4 
S. Schwegmann & J. Holfort: Baltic sea ice volume 1982-2019 
Meteorol. Z. (Contrib. Atm. Sci.) 
PrePub Article. 2020 
204 
205 
206 
207 
208 
209 
210 
211 
212 
213 
214 
215 
216 
217 
218 
219 
220 
221 
222 
223 
224 
225 
226 
227 
228 
229 
230 
231 
232 
233 
234 
235 
236 
237 
238 
239 
240 
241 
242 
243 
244 
245 
246 
247 
248 
249 
250 
251 
252 
Trend ASIC 
(%mVy) 
M-'OO 
-0 50 
I 1-0 23 
HZ]-0.10 
I l-O.OI 
□□ 0.01 -010 
010 -0 25 
0 26 - 0 60 
Hoso-060 
Trend ASIV 
(m/y) 
H<-2 
■■ -2 00- -100 
■I -100- -0 60 
-0 50--0 2S 
□D-02S--010 
I I -0 .10 • -0 01 
-0 01 -001 
CZD 0.01 -010 
CD o 10 - 0 25 
■■026 - 0 60 
Hoso-060 
Figure 3: Linear trend of accumulated sea ice concentration (left) and accumulated sea ice volume over the entire winter for the period 
1982 to 2019. Blue areas show negative, red areas positive trends. The trend in accumulated sea ice concentration varies between -2.98 and 
+3.26 % rrT 2 per year and the trend for accumulated sea ice volume ranges from -2.25 to +1.14 m per year. However, the majority of grid 
cells has a trend between -2 and +0.6 % rrT 2 per year (m per year) for ASIC (ASIV). 
data are available only for this period. To keep the com 
parison consistent, also the impact of air temperature 
changes on sea ice is analysed for this time span. The 
most important influencer is the air temperature (7^,), 
which also links to sea surface temperature (SST). Over 
simplifying, with SST well above freezing, no sea ice is 
present (or is melting). Around the freezing point sea 
ice formation can occur and with sea ice present SST 
is at the freezing point. So there is no direct linear re 
lationship between SST and sea ice, but we expect that 
monthly means arc negatively correlated. Therefore, we 
correlated the detrended SST and winter ASIC (ASIV), 
and, as expected, the correlation is usually negative. In 
regions, where a large amount of ice forms every win 
ter, coefficients of down to -0.9 arc reached. However, 
in regions where ice occurs only in few years (in par 
ticular the central Sea of Bothnia) and along the east 
ern coast of the Gulf of Bothnia, correlation is mostly 
weak. Trends for monthly SST arc shown in Fig. 5 for 
the months January to April. In January, SST increases 
by between 0.1 °C to 0.5 °C per decade in the Gulf of 
Bothnia. However, in the northernmost Bay of Both 
nia, along the coast, a light decrease of up to 0.1 °C 
per decade occurs. Further south, trends he between 
-0.1 °C per decade close to the coasts in the east and 
up to +0.25 °C per decade towards the Swedish coast. 
with the majority of grid cell having trends between 
-8 m and 0 m per decade over the winter season. As for 
ASIC, also in the ASIV product some grid cells show 
an increase. These positive trends arc proven to be not 
realistic (see discussion). 
In order to see, when the largest changes in ASIV oc 
cur within a season, trends were calculated for monthly 
ASIV. The results for the months January to April arc 
shown in Fig. 4. In December and May, any trends arc 
small ah over the Baltic Sea. From February to April, 
ASIV decreases strongest in the Bay of Bothnia while in 
the Sea of Bothnia, the highest negative trends occur in 
March and April. In the Gulf of Finland and Riga, ASIV 
changes show hardly any seasonal variability. Monthly 
trends vary between -3.5 m and +2.2 m per month. 
4 Corresponding changes in sea 
surface temperature and 
meteorological conditions 
from 1982 to 2011 
Basically, changes in sea ice concentration and sea ice 
thickness arc caused by atmospheric and oceanographic 
conditions. Here we analyse the impact of these condi 
tions on sea ice for the period 1982 to 2011, as the SST