2 
S. Schwegmann & J. Holfort: Baltic sea ice volume 1982-2019 
Meteorol. Z. (Contrib. Atm. Sei.) 
PrePub Article. 2020 
60 
61 
62 
63 
64 
65 
66 
67 
68 
69 
70 
71 
72 
73 
74 
75 
76 
77 
78 
79 
80 
81 
82 
83 
84 
85 
86 
87 
88 
89 
90 
91 
92 
93 
94 
95 
96 
97 
98 
99 
100 
101 
102 
103 
104 
105 
106 
107 
108 
109 
110 
111 
112 
113 
114 
115 
116 
117 
118 
119 
120 
121 
122 
123 
124 
125 
Figure 1: Mean values of accumulated sea ice concentration (left) and accumulated sea ice volume (middle) over an entire winter, averaged 
over the period 1982 to 2019. Highest values (blue) occur in the northern parts were a compact sea ice cover exists nearly throughout the 
entire winter every year. Lowest values (red) are found in the southern and western Baltic Sea, where sea ice at sea occurs only in strong 
winters. The right panel shows frequency of ice occurrence in % of all winters. In the northern Bay of Bothnia, nearly every year ice occurs 
(red fields) while in the southern central parts, ice occurs only in some few years with severe winters. 
ice charts, which were mostly drawn twice per week for 
the entire Baltic Sea over the winter season until 2011. 
Afterwards, ice charts were drawn at least once per week 
and in times with rapid sea ice changes twice per week. 
All the ice charts are based on visual observations from 
trained ice observers, on ship-based observations, on air 
borne observations (in former years) and on satellite data 
since about three decades. In the ice charts, sea ice con 
centration and sea ice thickness is not given in absolute 
values but rather in intervals. That means, ice concen 
tration is given in terms like 10-30 % and sea ice thick 
ness in intervals like 5-10 cm. For the regular grid, these 
values were averaged, i.e. for the mentioned example, 
sea ice concentration would be 20 % and sea ice thick 
ness 7.5 cm. The charts before 2005 have been manu 
ally digitalized onto a regular 0.5° x 0.5° grid (Feistel 
etal., 2008). Newer charts are available as vector data 
and were mapped in an automated fashion onto the same 
grid. Due to the better resolution of the newer charts, 
inconsistencies may appear sometimes in narrow, near 
coastal sections like the Vistula Lagoon, as the newer 
charts captured this near coastal ice, which (if present in 
the charts) were often not taken into account in the man 
ual gridding process. This data has already been used in 
a comparison between maximum ice volume and max 
imum ice extent for the Baltic (Schmelzer and Hol- 
fort, 2014), and both time series showed in general 
good correlation. However, in some years interesting 
differences appeared which showed us that it is reason 
able to study sea ice volume, too. 
In this study we use only data covering the period 
1982 to 2019, as ice thicknesses have not been collected 
consistently before 1982. So, in order to avoid high 
sea su 
uncertainties in the accumulated sea ice volume fields, 
e 1982 have not been taken into consideration, 
comparison of sea ice changes to those of the 
rface temperature only data from 1982-2011 arc 
used, as the SST data (described below) is only available 
for this time span. We analyzed the local cumulative 
ice concentration and ice volume sum, which arc the 
integrals over the season of the ice concentration and 
volume, respectively, here to be named as accumulated 
sea ice concentration (ASIC) and accumulated sea ice 
volume (ASIV). The idea of the cumulative ice volume 
sum was introduced by Koslowski (1989). Using the 
accumulated values instead of mean or maximum values 
for ice concentration and ice volume, we believe to 
get a more reliable mean for the winter severity. To 
account for the irregular sampling times (once or twice 
per week only), data between two subsequent ice chart 
dates were linearly interpolated to daily data and were 
then summed up over the entire winter or over single 
months. The ASIC is expressed as percent per square 
meter and the ASIV in volume per square meter, so its 
dimension is meter. The distribution of averaged ASIC 
and ASIV over the entire winter season is shown in 
Fig. 1 and is combined with the information of ice 
frequency, indicating how often a grid cell has been 
covered by ice in the observation period. Mean values 
for ASIC (ASIV) over the entire winter range from 
0.5 percent m 2 (0.5 m) at the German Baltic coast to 
about 152 percent m 2 (75 m) in the northern paid of the 
Bay of Bothnia where nearly the entire winter is covered 
by a compact sea ice cover. For single months, mean 
values of up to 30 percent m 2 (18 m) arc found (not 
shown). The temporal variability of ASIC and ASIV can