236 



[chap. 4 



specific humidity, and lifting surface air with these properties, produces a moist 

 adiabatic ascent curve for which the corresponding hydrostatic surface pressure 

 is just shghtly lower than the given jps. This may be, a posteriori, obvious, but 

 only in the framework of the model. The storm must thus be able to meet two 

 simultaneous constraints : accelerated heat pick-up and sufficiently con- 

 centrated release to maintain hydrostatically the surface pressure gradient 

 necessary to ensure this pick-up. There is thus both a boundary and an internal 

 constraint. The rarity of the phenomenon suggests that one or both of these is 



?4 26 



0(°A) 



Fig. 69. Diagram showing sub-cloud air properties as a function of surface pressure (ordi- 

 nate) in horizontal inflow into hvirricane as deduced from observations in hurricane 

 Daisy, 1958. Solid line shows potential temperature {6 in °A, lower abscissa). It 

 corresponds to isothermal expansion of mean tropical air at 26'^C. Dashed curve 

 shows specific humidity {q in g/kg, upper abscissa) (see text and Table XVIII, 

 page 244). 



difficult to satisfy. To pursue this point, let us examine the air-sea relationship 

 still more explicitly. 



If, and only if, the inflowing air can increase its heat content by 2-3 cal/g in 

 the core regions can the pressure gradients for a moderate hurricane be sus- 

 tained. The question is whether and how exchange can be augmented to this 

 degree. Can the necessary magnitudes be deduced from the ordinary form of 

 the transfer formulas with coefficients unchanged from normal, or must the 

 processes of exchange be radically altered? Is the ratio of sensible to latent 

 heat input of 34% deduced by aj^plying equation (51) to Fig. 69 necessary or 



