2. Open-Coast Tsunami Recording and and Negative Influence of Infragravity Waves A.B. Rabinovich 1,2, F.E. Stephenson 2, and D.G. Goring 3, 1 P.P. Shirshov Institute of Oceanology, Moscow, Russia 2 Institute of Ocean Sciences, Sidney, B.C., Canada 2 Institute of Ocean Sciences, Sidney, B.C., Canada 3 Mulgor Consulting Ltd., Christchurch, New Zealand 3 Mulgor Consulting Ltd., Christchurch, New Zealand

3. Present scheme of tsunami monitoring on the coast of British Columbia based on tsunami warning and PWLN stations Upgrated Tsunami Warning System (Canadian Hydrographic Service)

4. Location of tide gauges on the coast of British Columbia 3 Tsunami Stations 10 PWLN Stations Sampling: 1 min

5. Tsunamis recorded on the BC coast: 1999-2005 DateRegionMStations 2001-07-23Peru 8.4 Langara, Bella Bella, Queen Charlotte, Port Hardy, Winter Harbour, Tofino, Bamfield, Victoria 2001-10-12Q. Charlotte (Canada, BC) 6.3 Port Hardy, Winter Harbour, Tofino, Bamfield 2003-01-22Colima (Mexico) 7.4 Winter Harbour, Tofino, Bamfield 2004-11-02Vancouver I. (Canada, BC) 6.7 Tofino, Bamfield 2004-12-26Sumatra (Indian Ocean) 9.3 Bella Bella, Port Hardy, Winter Harbour, Tofino, Bamfield, Victoria 2005-06-15California 7.4 Tofino, Bamfield

6. In contrast to the spectra of other stations, spectra of Langara at frequencies >0.02 cpm (periods <50 min) are “white” or “blue ”. The corresponding topographic transfer function at these frequencies is going up. BC spectra and transfer functions

7. An attempt to identify the Hokkaido Tsunami of September 25, 2003 in the BC records. This plot clearly demonstrates the problem at Langara station (very high background noise). BC tide gauge records

8. Location of the New Zealand tide gauges Sampling: 1-minute

9. Storm of August 17-21, 2003 New Zealand Three different instruments at the same site gave almost identical results. That means that high background noise is related to physical factors rather than to instrumental problems.

10. “Long waves” here refers to periods between swell and tides, that is, from about ½ minute to 12 hours… The most conspicuous thing about waves in this frequency range is their absence.Munk [1962] Generation of long background waves (From Rabinovich [1993])

12. Formation of IG-waves in the ocean From Rabinovich [1993]

13. Map of Langara Island with positions of permanent tide gauge station Langara and temporary station Henslung

14. Langara

15. Stormy Langara…

16. …and quiet Henslung

17. Henslung and Langara records Simultaneous longwave records at the Henslung and Langara stations Zoomed segments

18. Location of the Sumner Head and Lyttelton tide gauges New Zealand

19. Lyttelton and Sumner Head records Simultaneous longwave records at the Sumner Head and Lyttelton stations Zoomed segments

20. Spectra or background noise in Sumner Head during “normal” (windy) and very calm weather Spectral ratio of “windy” and “calm” longwave spectra New Zealand

21. Spectra of simultaneous records at (a) Henslung and Langara BC, and (b) Lyttelton and Sumner Head, New Zealand. British Columbia (a) British Columbia New Zealand (b)

22. Spectral ratio and coherence between (a) Langara and Henslung, British Columbia; and (b) Sumner Head and Lyttelton, New Zealand. (a) British Columbia(b) New Zealand

23. Tsunami source function 2001 Peru Tsunami (M w = 8.4) Three consecutive days Six stations British Columbia New Zealand

24. October storm (27.10.2002) IG waves IGwaves British Columbia

25. Source functions reconstructed for four New Zealand stations during a storm passage New Zealand: August storm (17-21.08.2003)

26. Main results: (1)High-frequency background noise associated with IG-waves at open coast stations (such as Langara and Sumner Head) complicate the recording and detection of tsunamis and, consequently, causes troubles in using these stations for tsunami warning. (2)Despite common opinion, tide gauges installed in bays and harbours are more appropriate for detection of small tsunamis, even though harbours are affected by resonant seiche oscillations.

27. (3)Based on present analysis, the decision was made to relocate the Langara tsunami warning station from its present location to Henslung Cove. This decision was validated when the recent 2004 Sumatra tsunami was clearly recorded at six harbour-located BC stations, but was not identified at the Langara station.