Baltic Sea Ice Climate Workshop 2005 
50 
Nevertheless, most of the fully insecure cases in the period 1401-1500 are weak or moderate 
ice winters with an ice winter index of 0 or 0.1 and without any remarkable features, Figure 5. 
Therefore, the data series for the period 1401-1500 can be analysed by applying the Gaussian 
low-pass filter with a possible margin of error, Figure 6. 
1300 1350 1400 1450 1500 
Fig. 5. Ice winter index numerals 1301-1500. Most uncertain cases in the period 1401 to 1500 
are normal ice winters with an ice winter index of 0 and 0.1. 
Fig. 6. Ice winter index numerals 1401-1500. The solid curves represent smoothed ice winter 
index numerals series obtained by applying a Gaussian low-pass filter with a 20- year cut-off 
period. The shaded areas denote the possible margin of error. 
The ice winter index data series from 1401 
Figure 7 demonstrates the long-term variations in the ice winter index numerals in the period 
1401-1700. The thin and heavy solid curves represent smoothed ice winter index numerals 
series obtained by applying a Gaussian low-pass filter with a 20- and 40-year cut-off period, 
respectively. The horizontal solid line denotes the arithmetic mean of 0.529 for the period 1401- 
2005.