SECT. 2] LARUE-SOALK INTERACTIONS 161 



the ocean ; and, secondly, the bulk of the requirement for precipitation heating 

 falls in the 900-500 mb layer, to which nearly all convective clouds are confined 

 in this season. \Mthin computational accuracy, the heat deficit in the upper 

 troposphere is balanced by the flux convergence of h and, implicitly, conversion 

 of potential energy, ^4^72, into sensible heat energy, CpT, by the mean sinking 

 motion. Actually, 500 mb is considerably above the mean height of cloud tops 

 (800-700 mb) and some compressional heating by subsidence is also needed and 

 found in the middle layer, where the radiation sink is more than twice LP. 



As a whole the Caribbean ellipse region exports 6.1 x lO^^ cal/day latent heat 

 in December and imports 61% that much sensible heat ; the net export of total 

 heat energy {Q = CpT + Agz + Lq) to other parts of the globe thus amounts to 

 29% of the total transfer {Qs-\-Q,€) received locally from the ocean, in excellent 

 agreement with the Pacific work of Riehl et al. (1951). This result brings out an 

 important and paradoxical feature of the tropical atmosphere, illustrating the 

 complex linkage between planetary fluid dynamics and energy transformations. 

 Examining the whole trade -wind troposphere, we see that an outside source of 

 energy is needed to drive the circulation and offset radiational losses. This is so 

 despite the fact that a more-than-adequate source is readily at hand in the 

 form of latent heat. The latter is, however, exported rather than released and 

 processed locally. Therefore, while the region as a whole exports heat energy, 

 it is at the same time dependent upon processes elsewhere for its own main- 

 tenance I 



The final completion of heat and moisture balance for each layer requires 

 the upward-directed dotted arrows (residuals) across the 900 and 500 mb 

 surfaces. If real, these transfers must be achieved by processes on a time and 

 space scale much smaller than that of the mean circulation, namely eddy 

 turbulence or convection. Similar fluxes were obtained in the Pacific trade by 

 Riehl et al. (1951), who demonstrated quantitatively that the trade cumulus 

 chimneys (Fig. 2) were easily able to pump up the required moisture. Since 

 then, the Woods Hole expeditions have confirmed the existence of such fluxes. 

 Their individual cloud measurements and convection models (Malkus, 1958) 

 illustrate how the cumulus groups accomplish the net warming, moistening and 

 deepening of the cloud layer as the trade-wind air flows downstream. In the 

 Caribbean study, Colon used their results to show that the total upward energy 

 flux through 900 mb may be achieved by normal cumulus updrafts if the mean 

 cloudiness is 35% and 2% of the cloudy matter rises at 2 m/sec. The trade 

 cumuli serve more purposes than to decorate postcards of the tropical atolls 

 which they inadequately water! 



The link between heat source and circulation dynamics is forged when we 

 examine the pressure drop along the trade- wind trajectory. It is, similar to 

 the net warming, large at the ocean boundary and, also similar to the warming, 

 vanishes at the top of the moist layer, so that the 700-mb pressure surface does 

 not incline downstream. In fluid mechanics, the basic driving force is "pressure 

 head" or horizontal pressure gradient; on the rotating earth, its downstream 

 (ageostrophic) component is the intermediary by which the energy sources are 



