Columbia River

[harry]

(Note that this message was originally sent via email on July 12, 2011)

I have been thinking of the tsunami penetration up the Columbia River.
Although very-very small, the tides can be detected all the way up to Bonneville (Kukulka and Jay, 2003). But the wave period of the tide (12 hours) is oder-of-magnitude longer than the tsunami (1 hour). The recent East Japan tsunami penetrated the Kitakami River up to 50 km, according to Asahi Newspaper. So, it may be unlikely that our tsunami can affect the second domain that starts more than 60 km upstream (east of 123˚10'45", see Fig. 1 in http://isec.nacse.org/workshop/2011_ors ... tions.html). We have to be careful because the Columbia is not the Kitakami river. Did anybody simulate our case already?
If the foregoing is confirmed, then we should focus on the first domain (from the river mouth to 123˚10'45"). Let's discuss on this?

Harry

[stefant]

(Note that this message was originally sent via email on July 12, 2011)

Dear Harry,

I havn't done any simulations. A couple of observations might be pertinent, however:

(1) The tides in Vancouver, WA (~ km 160) are relatively large in the low-flow season. A tidal range of greater than 1 m is not uncommon (see for example http://tidesandcurrents.noaa.gov/data_m ... =Tide+Data)

(2) The tsunami from the Jan. 1964 Alaska earthquake was measured in Vancouver, and was still relatively large (compared to Astoria) in Beaver (km 87) and Kalama (km 111). See attached.

This suggests that modeling the upstream domain might be important (even if not feasible for the workshop).

thanks,

Stefan[attachment=0]ColumbiaRiver_1964Tsunami_analysis_of_5_gages_Nat_Ac_Sci_Report1968.png[/attachment]

[etolkova]

(Note that this message was originally sent via email on July 12, 2011)

Dear Harry,

So far, I've modeled the tsunami propagation up the floating river (7000
m^3/s discharge rate) and, for comparison, up the still 'river'. My
computational domain extends from the mouth (at x=333569) to just 300 m
past Beaver (x=403060), at 20 m resolution.

Please find in the attachment:

time histories (1-st column - time in hours, 2-nd column - elevation due
to tsunami in cm) as Matlab binaries, at Beaver and Skamokawa, in
floating river and in still river, file names are self-explanatory.
Please let me know if binaries won't work for you, I'll convert to text
files.

Pdf with figures shows the above time histories on the 1st page.

Apparently, flow can amplify the tsunami, so I'm looking forward to see
what tides will do, with stronger currents. I'm still tuning up a
combination of tides and flow in the river, but the soup is almost ready
for sending the tsunami in. Tides certainly penetrate far enough: ~3 m
range at the entrance, 1.5 m at Skamokawa, 0.75 m at Beaver (please see
the last figure in the attached pdf for the surface dynamics due to
tides). The tsunami came to Beaver with ten times lower amplitude than
it was at the entrance.

Best Regards,
Elena

[harry]

(Note that this message was originally sent via email on July 13, 2011)

Excellent preliminary results. Thank you.
The Columbia is definitely different from the Kitakami river in Japan. In hindsight, it's a difference between a large river and a huge river. In the Columbia River, there is very small variation in the riverbed elevation: in fact the riverbed elevation even at the upstream end (Bonneville) is still below the sea level. In some sense, the Lower Columbia can be considered as an "estuary" all the way up to ??? in terms of tsunami penetration. Any comments?

Harry