1062 
rough surface are used, the numerical values of the 
constants are: 
Ya Ta 
(Wooo < 650 em sec!) 0.030 0.085 
(Weo > 650 em sec+) 0.059 0.148 
With g = 0.623 e/p, where e is vapor pressure and p is 
atmospheric pressure, and p 1.25 X 10% and 
p = 1000 mb: 
Smooth surface 
Rough surface 
Smooth surface 
Rough surface 
issiics] 
Il Il 
i) 
79 X 107° (en — ea) Wa, (6) 
1 X 107° (€x — @a) Wa. (7) 
The hourly evaporation in grams or centimeters 
becomes: 
Smooth surface H = 2.8 X 10-® (ey — ea) Wa, (8) 
Rough surface E = 9.8 X 10° (€» — @a) Wa. (9) 
However, when climatological data are used, the 
evaporation should be computed from the equation 
El = 2:8) X09 trl Gare Wale econ (10) 
+ 3.5ml (ey — ea)Wal We>s50}, 
where n and m represent the fraction of hours with 
wind speeds less or greater than 650 em sec™!, respec- 
MARINE METEOROLOGY 
quantities H; and # could be expected to be equal only 
in the absence of lateral heat transport or if the entire 
ocean surface could be included in their computation. 
The average value of K for the four areas proves to be 
approximately 6 X 10°, that is, a value which lies 
about halfway between that applicable to a smooth 
surface and that which applies to a rough surface. Thus, 
if H is the 24-hr evaporation in millimeters or in 
decigrams per square centimeter and W, the wind speed 
in meters per second, equation (4) becomes 
J OAL} (On = G2) Wee (11) 
It would be expected that the values for H,, com- 
puted by means of the energy equations, would tend to 
be too high at the lower latitudes and too low at the 
higher latitudes because af the neglect of the advection 
term. This difference should be particularly noticeable 
in the North Atlantic where the dialatitudinal transport 
of surface waters from lower to higher latitudes is 
considerably greater than in the North Pacific. The data 
in Table III show that this is the case. Sverdrup has 
Tasie III. Vauurs or K, Hi, AND H FoR SELECTED AREAS 
Tbeaidtondle Homemade (10% a hr+) “easy (cm Aes) is (103 a hr) YB 
North Atlantic 
40°N-45°N 20°W-50°W 9.59 2.5 840 4.6 X 10° 12.51 0.77 
20°N-25°N 25°W-65°W 19.12 4.0 573 8.3 X 107° 13.58 1.41 
North Pacific 
40°N-45°N 140°W-160°E 7.73 1.9 760 5.4 X 1075 8.58 0.90 
20°N-25°N 130°W-140°E 16.92 5.3 540 5.9 X 1075 17.04 0.99 
*#H, = 
(2 = 
tively (n + m = 1), and where the notation (ey — €a)Wa 
represents the average of hourly values of the product 
(@w — @a)W.. When dealing with climatological data, 
only the average monthly or seasonal values (e» — eq) 
and W, are known, and even if the percentage of winds 
of speeds less than 650 cm sec“! (less than force Beaufort 
4) is determined, the corresponding value of (e — ¢2)Wa 
is still unknown. It remains to determine whether 
evaporation could be computed with a reasonable degree 
of accuracy, using a simple equation similar to (4), 
where for hourly values the factor K might be ex- 
pected to lie between 2.8 X 10 and 9.8 & 10-*, if H 
1s expressed in grams (or centimeters), @ in millibars, 
and W., in centimeters per second. 
The chmatological values for ¢.,, ga, Wa, and the com- 
puted values (hourly) for #; were substituted in equa- 
tion (5), and the factor K was determined for each of 
the four areas. The results are given in Table III. A 
considerable spread in the K values among the four 
areas is indicated but this is the expected result because 
it is obvious that the advection term cannot be neg- 
lected when computing evaporation by the energy 
equations for any restricted ocean area. The computed 
evaporation computed on the basis of the energy equations. 
evaporation computed by equation (4) with K = 6 X 10°°. 
further analyzed the significance of the heat advection 
term elsewhere in this volume! and he points out that 
the zonal differences in K are of the proper order of 
magnitude to be accounted for by the zonal differences in 
the heat advection term. Moreover, the computations 
of H made for the North Atlantic and North Pacific 
through use of equation (11), and averaged for the year 
for the entire area, give results which are in agreement 
with those previously reached through energy considera- 
tions. Nevertheless, there must still remain some un- 
certainty as to the average value of K. However, the 
time and space variations in evaporation over the 
oceans are so large that it is believed that the remaining 
uncertainty is relatively unimportant. 
Through the use of equation (11), which applies only 
to the particular marine climatic data which were used, 
computations of the seasonal and annual values of # 
have been made for each five-degree square in the North 
Atlantic and North Pacific. The results have been con- 
verted into their energy equivalents by taking L; to be 
585 cal g— and thus letting 
Q. = 585 E cal em day1, 
where # is given in centimeters per day. 
(12) 
