344 VON ARX [CHAP. 16 



to deal with slopes as small as 10 5 must be made under light winds and at 

 times when the air-sea temperature difference is less than 1°C. Unfortunately 

 the principal regions of the ocean where air-sea temperature contrasts tend to 

 be this small are also regions where currents, and hence sea-surface slopes, also 

 tend to be minimal. In the vicinity of strong but warm currents, such as the 

 Gulf Stream, air-sea temperature differences can be quite large. At best, then, 

 one may expect optical methods to be useful in the vicinity of strong currents 

 only under unusual circumstances. But given these, and a suitably precise 

 reference vertical, it seems to lie within the range of future possibility that 

 measurements of the regional slope of the sea surface can be made by optical 

 means where the slopes are as large as 10 -5 . 



3. The Problem of Time Variations 



Were any method to succeed in providing measurements of the regional 

 slope of an isobaric surface in the sea as a function of long periods of time, and 

 thus yield the time variation of the horizontal pressure field in the ocean, it 

 would be possible to deal much more completely than it is now practicable 

 with the question of the variability of the ocean circulation around the climato- 

 logical mean. Because of the very difficulty of this problem, we have not only 

 tended to become preoccupied with those simplified forms of the equations of 

 frictionless (geostrophic) motion which neglect accelerations but have further 

 restricted their uses to the components of the horizontal pressure field associ- 

 ated with the internal distribution of density. This limitation is the result of 

 the practical assumption, made in the method of dynamic sections, that the 

 sea surface departs from level only insofar as steric anomalies permit it to do so. 

 Yet it is well known that an entirely homogeneous ocean can develop sea- 

 surface slopes, possess horizontal pressure gradients and develop accompanying 

 barotropic circulations which accommodate themselves to environmental 

 changes much more quickly than any of the baroclinic processes (Veronis and 

 Stommel, 1956). That the barotropic circulation, in steady state, can resemble 

 the observed ocean circulation very closely has already been shown in experi- 

 ments with rotating models (von Arx, 1957). These experiments have also 

 revealed the conspicuous nature of the variability + hat may be expected in an 

 ocean responding barotropically to seasonal changes in the wind field. 



Since the slope of the sea surface is related to both barotropic and baroclinic 

 influences, direct measurements of this regional property could help to describe 

 the ocean circulation not only in absolute terms and at all depths but with 

 reference to volumes of water large enough to have significance in the general 

 circulation. 



References 



Brocks, K., 1950. Die Lichtstrahlkrummung in Bodennahe. Deut. Hydrog. Z., 3, 241-248. 

 Brocks, K., 1951. Refraktionsmessungen iiber See. Z. Vermess., Hft. 3. 



