SECT. 2] LARGE-SCALE INTERACTIONS 159 



Qe and Qs are taken over from the oceanographic study. Ea, the atmospheric 

 radiation sink for the vokime, is calculated using the methods and figures of 

 London (1957). 



In evaluating c„, the wind data available and method of analysis impose 

 limitations on the time and space scales of the motions whose fluxes can be 

 included in the budget. Colon's method of handling this problem is typical of 

 a simplifying hypothesis that can often be used in treating the tropical atmos- 

 phere. 



For each of the two months, the meteorological network data were analyzed 

 to obtain mean values of Cn and the energy parameters at twenty evenly 

 spaced grid points around the ellipse periphery at each of ten pressure levels 

 between 1000 and 100 mb (approximately sea-level to tropopause). Thus 

 contributions of "time eddies" of a scale less than one month and "standing 

 space eddies" of a size less than one-twentieth of the ellipse circumference 

 (about 3° latitude) were ignored. The steadiness of the flow and relative rarity 

 of synoptic disturbances in winter justify this, while the contribution of still 

 smaller scales of motion in maintaining the balances are deducible later. 



The two independent precipitation estimates came out within 50% of each 

 other for both months. Their average for December, 1956, which we use in 

 later work, is about 3 cm per month. This is less than half the mean monthly 

 rainfall rate for the region found by other authors. However, December is in 

 the dry season in a climatic regime where the seasonal precipitation cycle may 

 easily have a two-to-one amplitude about the mean (Riehl, 1954). 



The energy transactions, their role and mechanism, take physical life in 

 connection with process when these heat and moisture balances are performed 

 separately for three vertically superposed layers : surface to 900 mb, 900- 

 500 mb and 500-100 mb. This subdivision corresponds roughly to sub-cloud, 

 cloud and above-inversion layers, although cloud bases and tops average more 

 nearly 960 and 750 mb, respectively. In order to complete these balances, we 

 need to draw on mass continuity requirements to deduce mean vertical mass 

 flows through the 900- and 500-mb levels. Evaluation of the mass flow through 

 the lateral boundaries of the Caribbean ellipse indicated, as expected, a net 

 convergence near 200 mb (about 40,000 ft elevation) and divergence in the 

 trade regime below. Continuity calls for sinking motion of several hundred 

 meters per day throughout the troposphere, with a maximum descent of 700- 

 800 m day~i at the 300-mb level. As we shall see, this sinking motion proves 

 essential in the heat-energy budget of the upper layers. 



The budgets by layers are summarized in Fig. 30. The Caribbean ellipse is 

 5% area- wise of the Northern Hemisphere trade region. The numbers in the 

 figure when multiplied by lO^^ cal/sec give the extrapolated fluxes for the 

 entire 10°-20° latitude belt to compare with the equatorial study to follow. 

 When multiplied by 5 x lO^^ cal/sec they give the contribution of the elliptical 

 cylinder actually computed by Colon. The h = CpT + Agz terms are segregated 

 on the left side of the diagram, while the latent heat (Lq) terms are entered on 

 the right. The solid arrows show heat and water- vapor transports by the mean 



