EVAPORATION FROM THE OCEANS 
applies, because it gives high evaporation rates at very 
low wind velocities and, by extrapolation, an evapora- 
tion of 0.21 mm per 24 hr in absolutely calm weather. 
When one is dealing with a pan, the evaporation may 
indeed be considerable even at very low wind velocities, 
but the evaporation from the sea surface under this 
condition must be negligible. The question as to the 
relation between the evaporation and the meteorological 
conditions at sea will be discussed later. 
EVAPORATION FROM THE OCEANS COMPUTED 
FROM ENERGY CONSIDERATIONS 
Energy available for evaporation. The first attempt 
to determine the evaporation from the oceans on the 
basis of energy considerations was made by Schmidt 
[20]. In principle the procedure is quite simple. During 
the year the oceans gain energy by short-wave radia- 
tion from the sun and the sky (Q,) and lose energy by 
effective long-wave radiation to the atmosphere (Q,), b 
evaporation (Q.), and by conduction of heat to the 
atmosphere (Q;,). Other sources of energy gains or losses, 
such as conduction of heat from the interior of the 
earth, energy changes related to chemical and bio- 
chemical processes in the sea, and friction losses, are 
negligible compared to the former. Furthermore, it 
can be assumed that the average temperature of the 
Oceans remains so nearly constant from one year to 
another that, to a close approximation, the average 
energy gain must equal the loss: 
Q: = Q; + Q. =F Qh. (4) 
Introducing 
R= Q:/ Q. (5) 
and 
E = Q./L, (6) 
where J is the latent heat of vaporization, we obtain 
ee Q: ae Q, 
K= bE @ aR): (7) 
Schmidt determined the values of the radiation sur- 
plus Q. — Q,), using measurements of radiation at sea 
combined with climatological data as to cloudiness and 
sea-surface temperature (see Fig. 1). 
Mosby [17] undertook a new computation of the 
radiation surplus, using certain empirical relations be- 
tween the incoming radiation and the altitude of the 
sun. On the whole his values agree very well with those 
of Schmidt (see Fig. 1). The major discrepancy is 
found between latitudes 20°N and 20°S where Schmidt’s 
results show a minimum of radiation surplus near the 
equator, whereas Mosby obtains no such minimum. 
The reason for this is that Schmidt has introduced a 
considerably greater cloudiness at the equator than at 
30°N or 30°S (5.9, 4.2, and 4.0, respectively), whereas 
Mosby has used nearly the same values of cloudiness in 
all latitudes between 30°N and 30°S (values between 
5.6 and 5.2). It seems probable that Schmidt’s values 
1073 
give a more correct representation of the variation of 
radiation surplus with latitude. 
McEwen [13] has computed the radiation surplus 
between latitudes 20° and 50°N in the eastern North 
Pacific, using a very elaborate method. His results are 
in very good agreement with those of Schmidt and 
Mosby, as is evident from Fig. 1. 
—— W.SCHMIDT 
60°N 40° 20° o° 20° 40° 60°S 
LATITUDE 
Fic.1.—Average annual surplus of energy (in cal cm? 
min!) which the oceans receive in different latitudes by 
radiation processes. 
The Bowen Ratio. From the preceding discussion 
and the results shown in Fig. 1 it appears that the 
radiation surplus received by the oceans has been de- 
termined with considerable accuracy. For computing 
the evaporation it is in addition necessary to know the 
ratio R = Q;/Q., which is often referred to as the 
Bowen ratio because Bowen [2] has shown that it can 
readily be computed from meteorological observations 
on board ship if, in addition, the sea surface tempera- 
ture has been recorded. 
When computing the evaporation, Schmidt [20] did 
not introduce the ratio R, but a ratio R’ = Q./(Q: — 
Q,) from which R can be found from the relation R = 
(1 — R’)/R’. Schmidt had no means of determining 
Rk’ directly, but concluded from certain general con- 
siderations that it varied greatly with latitude. Ex- 
pressed as R, the range is from a value of about 0.28 
at the equator to a value of about 1.67 at 70°N or S. 
Angstrom [1] criticised Schmidt’s determination of 
R and pointed out that in middle latitudes Schmidt’s 
values were far too large. From the application of 
energy considerations to Swedish lakes and from meas- 
urements of actual evaporation and of meteorological 
conditions Angstrom concluded that the ratio R was 
only about 0.1, meaning that, of the net energy re- 
ceived by Tadiation (Q. — Q,), only about 10 per cent 
was given off to the atmosphere by conduction and 
about 90 per cent was used for evaporation. 
Mosby [17] accepted Angstrom’ s estimate of R and 
used a value of 0.1 for the computation of the average 
evaporation from all oceans. He was not aware of 
Bowen’s paper, according to which R can be computed 
if the air temperature 3, and the vapor pressure é, have 
been observed at the same height a above the sea sur- 
face, and if the temperature at the sea surface 3, has 
been recorded, from which the vapor pressure at the 
sea surface is obtained using equation (2). If we antici- 
