, e.g., from
the Baltic Environment Database (BED), which can be used through the Data Assimilation System (DAS, Sokolov et al. 1997). Observations from buoys and tide
gauges can be used for model verification and data assimilation (Myrberg et al.
2010). A big challenge is still the availability and accurate description of the open
boundary conditions in the Danish Straits where only a limited number of data are
available. These problems are often avoided through coupling the Baltic Sea models
with North Sea models.
2.5 Summary: How the Baltic Sea Can Be Replicated
by Numerical Models of Today
This Chapter has introduced the reader to the very specific environment of the Baltic
Sea, especially its topography, hydrography and circulation dynamics. Keeping in
mind the extensive ongoing modelling activities and the ever increasing use of various models to operationally forecast important features of the Baltic Sea dynamics,
a question arises: how accurately can the present 3D numerical models describe the
Baltic Sea physics? What kind of problems might still exist?
Proper numerical modelling of the physics and dynamics of the Baltic Sea is
a great challenge and a very demanding task even for the modern 3D hydrodynamic models because of a number of reasons. These problems and potential shortcomings in the model results should be kept in mind whenever studying particular
model-based analyses or conclusions, especially when interpreting the outcome of
the modelling-oriented chapters of this book. The first task for any 3D model to
solve is the proper description of the stratification. Regrettably, in the Baltic Sea
this has, strictly speaking, not exactly been solved to a satisfactory level. Even the
very best models tend to often exhibit problems to replicate the right position of the
halocline or a realistic vertical gradient of density. There is no unique opinion about
the reasons behind this feature that may substantially undermine the accuracy and
reliability of the simulations of water circulation and spreading of substances over
the entire water column. Fortunately this shortage only insignificantly affects the
results of modelling the velocities and transport in the uppermost layer.
59
resolution simulations the output of local atmospheric models such as different versions of the High Resolution Limited Area Model (HIRLAM) (www.hirlam.org)
or the DWD model (www.dwd.de), ERA-40 reanalysis (Uppala et al. 2005) or its
downscalings (Höglund et al. 2009; Samuelsson et al. 2011) are typically used as
meteorological forcing. As the resolution of such analyses is often too coarse for
local hindcast studies and several systematic features of air flow in some basins
are not captured (Keevallik and Soomere 2010), new, high-resolution reanalyses are
urgently needed.
One of the most important inputs in the Baltic Sea modelling is the voluminous river discharge. The relevant data are available, e.g., from the BALTEX Hydrological Data Centre in terms of monthly mean values (Bergström and Carlsson
1994). The initial conditions for hydrographic parameters are available, e.g., from
the Baltic Environment Database (BED), which can be used through the Data Assimilation System (DAS, Sokolov et al. 1997). Observations from buoys and tide
gauges can be used for model verification and data assimilation (Myrberg et al.
2010). A big challenge is still the availability and accurate description of the open
boundary conditions in the Danish Straits where only a limited number of data are
available. These problems are often avoided through coupling the Baltic Sea models
with North Sea models.
2.5 Summary: How the Baltic Sea Can Be Replicated
by Numerical Models of Today
This Chapter has introduced the reader to the very specific environment of the Baltic
Sea, especially its topography, hydrography and circulation dynamics. Keeping in
mind the extensive ongoing modelling activities and the ever increasing use of various models to operationally forecast important features of the Baltic Sea dynamics,
a question arises: how accurately can the present 3D numerical models describe the
Baltic Sea physics? What kind of problems might still exist?
Proper numerical modelling of the physics and dynamics of the Baltic Sea is
a great challenge and a very demanding task even for the modern 3D hydrodynamic models because of a number of reasons. These problems and potential shortcomings in the model results should be kept in mind whenever studying particular
model-based analyses or conclusions, especially when interpreting the outcome of
the modelling-oriented chapters of this book. The first task for any 3D model to
solve is the proper description of the stratification. Regrettably, in the Baltic Sea
this has, strictly speaking, not exactly been solved to a satisfactory level. Even the
very best models tend to often exhibit problems to replicate the right position of the
halocline or a realistic vertical gradient of density. There is no unique opinion about
the reasons behind this feature that may substantially undermine the accuracy and
reliability of the simulations of water circulation and spreading of substances over
the entire water column. Fortunately this shortage only insignificantly affects the
results of modelling the velocities and transport in the uppermost layer.