60
K. Myrberg and A. Lehmann
The reasoning behind this problem is manifold. The existing knowledge of the
water budget (in- and outflowing waters, precipitation, evaporation and river runoff),
albeit based on quite a large data set, is not very accurate. An implication from
the listed uncertainties is that the water cycle is not ‘closed’. There is always a
source of errors, which at times become evident as systematic bias in the water level
in operational models (Lagemaa et al. 2011). The performed measurements of the
different constituents of the water budget are not uniform. Moreover, they are highly
unsynchronized and biased, and have very different temporal and spatial resolution.
The meteorological data used for model forcing are often inaccurate and have a
low horizontal resolution, which does not allow a proper description of open sea
meteorological conditions and the land-sea differences due to too few grid points.
This problem still remains today even if the present operational models have high
resolutions, but there is a lack of reanalyses, which are necessary for long-term
hindcasts of spatial distributions of specific fields generated by the current driven
transport as described in Chaps. 4, 5 and 9–11.
A more generic problem is the description of vertical turbulence. Usually the existing and widely used parameterizations are tuned based on the data from the open
oceans (where the stratification conditions greatly differ from those in the Baltic) or
extracted from laboratory experiments that only partially replicate the complexity
of the marine environment. The choice of the proper (potentially spatially and vertically varying) parameters of the numerical scheme still remains a major challenge
for the Baltic Sea modelling. For these reasons also the closure of the energy budget
is a complicated task. This is not only due to the inaccuracies in the wind forcing but also due to challenges in the parameterizations of heat and moisture fluxes
in the rapidly varying atmospheric stability conditions, abrupt changes in surface
roughness conditions in the complicated coastal areas and due to local, instantaneous changes in sea surface temperature due to upwelling. Additional problems for
closing the energy budget appear because of the sparse network of radiation measurements and the locally varying surface albedos. Frankly speaking, the present
forcing data for models still suffers from several uncertainties, which are reflected
in the model results as well.
High demands exist for the simulation of the dynamics of currents—flow fields,
current-driven transport and spreading processes—which is from the pure modelling
perspective the main challenge of the technology presented in this book. In addition
to the previously mentioned problems with meteorological forcing, even more serious problems arise. The small baroclinic Rossby radius of deformation requires
high-resolution modelling in both horizontal and vertical directions. Together with
the request to perform calculations over at least several years, this leads to high computational requirements. In order to fulfil the two latter requirements, a nested grid
approach is often used, where a high-resolution local model get its open boundary
conditions from a large-scale model (the whole Baltic, Andrejev et al. 2011, or the
Baltic Sea–North Sea, Lu et al. 2012). These technical difficulties apparently can be
partially resolved thanks to parallelized model codes and ever faster computers.
There is, however, room for many more issues. As the Baltic Sea is one of the
most studied basins of the World Ocean, one might expect that the relevant basic information such as high-resolution bottom topography is readily available. The
K. Myrberg and A. Lehmann
The reasoning behind this problem is manifold. The existing knowledge of the
water budget (in- and outflowing waters, precipitation, evaporation and river runoff),
albeit based on quite a large data set, is not very accurate. An implication from
the listed uncertainties is that the water cycle is not ‘closed’. There is always a
source of errors, which at times become evident as systematic bias in the water level
in operational models (Lagemaa et al. 2011). The performed measurements of the
different constituents of the water budget are not uniform. Moreover, they are highly
unsynchronized and biased, and have very different temporal and spatial resolution.
The meteorological data used for model forcing are often inaccurate and have a
low horizontal resolution, which does not allow a proper description of open sea
meteorological conditions and the land-sea differences due to too few grid points.
This problem still remains today even if the present operational models have high
resolutions, but there is a lack of reanalyses, which are necessary for long-term
hindcasts of spatial distributions of specific fields generated by the current driven
transport as described in Chaps. 4, 5 and 9–11.
A more generic problem is the description of vertical turbulence. Usually the existing and widely used parameterizations are tuned based on the data from the open
oceans (where the stratification conditions greatly differ from those in the Baltic) or
extracted from laboratory experiments that only partially replicate the complexity
of the marine environment. The choice of the proper (potentially spatially and vertically varying) parameters of the numerical scheme still remains a major challenge
for the Baltic Sea modelling. For these reasons also the closure of the energy budget
is a complicated task. This is not only due to the inaccuracies in the wind forcing but also due to challenges in the parameterizations of heat and moisture fluxes
in the rapidly varying atmospheric stability conditions, abrupt changes in surface
roughness conditions in the complicated coastal areas and due to local, instantaneous changes in sea surface temperature due to upwelling. Additional problems for
closing the energy budget appear because of the sparse network of radiation measurements and the locally varying surface albedos. Frankly speaking, the present
forcing data for models still suffers from several uncertainties, which are reflected
in the model results as well.
High demands exist for the simulation of the dynamics of currents—flow fields,
current-driven transport and spreading processes—which is from the pure modelling
perspective the main challenge of the technology presented in this book. In addition
to the previously mentioned problems with meteorological forcing, even more serious problems arise. The small baroclinic Rossby radius of deformation requires
high-resolution modelling in both horizontal and vertical directions. Together with
the request to perform calculations over at least several years, this leads to high computational requirements. In order to fulfil the two latter requirements, a nested grid
approach is often used, where a high-resolution local model get its open boundary
conditions from a large-scale model (the whole Baltic, Andrejev et al. 2011, or the
Baltic Sea–North Sea, Lu et al. 2012). These technical difficulties apparently can be
partially resolved thanks to parallelized model codes and ever faster computers.
There is, however, room for many more issues. As the Baltic Sea is one of the
most studied basins of the World Ocean, one might expect that the relevant basic information such as high-resolution bottom topography is readily available. The