175° 170° 165° 160° 155° 150° 145° 140° 135° 130° 125° 120° 



35° 



I i n 



4 



till 



l 1 



1 









NUNIVAK ^ i 

 ISLAND 







- 



*$ 



^/0- 



\ 





<?&' 





VANCOUVER 



-o 







^ 



~9 

 O 

 / 



/ 



C 







• 



OCEAN 

 i i i i i il 



1 1 1 1 \ 



Figure 3. Boundary of the finite difference grid 



arclength of 2.77 miles. The along-parallel spacing varies with latitude and 

 takes on a minimum arclength of 2.18 miles along the northern edge of the grid 

 and a maximum arclength of 3-06 miles along the southern edge of the grid. 

 The grid spacing used provided resolution of such topographic features as 

 Resurrection Bay, Port Valdez, and Sitka Sound. 



22. The model was calibrated by adjusting the friction coefficient k 

 in order to adequately reproduce the tide gage recordings made at Sitka and 

 Yakutat during the Alaskan tsunami of March 1964. The comparisons of the 

 actual tide gage recordings (with the astronomical tide subtracted out) and 

 the computed tsunami elevation time-histories for Sitka and Yakutat are shown 

 in Figures 4 and 5, respectively. The main tsunami wave, having a period of 

 aproximately 1.7 hr, is modeled quite well. The higher frequency oscillations 

 in the actual tide gage records represent the effects of the local scattering 

 of the tsunami wave. They may also represent some locally generated waves 

 caused, for example, by submarine landslides induced by the earthquake. 



12 



