170 MALivUS [chap. 4 



slightly smaller than the calculated mean h export of the trough of 0.94 units, 

 particularly since the Caribbean is within one of the major high-level transfer 

 channels. However, data of Palmen et al. (1958, Fig. 6) suggest that the other 

 two channels may be far stronger, and due to omission of Qvo our atmospheric 

 export may be overestimated. In any case, the discrepancy is not surprising in 

 view of the handicaps facing both determinations, nor does it alter the main 

 physical deductions drawn from either study. Taken together, the trades and 

 trough pick up a total {Qs + Qe) of 3.18 units from the sea and lose 2.32 units 

 by radiation. Neglecting the 1% or so (Kraus, 1959) converted into kinetic 

 energy of tropical winds, this leaves about 0.86 units, or 27% of the input, to 

 be exported to mid-latitudes, in good agreement with the figures in Table I. 

 As well as an inefficient combustion engine, the atmospheric machinery operates 

 a very leaky fuel pump. 



To evaluate the integral terms in (27c) and (28c), cross-sections relative to 

 the trough were prepared of Lq and h using the monthly Climatic Data for the 

 World (U.S. Weather Bureau) and other climatological information. The object 

 was to composite one vertical cross-section for each property, extending from 

 20° latitude on the summer side to 20° on the winter side of the trough. Longi- 

 tudinal variations proved small enough to do this, except that land and water 

 areas could not be combined in a single cross-section. Highlights of the results 

 for ocean areas are shown in Fig. 33 (land sections are qualitatively similar). As 

 in the evaporation calculations of Table XIII, the outstanding feature is the 

 symmetry with respect to the trough. Fig. 33a shows that, in spite of the sun's 

 wider migration, the warmest temperatures are situated on the troughline 

 from the surface to 200 mb. This warm core is a result of condensation of the 

 imported water vapor, in a manner we shall see presently, and it not only 

 migrates with the trough but is responsible for its existence and function. Its 

 dynamic consequences arise by way of the hydrostatic pressure field. While 

 warm fluid columns in total weigh less, and thus exert lower surface pressure 

 than cold ones, their expanded condition leads to a slower pressure decrease 

 with increasing altitude. At the equatorial troughline, the temperature excess 

 is strong enough actually to reverse the horizontal pressure gradient from the 

 low to high troposphere (Fig. 33b). The pressure head is thereby directed 

 equatorward at the surface and poleward aloft, where it drives the antitrade 

 westerlies of Colon. It is thus established that the mass distribution is suitable 

 to maintain the tropical cell by a simple "heat engine" type circulation. As 

 mentioned by Riehl (1954), a warm core is not found in the equatorial trough 



Fig. 33. Vertical cross-sections of properties distributed relative to equatorial trough 

 over oceans. Horizontal co-ordinates in degrees latitude relative to troughline; 

 vertical co-ordinates in lOO's mb. 



(a) Vertical cross-section of temperature departure (°C) from horizontal means 

 20°S-20°N relative to trough. 



(b) Vertical cross-section of departure of height of isobaric surfaces (ten's of feet) 

 from horizontal means 20°S-20°N relative to trough. 



(c) Vertical cross-section of departure of specific humidity (g per kg) from horizontal 

 means 20°S-20°N relative to trough. 



