116 



OUVKR AND DORMAN 



[chap. 8 



curves for multi-layered models, are much more complex. However, certain 

 features of the complex models and curves which we shall consider can be 

 understood if the results for the simple two-media case are borne in mind. 



In order to place in the proper perspective the detailed studies of surface- 

 wave dispersion for oceanic paths, let us consider first a general summary of 

 observed surface-wave dispersion, i.e. in this case, group velocity as a function 

 of period. The data are summarized in Figs. 4 and 5 but various portions of the 



20 30 40 50 

 Period (sec) 



500 1000 



Fig. 



5. Suramary of Love-wave group -velocity observations. Regional differences within 

 continents and oceans are associated with observed departures from these average 

 curves out to periods probably as great as 100 sec. The relatively high group velocity 

 of the oceanic first mode is due to the shallow depth of high-velocity rock beneath the 

 oceans. The absence of observations of the oceanic second mode is probably due to 

 attenuation of the very short period waves which would necessarily prevail in this 

 mode. 



curves in these figures will be discussed in more detail later. Consider the 

 observations of waves of the Rayleigh type first (see Ewing and Press, 1956). 

 For periods greater than about 75 sec, i.e. for waves whose lengths are greater 

 than about 300 km, differences between continents and oceans are very small, 

 and the waves are represented by a single line in Fig. 4. The minimum value 

 of the group velocity at a period of about 225 sec results primarily from the 

 effect of the velocity increase at depths of the order of hundreds of kilometers 

 in the earth's mantle. 



For periods less than 75 sec, the curve splits and the distinction between the 

 continental and oceanic types of crust becomes readily apparent. In the period 

 range 70 to 17 sec, the ocean curve falls well above the continental curve. This 



