Five Great Tsunamis of the 20th Century as Recorded on the Coast of British Columbia

Abstract

The five great trans-Pacific tsunamis of the 20th century that occurred in 1946, 1952, 1957, 1960 and 1964 were accurately recorded by analogue tide gauges on the coast of British Columbia. All available pen-and-paper records of these events were collected, digitised, de-tided and analysed. The 1946 Aleutian Islands event was recorded at two stations, Tofino and Victoria, where maximum trough-to-crest tsunami wave heights were 55 and 27\xa0cm, respectively. These two gauged stations, as well as Prince Rupert, Alert Bay and Kitimat, also recorded the 1952 Kamchatka tsunami, which generated a maximum wave height of 77\xa0cm at Tofino. The 1957 Andreanof Islands tsunami was recorded at six primary stations and the 1960 Chile tsunami by 17 primary and temporary tide gauges. For both of these events, the maximum tsunami wave heights also occurred at Tofino: 48\xa0cm (1957) and 132\xa0cm (1960). The 1964 Alaska tsunami remains the strongest tsunami yet instrumentally recorded on the coast of British Columbia. Our examination of 16 records from this event shows that maximum wave heights at eight stations were higher than 1\xa0m, including Port Alberni (770\xa0cm), Ocean Falls (376\xa0cm), Tofino (237\xa0cm) and Alert Bay (222\xa0cm). We also find that the maximum wave at all stations for this event was among the first three waves. Subsequent wave heights rapidly attenuated following this group of waves. Frequency-time (f–t) analyses of the tsunami waveforms reveal that, for each station, the dominant frequencies of the waves and their evolution with time were very similar for different tsunamis, but differed considerably among sites for a particular tsunami, indicating the strong influence of local/regional topography on the incoming waves. From the latter point of view, the 1964 tsunami was exceptional. The epicentre of the 1964 Alaska earthquake was located much closer to the BC coast than for the other events and, therefore, the influence of the source was much stronger. The “ringing” of this tsunami was substantially shorter (<\xa01.5\xa0days), the energy decay much faster, and the f–t diagrams at all sites more similar than for the other major events. During the 1964 event, energy associated with the dominant period of 2\xa0h rapidly decayed. All of the tsunamis examined penetrated deep inside the narrow channels, fjords and inlets, typical of the BC coast. Waves also propagated far into some of the rivers. The 1952 Kamchatka tsunami was recorded at Kitimat, located at the head of Kitimat Arm about 80\xa0km from Hecate Strait, while the 1957 Andreanof Islands tsunami was measured at Bella Coola, located at the remotest part of Burke Channel, a distance of ~110\xa0km from the open ocean (Queen Charlotte Sound). The 1960 Chile tsunami was recorded at five stations located in the complex network of inlets and channels that make up the Seymour–Belize Inlet system on the central mainland coast of British Columbia, while the 1964 Alaska tsunami was observed at many stations in the Fraser River, including Pitt Lake that is connected to the Strait of Georgia through a 64-km route upstream in the Fraser and Pitt rivers. Our results further show that occasionally the fjords and inlets of the BC coast not only do not hinder the incoming waves but can strongly enhance them through resonant amplification. This was the case for the 1964 tsunami waves in Alberni Inlet, where the resonant response to the incoming waves resulted in severe damage to Port Alberni at the head of the inlet. The results of our analyses make it possible to isolate the tsunami signal and evaluate the principal parameters of tsunami waves. This, in turn, is of considerable value for three major problems: (1) Cataloguing of tsunami events; (2) estimation of the potential tsunami risk to the BC coast; and (3) verification and calibration of numerical tsunami models.