Baltic Sea Ice Climate Workshop 2005 
52 
Table 2. Means and variances of the ice winter index numerals, ice winter severity related to 
the arithmetic mean of 0.529 for the time from 1401-2005, and number of normal and 
anomalous ice winters per decade in specific time intervals. 
Period 
Number 
of years 
Ice winter index 
numerals 
Ice winter 
severity related 
to mean 
ice winter index 
1401-2005 
Number of ice winters 
per decade 
Means 
Variances 
Normal Anomalous 
1401-1424 
24 
0.342 
0.749 
Reduced 
8.33 
1.67 
1425-1443 
19 
0.816 
1.137 
Increased 
5.79 
3.68 
1444-1459 
16 
0.231 
0.294 
Reduced 
8.12 
1.25 
1460-1468 
9 
0.633 
0.958 
Increased 
5.56 
3.33 
1469-1503 
35 
0.426 
0.471 
Reduced 
5.71 
1.71 
1504-1519 
16 
0.725 
0.847 
Increased 
5.00 
4.37 
1520-1547 
28 
0.339 
0.319 
Reduced 
6.43 
1.79 
1548-1578 
31 
0.639 
0.864 
Increased 
5.48 
3.23 
1579-1592 
14 
0.293 
0.321 
Reduced 
7.14 
1.43 
1593-1634 
42 
0.679 
0.754 
Increased 
4.76 
3.33 
1635-1656 
22 
0.355 
0.371 
Reduced 
6.36 
1.82 
Total period 
1401-2005 
605 
0.529 
0.683 
5.80 
2.45 
The severity of ice winters in the period 1301-1400. Anomalous ice winters. 
Because of the large number of insecure reconstructed cases in the period 1301-1400, the 
severity of ice winters has been derived from the number of anomalous winters. Koslowski and 
Glaser (1999) defined anomalous ice winters as follows: 
"The severity types weak and moderate, (index 0 and 0.1) with a frequency of 70% in the period 
1501-1995, which lead only to ice formation in the inner coastal waters, are the norm. Strong + , 
very strong and extreme severity types (index >1 with a frequency of 25.2 %), causing total ice 
cover in the Western Baltic with an ice thickness greater than 20 cm, can be regarded as 
anomalous". 
In Figure 9, the occurrence of anomalous ice winters (anomalous ice winter = 1, normal ice 
winter = 0) is shown for the period 1301-2005. The thin and heavy solid curves are smoothed 
frequency curves obtained by applying a Gaussian low-pass filter with 20- and 40-year cut-off 
periods, respectively. The horizontal solid line denotes the arithmetic mean of 0.228 for the 
period 1301-2005. The areas above the mean line indicate phases with an increased number of 
anomalous ice winters (increased ice winter severity) and are comparable to those obtained for 
the period 1401-2005 as shown in Figure 8.