limit for icing. The diagrams were constructed from the sea level temperatures by 

 assuming a moist adiabatic lapse rate above the sea surface. This assumption is a 

 reasonable one in cases of fresh winds, showery weather conditions, cyclonically 

 curved isobars and for air which is being heated from below. When air is cooled 

 from below or when the winds are light, the average lapse rate is probably less than 

 the moist adiabatic. In general, any evidence of atmospheric stability will mean 

 that the existing lapse rate will be less than the moist adiabatic. Whenever the lapse 

 rate is less than the moist adiabatic, the computed height of the 0° C isotherm will 

 always be too low. 



In Figure 25 the mean heights are probably approximately correct, due to 

 the strong winds and frequent deep cyclonic systems in winter. In low latitudes 

 near the subtropical anticyclone, the estimated heights may be too low. In Figure 

 27, when we are considering direct outbreaks of polar continental air, the values 

 should be approximately correct. 



During the summer months, general subsidence in the subtropical anti- 

 cyclone and light winds will result in a lapse rate which is less than the moist adia- 

 batic. An attempt has been made to apply a correction to the summer values. 

 This was achieved by extrapolating from the 10,000-foot temperatures at New- 

 foundland. On the basis of the Newfoundland check, the corrected height lines 

 were drawn on Figure 26. In all likelihood these average figures are the least 

 accurate of the four icing charts. In Figure 28 the moist adiabatic lapse rate was 

 taken as the average lapse rate. This appears a reasonable assumption for those 

 situations when direct outbreaks of polar air occur in summer. 



Only the lower limit of icing has been indicated since the upper limit will 

 depend greatly on the particular synoptic situation. An approximate temperature 

 of —15° C might be assumed as a lower temperature limit, and with this tempera- 

 ture we should find the icing region to be about 8,000 feet thick. 



In order to illustrate the figures presented in the above diagrams, let us 

 consider the cloudy region 55° N 30° W. In winter the average height of the lower 

 limit of icing would be about 4,000 feet with extreme low values of about 1,000 feet. 

 In summer the average height may be about 10,000 feet with extreme low values 

 of about 6,000 feet. 



Due to the complex nature of the icing problem, no attempt has been made 

 to estimate the severity of the icing. On the other hand, more severe icing would, 

 in general, be expected in winter because of the frequent occurrence of unstable 

 weather conditions. 



12 



