6. Tsunami Simulations
This part of our project conducts simulations of the March 11, 2011, M 9.1 Tohoku and the February 27, 2010 M 8.8 Chile tsunami events. Details on historic tsunami events are available from the NOAA/WDS Global Historical Tsunami Database at NGDC.
For our simulations, we use the respective epicenter as projection center for our map projections. In preprocessing[1] the map projections project the data, typically given in terms of a reference ellipsoid, to a planar domain. Using the epicenter as projection center reduces distortions of the used cylindrical map projection (transverse Mercator) in the vicinity of the epicenter. Additionally, we use a scale factor of 0.9996 following the UTM mapping system.
For both events, preprocessed input data in the form of binary netCDF-files are provided. As an example, Fig. 6.1 illustrates the bathymetry data for the Tohoku event.
6.1. 2010 M 8.8 Chile Event
Bathymetry and displacement data for the February 27, 2010 M 8.8 Chile earthquake and tsunami event are available as part of the provided archive.
You should be able to read them with Ch. 5.2’s setup::TsunamiEvent2d
.
Tasks
Make yourself familiar with the input data and visualize it.
Simulate the tsunami event and visualize the output. Use different grid resolutions, e.g., 1,000m, 500m or 250m, and outflow boundaries for your simulations.
How long (simulated time) do you have to run the simulations until the first waves leave your computational domain?
What are the computational demands of your simulations (number of required cells and cell updates)?
Hint: Use proper color maps for visualizing the bathymetry, e.g., the ParaView XML color maps (converted from GMT). Note, newer ParaView versions additionally require wrapping the color maps in the XML-tag
ColorMaps
.
6.2. 2011 M 9.1 Tohoku Event
Analogue to Ch. 6.1, bathymetry and displacement for the March 11, 2011, M 9.1 Tohoku earthquake and tsunami event are available in the linked archive.
Tasks
Simulate the tsunami event and visualize the output. Use different grid resolutions and outflow boundaries for your simulation.
How long (simulated time) do you have to run the simulation until the first waves leave your computational domain?
What are the computational demands of your simulations (number of required cells and cell updates)?
Sõma is a town in Japan about 55 km south and 128 km west of the March 11, 2011, M 9.1 Tohoku event’s epicenter. We are interested in the time between the earthquake rupture and the arrival of the first tsunami waves in Sõma.
Find measured data for Sõma for the March 11, 2011, M 9.1 Tohoku earthquake and tsunami event.
A cut (see Fig. 6.1) of the bathymetry is located at the linked location.
Points:0
andPoints:1
are the \(x\) and \(y\) coordinates andz
is the corresponding value for the bathymetry. The cut contains Sõma and the epicenter. Use the following rule of thumb to approximate the wave speed \(\lambda\) and the travel time until the first waves reach Sõma:\[\lambda \approx \sqrt{gh}\]Add a station close to Sõma and measure the \(h\), \(hu\) and \(hv\) over time. Compare the results with your simulated tsunami arrival times.