Effects of Processing Techniques 

 on the Heat Exchange Results 



Ideally, if interest lies in the total of a heat 

 exchange process for an area such as a 5° square 

 about the Hawaiian Islands, for a month, an area 

 integral should first be obtained for each day (or 

 more frequently) and then be summed for the 

 month. If interest is concerned with a process 

 at a single location, values should be obtained 

 for each day (or more frequently) at that loca- 

 tion and then be summed for the month. With 

 some exceptions, the spatial and temporal dis- 

 tribution of meteorological data has been inad- 

 equate to follow these processing procedures. 

 It has therefore been necessary in the past as 

 well as here, to estimate monthly mean values 

 of the required meteorological properties for 

 relatively large areas and then to compute the 

 exchange processes. 



This procedure involves the question of how 

 large is the difference between the mean of the 

 product and the product of mean values, since 

 all the empirical heat exchange expressions are 

 products of independently varying properties. 

 Malkus (1962) discussed this problem and cited 

 examples where the conduction of sensible heat 

 and the heat of evaporation were computed by 

 the two methods. The use of mean meteorolog- 

 ical properties resulted in an underestimate of 

 about 7 percent. 



Table 2 lists for the three locations where 

 daily observations are available--lat. 2° N., 

 long. 157° W. (Christmas Island), lat. 16° N., 

 long. 169° W. (Johnston Island), and lat. 30° N., 

 long. 140° W. (Ocean Weather Station Novem- 

 ber) — for each month, the exchange processes 

 as based on the values computed daily (subscripts 

 1), and those computed from mean monthly 

 meteorological properties (subscripts 2). In 

 general, for each of the sample locations, Q(S|) 

 is smaller than Q(S2), Q(B|) is smaller than 

 Q(B2), Q(E|) is larger than Q(E2), andQ(C|)is 

 about the same as QiCg). In consequence of 

 these differences, Q(N|) would be smaller than 

 Q(N2) due to Q(S|) and Q(E|) but larger due to 

 Q(B|) and Q(C|). At Christmas Island the dif- 

 ferences in the exchange processes are small 

 both in absolute and in relative terms. At John- 

 ston Island and Ocean Weather Station Novem- 

 ber, the differences (in the exchange processes) 

 are about the same as at Christmas Island ex- 

 cept for the heat of evaporation, which is signif- 

 icantly larger both in absolute and in relative 

 terms. The month-to-month variation in the 

 difference between Q(E,) and Q(E2) can also be 



relatively large. Values of Q(C|) - Q(C2) are 

 unimportant. The differences in each of the ex- 

 change processes due to the manner of compu- 

 tation are reflected in the 2-year mean values 

 which are also tabulated. 



It is evident from the tabulation that at John- 

 ston Island and at Ocean Weather Station No- 

 vember the relatively large values of Q(E|) - 

 Q(E2) have a pronounced effect on the differ- 

 ences between Q(N,) and Q(N2). An important 

 cause for the relatively large differences in the 

 heat of evaporation is the use of a variable drag 

 coefficient. 



Consider the portion of the evaporation equa- 

 tion which depends on the wind speed, the heat 

 of evaporation per unit difference in sea-air 

 vapor pressure. 



G(W) = 3.77 



arctan (W - 8) 



cal. cm 



1.96 

 ■2 day-' 



+ 1.6 W 



mb.' 



It is apparent from figure 2, where G(W) is 

 plotted as a function of wind speed, that the 

 evaporation factor changes most rapidly be- 

 tween 6 and 10 m. sec."' Below and above these 

 speeds the change is about linear. Wind speeds 

 in the trade wind zone generally fall within the 

 range where .the change of G(W) is nonlinear. 

 One would therefore not expect the wind factor, 

 G(W), of the mean wind to be the same as the 

 mean wind factor, G(W). For example, in Feb- 





G(w) = 3.77 [^^ecM::81+ 1.6] w,^ 



WIND SPEED (M. SEC."') 



Figure 2. — The heat of evaporation per unit dif- 

 ference in sea-air vapor pressure as a function 

 of the wind speed. 



