286 MALKUS [chap. 4 



opposing our efforts to understand, predict or control our planetary environ- 

 ment? 



An honest answer to the last question must be in the negative : and therein 

 lies the connection between the data problem and the other partly true con- 

 clusion. The times that our studies were hampered by limitations due to data 

 alone are, in fact after careful consideration, not so very large. That limitation 

 is more often coupled to inadequate physical laws, or inadequate concepts, or 

 inadequate formulation of the inquiry in the face of geophysical complexity. 

 We are forced to accept limited "solutions" to our problems because we must 

 ignore or inadequately parameterize processes, energy sources, boundary 

 effects, nonlinear interactions and scales of motion that our present framework 

 of turbulent fluid dynamics does not tell us how to include. Even if we could 

 easily and cheaply measure anything we chose, we should often not be sure 

 what to measure. What become worthwhile "facts" to measure depends upon 

 expressible or visualized relationships they bear to one another; "facts" are 

 rarely useful in isolation. 



Re-examining the material in this chapter upon which our interaction picture 

 is based, namely turbulent boundary exchange, radiative processes, buoyant 

 convection, ocean and atmosphere dynamics and stability of flows on all 

 scales, we must conclude that further progress in our knowledge of air and sea 

 in interaction depends essentially upon growth in the fundamental physics of 

 turbulent, heat-driven fluids. Evaporation, for example, was shown to be one 

 of the most significant sea-air exchariges. To determine evaporation at point A, 

 time t, to an accuracy of n per cent becomes a definable worth-while endeavor 

 primarily when we seek its fundamental functional relation to other processes 

 and phenomena. Furthermore, improved data alone will hardly help us even to 

 determine evaporation better when the transfer equations may give results 50% 

 off due to inadequacies in their formulation of events at the turbulent boundary. 

 Nor will improved data alone aid us to understand the heat supply and release 

 by tropical circulations, until we construct some framework to relate their 

 dynamic instability to convection, transfer and planetary energetics. Further 

 data alone will scarcely help us to understand North Atlantic warming or 

 climatic change until we develop meander theories and physical models of 

 water-mass production, turbulent processes on several scales interacting with 

 one another. 



Thus, to say that the ocean and air must be treated together in their entirety 

 is even more futile than saying that all scales of motion in the atmosphere, 

 from microscopic eddy to planetary jet, must be dealt with together. Quite on 

 the contrary, in fact, we found that the more comprehensive the scope of 

 phenomena an investigator attempted to treat, the less happy we felt with his 

 resuJts, and the less we felt they contributed to our explicit understanding of 

 natural phenomena and their relationships. 



At the risk of being reactionary in this age of scientific optimism, we must 

 conclude that the head-on approach to the overall air-sea interaction problem 

 is almost surely doomed to failure. Geophysicists must discipline themselves to 



