of generation. This indicates that the uplifted water surface at the 

 source formed a series of solitary-type waves. The multiple crest can 

 be accounted for by initial instabilities in the waveform caused by the 

 generating mechanism, and the effect of the varying bathymetry of the 

 ocean basin through which the wave passes. The lower waves following 

 the initial series of wave crests correspond to the expected oscillations 

 from a disturbance in the water surface as the disturbance is damped out. 



Wilson, Webb, and Hendrickson (1962) showed that the height of a 

 tsunami at a coastal point near the source of generation could be given 

 as a first approximation by the empirical equation 



log 1Q H = 0.75 M - 5.07 (19) 



where H is the height in meters and M the Richter magnitude. Using 

 the value of M = 8.3 given by Berg, et al . (1970) for the March 1964 

 tsunami, H = 14.29 meters (46.9 feet). However, this is an empirical 

 relationship which does not completely account for the characteristics 

 of the generating mechanism or the coastline. Although equation (19) 

 might provide a first rule-of-thumb estimate of wave heights, the actual 

 heights could be above or below that estimate. Determination of actual 

 heights would require computation by numerical or empirical means. 



Wilson (1969) gives the relationship of Housner (1969) for the fault 

 length L.p in kilometers as 



log 1(} L = 0.87 M - 4.44 (20) 



giving a fault length for the March 1964 tsunami of L-p = 604 kilometers 

 (1.98 x 10 6 feet). This is within the range of estimates given in Berg, 

 et al . (1970) and approximates the length of the generating area, i.e., 

 the length along the initial wave crest . 



Wilson and T0rum (1968) give a relationship for the period T (in 

 minutes) of the primary tsunami (carrying maximum energy) as 



log T = 0.625 M - 3.31 (21) 



For the March 1964 tsunami, this equation gives a period of T = 75.4 

 minutes (using the Richter magnitude M = 8.3). This is very close to 

 the period of the positive surge noted by Van Dorn (1964) at Wake Island, 

 and is equal to that period if the crest of the initial oscillatory wave 

 at the trailing edge of the surge is neglected. 



The initial deformation of the water surface, for any tsunami, will 

 collapse into some system of waves which must be defined. The resulting 

 wave system depends on the shape of the seabed deformation and the water 

 depth above the deformation. The simplest means of analysis is to assume 



36 



