12 



BJ0RN HELLAND-HANSEN 



[REP. OF THE "AUCHAEL SARS" NORTH 



five of a high temperature in water and air [Helland- 

 Hansen 1916]. 



During the "Michael Sars" Expedition fog was met 

 with several times. It is noted in the Tables on June 

 30th, July 1st, 2nd, 8th to 11th (all of them near New- 

 foundland), 18th, 26th, and August 7th. The observations 

 of fog are marked in the Figures on pp. 78* and 80-' 

 above the A-curves, and it will readily be seen, that the 

 fog appeared almost exclusively when the temperature of 

 the air was higher than that of the sea surface. The 

 only exception was on the outskirts of the Labrador Current 

 on the 11th of July for a couple of hours when a north- 

 westerly wind probably brought fog. '^w\Atn{\y the occurrence 

 of fog is chiefly connected with a prominent stable equili- 

 brium in ''old sea-air", i. e. air which has passed over 

 relatively warm parts of the ocean for such a long time 

 that it has attained a high value not only of absolute, 

 hut also of relative humidity to a greater height than 

 that reached by the turbulent motion in the place where 

 the fog is formed. If not, the wind observed on foggy 

 days would dissipate it. In other words, it is not only 

 a question of the difference of temperature between water 

 and air, but also of the vertical distribution of moisture 

 in relation to the actual temperature of the sea surface. 

 The greater the negative temperature gradient and the 

 longer the previous air-path over a warm part of the 

 ocean are, the more favourable will be the conditions for 

 the formation of thick fog. The unusually great contrast 

 between warm and cold areas near the Newfoundland 

 Bank is apt to cause unusually thick fogs in this region. 

 In other regions where fog frequently occurs as e. g. in 

 the Northern part of the Norwegian Sea (at Spitsbergen 

 etc.) the contrast is much smaller, and we may therefore 

 suppose that the fog in such region is, as a rule, not so 

 dense, h must be remembered that the air may often 

 be supersaturated with water-vapour before the conden- 

 sation takes place, and the actual conditions for the 

 formation of water-drops (the existence of nuclei) must 

 therefore be considered too. It may be possible that 

 there is a connection between strong wind (much spray 

 giving minute water-drops or salt-particles to the atmos- 

 phere) and a subsequent formation of fog. -- 



We shall now turn to the B-curves, demonstrating 

 casual and short periodic variations in humidity. It is 

 evident that even these variations were common to the 

 absolute and the relative humidity in the great majority 

 of cases. This means that the general conclusions arrived 

 at in regard to the averages also hold good in regard 

 to the more detailed variations, within short intervals of 

 time. The correspondence with the variations both in 

 air temperature and in the temperature-difference between 



water and air is also quite evident; the divergencies may 

 in most cases be explained by the wind-conditions. 



In most cases, then, we also find a maximum of 

 absolute as well as of relative humidity by day and a 

 minimum by night corresponding to the daily period of 

 the air temperature. There are many exceptions from 

 this rule, but it holds good in the large majority of cases. 

 /// the middle of the day (in summer) the air was mostly 

 warmer than the sea (cf. sect. 8) so that a stabile 

 equilibrium was established, with an increase of humidity 

 near the sea surface. At night the reverse was as a 

 rule the case, giving labilitv with a decrease of humidity 

 near the surface. Many of the exceptions are easily 

 explained when the various elements are examined in 

 detail. 



10. Cloudiness. 



The observations of clouds were restricted to an 

 estimation of the extent of the cloudiness, and did not 

 embrace the cloud-forms. It is not possible, therefore, 

 to treat the variations in cloudiness in a satisfactory manner. 



In the 3rd series of observations, between the New- 

 foundland Banks and the British Isles (curve p. 80-), the 

 observations were so scanty and doubtful within the foggy 

 region near Newfoundland that the averages for two or three 

 days could not be calculated. The sea temperature 

 increased on the whole until the 18th of July, as did also, 

 correspondingly, the air temperature; but on a wide 

 average the cloudiness continued to sink until the 17th. 

 Taking the minor variations, however, it is quite obvious 

 that the cloudiness varied with the variations of temperature, 

 and still more with the humidity. An increase of temperature 

 and humidity corresponded to an increase of cloudiness, 

 and a decrease of the various quantities was mutual. There 

 was obviously some retardation of the variations from air 

 temperature to humidity and further to cloudiness. 



This parallelism may probably be accounted for by the 

 fact that temperature and humidity increase with the 

 beginning of southerly winds in front of the depression, 

 and that cloudiness soon increases loo on the further 

 approach of the depression. 



The cloudiness in depressions has been explained by 

 the polar front theory of Professor V. Bjekknes and his 

 collaborators. This theory sets out from the fact that the 

 temperature distribution in depressions often exhibits a 

 prominent line of discontinuity forming the boundary 

 between air of polar and tropical origin. This boundary 

 line, the "polar front", corresponds in the free atmosphere 

 to a boundary surface directed obliquely upwards with 

 an inclination from the horizontal of about 1 : 100. All 

 the clouds giving continuous rain (ASt and high Nb) seem 



