Language selection

Search

Research Document - 2012/115

Numerical simulation of tsunamis generated by submarine slope failures in Douglas Channel, British Columbia

By Thomson, R., Fine, I., Krassovski, M., Cherniawsky, J., Conway, K. and Wills, P.

Abstract

Multibeam bathymetric surveys by the Canadian Hydrographic Service and Natural Resources Canada have revealed the presence of two massive (~65 million cubic meter) submarine landslides along the southeastern side of Douglas Channel in northwestern British Columbia. Although the landslides likely date from the early to mid Holocene, Conway et al. (2012) suggest that these failures could have forced major landslide-generated tsunamis and that the risk of similar events in the channel in the future cannot be ruled out. We characterize this risk using a fully nonlinear, non-hydrostatic numerical mathematical model to simulate the tsunami waves that would have been generated by the two slides were they to occur during present-day marine conditions in southern Douglas Channel. Based on the multibeam data, the slides moved a distance of roughly 300 to 400 m before stopping near the base of the slope in water depths of around 400 m. A reconstruction of the slide regions immediately prior to failure indicates the slides were wedge-shaped. The head of the more northern slide (Slide A) began at a depth of around 60 to 100 m while that of the more southern slide (Slide B) at a depth of 75 to 120 m. Depending on the friction between the slide and the underlying seabed, the slides would have moved downslope with a peak velocity of approximately 25 m/s before coming to rest after a duration of about 30 seconds. The numerical simulations show that submarine landslides with these characteristics would generate tsunami waves with peak amplitudes of 30 to 40 m, current speeds of up to 15 m/s (roughly 30 knots), wavelengths of the order of 1 km, and periods of tens of seconds to several minutes. Highest waves and strongest currents would occur along the shoreline opposite and adjacent to the failure regions. Because of their relatively short wavelengths, the tsunami waves undergo multiple reflections and a high degree of scattering from the complex shoreline and bottom topography in Douglas Channel. These effects, combined with the flux of tsunami energy through adjoining waterways and channels, cause rapid attenuation of the waves with distance south and north of the source region. At the estimated propagation speeds of ~65 m/s, it takes roughly 10 to 15 minutes for the simulated waves to propagate the roughly 40 to 45 km to the intersection of Douglas Channel and Kitimat Arm, where peak wave amplitudes would be diminished to less than 1 m. It then takes another 15 minutes for the waves to reach sites near the proposed Enbridge facilities in Kitimat Arm where wave amplitudes would be reduced to a few tens of centimetres and associated currents to speeds less than a few tens of centimetres per second. As with the tsunami generation regions, the highest waves and strongest currents in any particular region of the coastal waterway would occur near the shoreline. Based on the numerical findings, tsunamis generated by submarine landslides of the form identified for the southern end of Douglas Channel would have heights and currents that could have major impacts on the coastline and vessel traffic at the time of the event throughout much of Douglas Channel, but a minor impact on water levels, currents and hence vessel traffic in Kitimat Arm. Hartley Bay, at the southern end of Douglas Channel, would be impacted by high waves and strong currents, whereas Kitimat, at the northern end of Kitimat Arm, would experience negligible wave effects. Additional modeling would be required to assess the characteristics of possible tsunamis originating beyond the area of the two identified slope failures.

Accessibility Notice

This document is available in PDF format. If the document is not accessible to you, please contact the Secretariat to obtain another appropriate format, such as regular print, large print, Braille or audio version.

Date modified: