S. Schwegmann & J. Holfort: Baltic sea ice volume 1982-2019 
Meteorol. Z. (Contrib. Atm. Sci.) 
PrePub Article. 2020 
345 
346 
347 
348 
349 
350 
351 
352 
353 
354 
355 
356 
357 
358 
359 
360 
361 
362 
363 
364 
365 
366 
367 
368 
369 
370 
371 
372 
373 
374 
375 
376 
377 
378 
379 
380 
381 
382 
383 
384 
385 
386 
387 
388 
389 
390 
391 
392 
393 
394 
395 
396 
397 
398 
399 
400 
401 
402 
403 
404 
405 
406 
407 
408 
409 
410 
411 
412 
413 
414 
415 
416 
417 
418 
419 
420 
421 
422 
423 
424 
425 
426 
427 
428 
429 
430 
431 
432 
433 
434 
435 
436 
437 
438 
439 
440 
441 
442 
443 
444 
445 
446 
447 
448 
449 
450 
451 
452 
453 
454 
455 
456 
457 
458 
ice (>90 %) and fast ice (100 %). The higher the ice con 
centration is the smaller the intervals arc chosen. The 
same applies for sea ice thickness. 
In addition to the inaccuracies of the data themselves, 
we have also found some inconsistencies in data sam 
plings near the coast. This occurs particularly in the 
southern areas, e.g. in the Vistula Lagoon, where ship 
traffic was not important enough to consider ice situa 
tion in these narrow and shallow waters in the older ice 
charts but are included in the newer charts, which results 
in a positive trend in ice conditions. Fig. 6 shows an ex 
ample for the temporal evolution of ice in the Vistula 
Lagoon, where positive trends in both ASIC and ASIV 
were found, and for the grid cell at the Finnish coast 
near Kaskinen, where ASIC has increased but ASIV has 
decreased. 
For the Vistula Lagoon, it is clearly visible that the 
positive trend in ASIV is caused by some few low values 
in the beginning of the period and some few higher 
values at the end, with nearly no data in between. We 
found the same circumstances for the other grid cells 
showing an increase in ASIV. For the grid cell near 
Kaskinen, on the other hand, data coverage has been 
consistent over the 30 year period and the positive trend 
in ASIC may be a true signal. 
In some regions, SSTs reveal a negative trend, which 
was also found by, e.g., Bradtke etal. (2010), who ar 
gued that the decade 1996 to 2005 shows annual mini 
mum SSTs lower than those in the decade 1986 to 1995. 
Also Siegel et al. (2006) reported slight negative bends 
in SST for February and March. A weaker response of 
SST to the increasing T Lur in winter can be explained 
by oceanographic changes (see Bradtke etal., 2010; 
Meier, 2006), but it is much hairier to explain a negative 
SST trend if at the same time T Lur is increasing and sea 
ice is decreasing (as SST is set to -1 °C if ice is present). 
It might be that the effect only arises from uncertainties 
of the trend statistics or because the SST data may be 
still influenced a bit by the sea ice signals, as the struc 
ture in the bends resemble the mean sea ice disbibution. 
The local atmospheric forcing relevant for the to 
tal volume of sea ice in a winter also depends on the 
larger scale atmospheric patterns. The dominant patterns 
of atmospheric circulation over the north Atlantic and 
the north European area arc the North Atlantic Oscil 
lation (NOA, Jones etal., 1997), the Arctic Oscillation 
(AO, from the National Weather Service Climate Predic 
tion Center: https://www.cpc.ncep.noaa.gov/products/ 
precip/CWlink/daily_ao_index/ao.shtml) as well as the 
Scandinavian pattern (SCAND, from https://www.cpc. 
ncep.noaa.gov/data/teledoc/scand.shtml). High values 
of the NAO Index, with increasing westerly winds, bring 
warmer air into the Baltic region. Therefore a negative 
correlation with the ice volume is expected. As we look 
at the cumulative ice volume sum, we are using the sum 
of the NAO index from January to March and to corre 
late it with the ASIV. As can be seen in Fig. 7, the cor 
relation value varies across the region and reaches max 
imum absolute values above 0.6, but the mean correla 
tion value is -0.40. So, the correlation is not very strong, 
but at least of the right sign. Correlation with the Arctic 
Oscillation gives a similar picture (not shown), with an 
only slightly lower mean correlation value (-0.36). The 
mean absolute value of the correlation with SCAND is 
lower (0.31) but is stronger than with the NAO in the 
Bay of Bothnia. 
All three indices show a decreasing trend in the time 
span used. Considering the negative correlation of NAO 
and AO with the ice volume sum, this would imply an 
increasing ice volume over time whereas we observe 
a decrease in ASIV in most areas. The SCAND trend 
does not conbadict the ASIV trend, so that the long 
term ASIV bend is probably better related to the more 
regional SCAND atmospheric pattern than to the larger 
scale NAO and AO patterns. 
er relate 
ttern thai 
6 Summary and conclusions 
In this study we showed for the first time the regional 
distribution of accumulated sea ice concenbation and 
sea ice volume in the Baltic Sea and their bends from 
1982 to 2019. Both sea ice parameters have been com 
pared to changes in SST and T Lur for the 30 year period 
1982-2011. The conclusion can be summarized as fol 
lows: A 
As expected, the highest mean ASIC and ASIV occur 
in the northernmost regions and decreases towards the 
central, southern and western Baltic Sea. 
Most regions show a decrease in both ASIC and 
ASIV, which generally compares well with the positive 
bends in SST and T Lur . However, both arc not the only 
drivers of sea ice changes, as locally low correlation 
coefficients indicate. But changes in the atmospheric 
patterns described by NOA, AO and SCAND could also 
not better explain the observed changes in ASIV. 
Some few regions reveal a negative bend in SST 
which is not reflected in the sea ice data. In these re 
gions, also the correlation between debended SST and 
ASIC/ASIV is low or even positive. The reason for this 
is still unknown and could also be the result from statis 
tical errors. This has to be considered in more detailed 
investigations in the future. 
The positive trends in most ASIC and all ASIV grid 
cells arc found to be unrealistic. They arc mainly caused 
by a low data record in the early years of the data set, 
where unfortunately also low concentrations and thick 
nesses were observed. All these data points arc close to 
the coast, in areas where ship traffic was certainly not 
important enough to include ice information in the hand 
made historical ice charts, which is expected to be the 
reason for this inconsistent data distribution. 
The regional disbibution of accumulated sea ice con 
centration and sea ice volume in the Baltic Sea will be 
updated after each ice winter. For a geographical bet 
ter resolved product, the ice charts after 2006 can eas 
ily be analyzed with higher spatial resolution. Older ice