Fourth Study Conference on BALTEX Scala Cinema Gudhjem

- 85 -
Influence of Atmospheric Forcing on Simulations with a General
Circulation Model of the Baltic Sea
Frank Janssen, Torsten Seifert
Baltic Sea Research Institute, Seestrasse 15, D-18119 Rostock, Germany, frank.janssen@io-warnemuende.de
1. Introduction
Numerical models play an important role in enhancing our
knowledge of the functioning of the Baltic Sea system. A
hierarchy of models from simple box-models to
sophisticated 3D general circulation models has been
applied to all kinds of problems and the influence of the
chosen model type on the results has often been discussed.
But irrespective of the model type selected the results
depend strongly on the upper and lateral boundary
conditions. This study deals with the influence of the upper
boundary condition, i.e. the atmospheric forcing, on multi
year simulations of the Baltic Sea physical environment with
a 3D circulation model. Four atmospheric datasets have been
selected and the ocean model is forced with the different
atmospheric data during the 10 year simulation period 1981-
1990. An intercomparison of the simulation results as well
as a validation with observations is carried out.
2. Material and Methods
From the several available atmospheric datasets which cover
the Baltic four have been selected for this study. Two of
them (ERA-15, NCEP) are extracted from global re-analysis
datasets. A regionalized version of the ERA-15 data comes
from the regional atmospheric REMO and the fourth
candidate is the dataset compiled within the BALTEX
project.
The ocean model is the general circulation model MOM3
adapted to the Baltic Sea with a horizontal resolution of
3’ x 6’ and 77 levels.
3. Discussion
Figure 1 shows a comparison of 2m air temperature and
wind speed in the central Baltic Sea. Both datasets have a
pronounced annual cycle in wind speed but the REMO data
are more than 1m/s higher throughout the year. In contrast to
the wind speed the air temperature is significantly lower in
the REMO dataset with the largest differences during winter.
[m/s]
S
W
]
[°C
T AIR
10
5
2
1
0
15
10
5
-2 0
1 2 3 4 5 6 7 8 9 10 11 12
REMO
ERA-15
REMO - ERA-15
mean annual wind speed
mean annual air temperature
1 2 3 4 5 6 7 8 9 10 11 12
time [month]
Figure 1: Mean annual cycle during the period 1980-1990
of wind speed and air temperature in the central Baltic Sea
(57°N, 20°E) from two different forcing datasets.
Both atmospheric variables contribute to the oceanatmosphere
heat fluxes and therefore a strong impact on the
ocean surface can be anticipated. The salinity distributions
in figure 2 clearly indicate that the influence is not
confined to the ocean surface layer.
y sa _ y era15_bio_hv_p2c
p a _J0 pos o 5 8 , 0 3
ra15_bio_hv_p2c era15_bio_hv_p2c
remo_bio_hv_p2c remo_bio_hv_
12
12
12
12
50
12
12
50
50
50
50
50 11
11
11
11
100
100
10 100 10 100
10 100
10
150
150 9
150 9
150
9
150 9
200
200
8
7
200
8
7
200
8
7
200
8
7
83 84 80 85 81 86 82 87 83 88 89 84 90 85 80 86 81 87 82 88 83 89 84 90 85
80
86
80 81
87
82
88
83
89
84
90
85 86 80 87 81 88 82 89 83 90 84 85 80 86 81 87
depth [m]
depth [m]
remo_bio_hv_p2c remo_bio_hv_p2c - era15_bio_hv_p2c - era15_bio_hv_p2c
depth [m]
depth [m]
[psu]
S [psu]
[psu]
S [psu]
hv_p2c - era15_bio_hv_p2c
2
[psu]
depth [m]
50
50
1
-0.2
1
-0.2
1
-0.2
100
100
0
-0.4
0
-0.4
0
-0.4
150
150
-0.6
-0.6
-0.6
200
200
-1
-0.8
-1
-0.8 -1
-0.8
-2
-2
83 84 80 85 81 86 82 87 83 88 89 84 85 86 87 88 89 90
80 81 -2 82 83 84 85 86 87
time [a] time 90 85 86 87 88 89 90
80 [a] 81 82 83 84 85 80 86 81 8782 888389849085 86808781 -2
83 84 85 86 87 88 89 90
80 81 82 83 84 85 86 87 88 8882 89 8983 90 9084
85 86 87
time [a]
80 81 82 83 84 85 86 87 88 89 90
time [a]
time 80 [a] 81
time [a]
time [a]
80 81
[psu]
2
depth [m]
[psu]
∆ S
bott
2
[psu]
depth [m]
S [psu]
[psu]
∆ S
bott
Figure 2. Upper panel: Salinity in the central Gotland
Basin (BMP Station J01) from two model runs. Lower
panel: Difference between the two simulations.
4. Conclusions
Preliminary model results show that the atmospheric
forcing has a strong impact on the quality of the
simulation. This impact affects not only the upper layer
but the whole system. A comprehensive analysis of the
forcing is therefore needed in order to decide which
properties of a forcing dataset are reliable before the
parameterizations of the ocean model are re-calibrated.
Furthermore, not only the differences between the type of
the used models should be discussed when model results
are compared. Differences in the forcing data should be
discussed with the same emphasis.