SECT. 5] 



SURGES 



615 



southernmost (San Diego), separated by a distance of about 800 km. The period 

 encompassed by the records was a typical fall-winter in that region — there 

 were no exceptionally violent storms. Records such as these in other parts of 

 the world indicate a general pattern : The over-all amplitude of the sea-level 

 variation is greater (1) in regions of stormy weather, (2) along the edge of 

 continents (rather than at islands), and (3) where the continental shelf is wide 

 and shallow. 



Fig. 3 shows a comparison between sea-level and (inverted) atmospheric 

 pressure for two of the ports. The vertical scales are such that, if the variable 

 atmospheric pressure were the only disturbing agent acting and if the sea 



San Francisco 



Son Diego 



NOV. I 

 1952 



FEB 1 

 1953 



Fig. 3. Two comparisons of sea-level and inverted atmospheric pressure. (After Groves, 

 1957. By courtesy of the American Meteorological Society.) 



surface were in hydrostatic equilibrium at each instant, the sea-level and 

 pressure plots would be identical. Actually, they are strikingly similar but not 

 identical. The difference arises either from transient effects or from other 

 agents, the latter being the more probable. The sea-level minus the atmospheric 

 pressure has been called "corrected" or "adjusted" sea-level, plots of which 

 are shown for the same stations in Fig. 4. Considerable reduction in over-all 

 amplitude has been achieved, and some coherence between adjacent stations 

 has been lost. This may be because the hydrostatic response to atmospheric 

 pressure does not depend on bottom topography, whereas the horizontal wind- 

 stress effect, remaining in the corrected sea-level plots, does. 



Now let us consider the effect of horizontal wind stress. The surface wind is 

 seldom constant throughout a whole day — the higher frequencies predominate. 

 Nevertheless there is an appreciable wind stress contribution to the day-to-day 



