92 MALKUS [chap. 4 



The more complex, extensive or difficult the geophysical problem, the more the 

 temptation to resort to pure description and inference, particularly where 

 data are scarce. Climatic fluctuations in the past, which have been deduced 

 from information ranging from lake levels, through tree rings, to human habits, 

 have largely had to be treated in this way. The danger is that two entirely 

 opposite effects may often be equally deduced from the same cause : for ex- 

 ample, reduction of the transport of the Gulf Stream may be argued to cool the 

 European climate, but it may equally soundly be argued at this level that 

 reduction of the Stream's flow would lead to a warmer Europe (see Stommel, 

 1958, pp. 175-177)! The danger of misleading the layman, who is unaware of the 

 level of inquiry, is vast and concomitant with that of discrediting sounder geo- 

 physical studies. On the other hand, when recognized for what they are, these 

 descriptions may pave the ground for more quantitative formulations and can 

 help the mariner or the pilot, for example, to interpret coherently what he sees, 

 as did the so-called "bubble model" of atmospheric cumulus, with its attractive 

 and visualizable concepts of cloud-tower "erosion" and dilution of the warm, 

 wet cores by "entrainment" (see, for example. Scorer and Ludlam, 1953). 



This chapter on air-sea interaction will draw heavily on the first three types 

 of study, occasionally joining together their results by the fourth method. We 

 shall begin in this way by describing how the whole system operates, trying to 

 indicate the level of inquiry and the present state of the knowledge upon which 

 the component parts of the description are made. For the most part, we shall 

 proceed using the following assumptions : 



(a) The solar radiational input to the planet Earth does not vary from one 

 year to the next, so that the ocean-atmosphere fluctuations are due to seasonal, 

 diurnal and spatial variations in input, and internal instabilities. 



(6) The mean condition of the air-sea system does not vary from year to 

 year, so that in its main climatological features one July, for example, is just 

 like any other in the major global regions. 



These assumptions are obviously not strictly true, particularly the second 

 which ignores internal instabilities with periods of several months or more 

 (possibilities of which are briefly suggested at the end of the chaiDter). They 

 place outside our scope the important topics of forced variations from solar 

 changes, secular storage in the ocean, and the entire fascinating labyrinth of 

 climatic fluctuation and control which are still beyond the bounds of being 

 tractable to our science. While the explorers' day in geophysics is ending, and 

 that of the model maker, the theoretician, and the controlled experimenter is 

 just dawning, that of the climatic engineer probably awaits another generation. 



2. How the Whole System Works 



The physical state of a fluid is intimately tied to how it moves or its "general 

 circulation". Planetary fluid circulations are governed by their energy sources 

 (which may in turn be regulated by their motions, as for example the cloudiness 



