Evaporation from the Surface of the Sea and the Water Budget of the Earth 221 



If 30%o < s < 40%o and — 2°C < t^ < 30°C, then with sufficient accuracy 

 pIP = 1 -027 and with the above values for o and / 



So — s-i 

 h, - 88-3 . 



For mean conditions the accuracy of //^ is 3-4%, which is quite sufficient. If systematic 

 errors in the measurement are avoided (such as spray from the sea, water drips, and 

 inflow of ash, etc.), the difficulty at the present day is not in getting comparable mea- 

 surements of evaporation but in the correction of the values in order to obtain sea 

 surface values. The value required is not the evaporation height in a vessel on board 

 but the considerably different values at the surface of the sea. For this it is necessary 

 to know: (1) the factors on which in the most general case the magnitude of the 

 evaporation depends; and also (2) the differences between these factors in the vessel 

 on board the ship and at the free sea surface. For an answer to these questions it is 

 thus essential, during long-term series of observations on board ship, to perform 

 special additional measurements; for instance, of the temperature of the surface of 

 the water in the evaporation vessel, etc., in addition to the self-evident meteorological 

 observations on board. Measurements of the evaporation in this way have not been 

 made very often. They were first carried out by WiJST (1920, see also Schmidt, 1921) 

 in a fundamental investigation, and these values after critical interpretation were used 

 to derive more correct average zonal values of the evaporation at the surface of the 

 ocean. From all the formulae which have been used many times to calculate the effect 

 of the meteorological factors, the best is the expanded Dalton evaporation formula 

 in the form : 



he = cf{u){\ + at){0-9Se, - e„). 



e^ is the height of water evaporated in 12 or 24 h, c is a constant and f{u) takes into 

 account the eff'ect of wind velocity. The last two expressions in brackets, which were 

 termed the "evaporation potential p" by Marvin (1909), take into account the effect 

 of the air temperature / and of the difference between the saturation pressure of water 

 vapour at the temperature of the evaporating water eg and the water- vapour pressure 

 in the air e^. The factor 0-98 takes into account the effect of salinity which hinders 

 evaporation. As it is known if 30%o < s < 50%o at the sea surface then e^ can 

 be put equal to 0-98^5 in the atmosphere whereby the evaporation is almost indepen- 

 dent of the salinity. 



From reliable measurements on board a moving ship single values of the quotient 

 ejp can be computed and can be related with the motion of the air at the time of 

 measurement, which is identical to the actual wind measured on board the moving 

 ship. A conversion of the evaporation measured on a moving ship into that which 

 was measured at deck height with an evaporation vessel at rest and at the true wind 

 speed over the sea can be done with sufficient accuracy. 



For correction of the true evaporation obtained from the instrument on board ship 

 to that at the surface of the sea, i.e. of the free ocean, Wiist used the gradient of the 

 meteorological factors between the evaporation vessel and the sea surface. A basis for 

 estimating the gradients of air temperature, humidity and wind speed immediately 

 above the sea surface was obtained from observations in the Baltic Sea (September 



