ATLANT. DEEP-SEA EXPED. 1910. voi.. i) PHYSICAL OCEANOGRAPHY AND METEOROLOGY 



11 



With increasing southerly winds in the North Atlantic 

 we generally find an extra increase of relative iuiniidity 

 (cf. e. g. June 24th- 26th, 29th, July 13th, 20th, 26th). 

 It must, of course, be borne in mind that our observations 

 only comprise certain parts of the North Atlantic and that 

 we lack observations from some important areas. Our 

 conclusions are therefore limited in regard to locality as 

 well as time. But for those parts of tiie Ocean that we 

 have investigated in the "Miciiael Sars" expedition we find 

 tiiat as far as the averages are concerned (the A-curves) 

 the variations in relative humidity did upon the whole 

 go parallel to the variations in xaind velocity with 

 southerly winds and inversely with northerly winds, 

 the variations being generally felt earlier in the wind 

 than in the humidity. 



In ordcT to explain these conditions we must call 

 attention to the conditions of equilibrium in the lowest part 

 of the troposphere. By vertical movements of the air its 

 temperature varies adiabatically. When the air ascends it 

 becomes cooled by 1°C pr. 100m. (the "dry adiabatic"), pro- 

 vided that the water-vapour does not condense and liber- 

 ate its latent heat of evaporation. When the air is perfectly 

 mixed we find below the level of condensation a vertical 

 temperature gradient corresponding to the dry adiabatic. 

 A stronger gradient (a quicker decrease of temperature 

 upwards) means a state of lability, and a smaller gradient 

 a state of stability. As the sea surface is upon the whole 

 warmer than the air, we generally find that the air above 

 the ocean is heated from the under boundary, /. e. there 

 is a state of instability in the lowest layer of the tropos- 

 phere. When the air is stirred by winds it ascends, and 

 up to the height of its ascension it tends to produce a 

 gradient corresponding to the dry adiabatic. This height 

 will, with continuous spells of wind, ultimately become the 

 height of condensation, or at any rate a height above which 

 a smaller gradient (may be inversion) rules. By this ascen- 

 sion the water vapours are carried upwards. The air 

 ascending from the sea surface must be replaced, and 

 the new air coming near the sea surface as a substitute 

 must obviously as a rule have a relatively low humidity 

 when the air is colder than the sea. With a positive 

 temperature gradient in vertical direction (temperature 

 decreasing upwards) the absolute humidity must upon 

 the whole decrease upwards, the more so as the moisture 

 comes from the sea surface. Aerological observations 

 have shown that even the relative humidity very often 

 decreases upwards. As long as the heating from the sea 

 surface takes place the turbulence is apt to provoke a 

 reduction in the absolute humidity observed at quite low 

 heights and an increase at higher levels. 



In the case of a labile equilibrium the descending air 

 doubtless reduces the temperature observed, but as long 



as a heating of the advancing air really takes place the 

 relative humidity must sink, provided that the air (as is 

 generally the case) has not been saturated all through. 



Such conditions are especially prevalent with northerly 

 winds in our areas of investigation. With northerly winds, 

 therefore, we may lay down as a general rule that the 

 records of air-temperature give sinking values at the same 

 time as the observed difference of temperature between 

 water and air increases and both the absolute and the 

 relative humidity observed decrease. 



With southerly winds (in the sense defined above) the 

 air is cooled from below so that an increasing state of stabi- 

 lity is established. The temperature increases from the sur- 

 face and for some distance upwards, and provided that we 

 have to deal with "old" sea-air and not relatively dry air 

 from the continents or air which has recently come from 

 Arctic regions, the relative humidity will increase. The 

 absolute humidity will eo ipso be greater in warm sea-air 

 from the south than in cold air from the north. In a state of 

 pronounced stability the mixing of tiie air proceeds more 

 slowly upwards than in the case of lability, thus limiting 

 the effect to lower heights. In this manner the increase 

 of absolute and relative humidity is intensified in the 

 lowest air-layer. 



In his important memoir on the "Scotia"-observations 

 in 1913 G. I. Taylor |1914] has arrived at some very 

 interesting results, obtained by means of kite and balloon 

 ascents. He studied the "life-history" of the air in which 

 his aerologic observations were made, and found a close 

 correspondence between the path of the air along the 

 sea surface during the preceding days, the surface tempe- 

 rature, and the vertical distribution of air temperature 

 and humidity. How far, and how rapidly the influence 

 of the sea surface and the casual wind will make itself 

 felt upwards depends upon this "life history", because it 

 is the decisive factor governing the variations in the 

 vertical gradients and the state of equilibrium in the lowest 

 part of the troposphere. 



In calm weather, the conditions are quite peculiar. 

 When the sea is warmer than the air, this ascends 

 with its content of water vapours, and the evaporation 

 may then go on permanently because new unsaturated 

 air appears at the sea surface. When, on the other hand, 

 the air is warmer than the sea, the state of stability is 

 apt to become more and more prominent. A film of 

 air along the sea surface may become saturated witli 

 moisture and fail to ascend on account of its compara- 

 tively low temperature. As a result, evaporation will 

 cease. Such a case may easily occur in summer time, 

 especially in the middle of the day. Under these circum- 

 stances the evaporation may have a minimum irrespec- 



