8 Trajectories and Spreading of Observed and Simulated Drifters in the Baltic Sea
277
water particle, would, on average travel ∼125 km in those 10 days. By the same
argument, if simulated particles are estimated to reach the coast in 10 days, real water particles would make the same journey in <8 days. Furthermore, water particles
contained within a <1 km radius initially would spread over an area of 12 km radius
in 25 days, while their model counterparts would disperse to cover <10 km radius.
Such conclusions would have impacts when estimating the fate of oil spills or other
pollutants. However, we wish to stress that we have compared observations from
2010–2011 with model data from 1962–2004, and that the drifter data is limited.
To make the comparison of modelled and observed properties fair, and to yield
more confidence in the magnitudes of the tuning needed would require model data
for the years 2010–2011. This could, however, take a few years as the RCO model
is being decommissioned in favour of a new regional ocean model, based on the
Nucleus of European Modelling (NEMO) ocean circulation model (Madec 2009).
The wind forcing will most likely also need updating, as the ERA-40 data set is
to be replaced by ERA-Interim and eventually by ERA-75, while the RCA model
used for downscaling is also likely to be decommissioned. If SVP-B drifter data is
continued to be gathered for the Baltic Sea, this data could be used to validate and
perhaps tune the next generation of Baltic Sea models.
The presented results from the drifters in the uppermost layer of the Gulf of
Finland indicate substantial differences in the dynamics of the vertically integrated
relatively thick subsurface layer and the uppermost layer with a thickness of 1–
1.5 m (which is where, e.g., oil spills or smaller lost items are transported). Although
the average spreading rate generally increased with time or the distance between
drifters, the well-known Richardson’s law did not satisfactorily explain the transport
in the uppermost layer of the Gulf of Finland. For the separation d of drifter pairs
on short time scales (less than 8 hours) or separation distances (from the first tens
of meters up to about 100–150 m), a power law d ∼ t 0.27 described the spreading
much better. Starting from this threshold, the distance then increased, in average,
according to a power law d ∼ t 2.5 . The spreading rate was about 200 m/day for
separations below 0.5 km, 500 m/day for separations below 1 km and in the range
of 0.5–3 km/day for separations in the range of 1–4 km.
As a considerable part of the drifters following the uppermost 1.5 m thick layer
was above the water surface, their drift was impacted by wind properties to some
extent. For example, a wind speed of 5 m/s may, technically, yield a contribution of
about 10 cm/s to the drift speed (Soomere et al. 2011). Although this value is on the
order of the current speed, it does not significantly effect the separation of drifters,
as the wind patterns over sea surface are much more homogeneous compared to
similar winds over the mainland. Therefore, it is natural to expect that the impact of
wind on closely located drifters mostly resulted in their concurrent downwind drift.
This would imply that winds can influence absolute dispersion while their impact
on relative dispersion would be negligible.
The results from the drifters in the uppermost layer suggest that a realistic parameterization of subgrid-scale processes in the Gulf of Finland strongly depends on the
resolution of the ocean model. It is well known that models with spatial resolution
coarser than 2 km cannot resolve mesoscale dynamics in this region (see Chap. 6
277
water particle, would, on average travel ∼125 km in those 10 days. By the same
argument, if simulated particles are estimated to reach the coast in 10 days, real water particles would make the same journey in <8 days. Furthermore, water particles
contained within a <1 km radius initially would spread over an area of 12 km radius
in 25 days, while their model counterparts would disperse to cover <10 km radius.
Such conclusions would have impacts when estimating the fate of oil spills or other
pollutants. However, we wish to stress that we have compared observations from
2010–2011 with model data from 1962–2004, and that the drifter data is limited.
To make the comparison of modelled and observed properties fair, and to yield
more confidence in the magnitudes of the tuning needed would require model data
for the years 2010–2011. This could, however, take a few years as the RCO model
is being decommissioned in favour of a new regional ocean model, based on the
Nucleus of European Modelling (NEMO) ocean circulation model (Madec 2009).
The wind forcing will most likely also need updating, as the ERA-40 data set is
to be replaced by ERA-Interim and eventually by ERA-75, while the RCA model
used for downscaling is also likely to be decommissioned. If SVP-B drifter data is
continued to be gathered for the Baltic Sea, this data could be used to validate and
perhaps tune the next generation of Baltic Sea models.
The presented results from the drifters in the uppermost layer of the Gulf of
Finland indicate substantial differences in the dynamics of the vertically integrated
relatively thick subsurface layer and the uppermost layer with a thickness of 1–
1.5 m (which is where, e.g., oil spills or smaller lost items are transported). Although
the average spreading rate generally increased with time or the distance between
drifters, the well-known Richardson’s law did not satisfactorily explain the transport
in the uppermost layer of the Gulf of Finland. For the separation d of drifter pairs
on short time scales (less than 8 hours) or separation distances (from the first tens
of meters up to about 100–150 m), a power law d ∼ t 0.27 described the spreading
much better. Starting from this threshold, the distance then increased, in average,
according to a power law d ∼ t 2.5 . The spreading rate was about 200 m/day for
separations below 0.5 km, 500 m/day for separations below 1 km and in the range
of 0.5–3 km/day for separations in the range of 1–4 km.
As a considerable part of the drifters following the uppermost 1.5 m thick layer
was above the water surface, their drift was impacted by wind properties to some
extent. For example, a wind speed of 5 m/s may, technically, yield a contribution of
about 10 cm/s to the drift speed (Soomere et al. 2011). Although this value is on the
order of the current speed, it does not significantly effect the separation of drifters,
as the wind patterns over sea surface are much more homogeneous compared to
similar winds over the mainland. Therefore, it is natural to expect that the impact of
wind on closely located drifters mostly resulted in their concurrent downwind drift.
This would imply that winds can influence absolute dispersion while their impact
on relative dispersion would be negligible.
The results from the drifters in the uppermost layer suggest that a realistic parameterization of subgrid-scale processes in the Gulf of Finland strongly depends on the
resolution of the ocean model. It is well known that models with spatial resolution
coarser than 2 km cannot resolve mesoscale dynamics in this region (see Chap. 6