2011 Japan tsunami

[jgalazm]

From @jgalazm on January 14, 2018 23:22

Earthquake model

I saw Rodrigo Pedreros at the Numwave workshop use this fault model: http://iisee.kenken.go.jp/staff/fujii/OffTohokuPacific2011/tsunami_inv.html The description they use for v8.0 is

Ver. 8.0 [Satake et al., 2013]: 55-subfault model. The near-trench subfaults are halved for the width (i.e. 25 km) from Ver. 7.0. The other fault parameters are the same with Ver. 7.0. The top depths of the subfaults were assumed to 0 km, 3.5 km, 7.0 km, 15.6 km and 26.0 km from near-trench to deep subfaults.

Wave gauges

On this paper they analize the waveforms of different tide and DART gauges.

One of the figures there presented is this one:

I think I like DART gauges 32401, 52403 and 51407, depending on the computational cost of the simulation.

Expected results:

[edit july 16th]

• Domain = whole world
• Show max heights and isochrones (using d3)
• Show plots for time series with lines pointing to their original locations

Copied from original issue: Inria-Chile/tsunami-lab#132

[jgalazm]

Satake et al (2013) use 0 depth in some patches. They explicitly say:

The top depths of the subfaults were assumed to 0 km, 3.5 km, 7.0 km, 15.6 km and 26.0 km from near-trench to deep subfaults

I'm assuming that these are the real depths and nothing should be added, but need to verify later.

Sending this to issue #136

[jgalazm]

This domain was too big, I'm constraining to [-120,-180] x [-60,60]

![Uploading image.png…]()

[jgalazm]

Satake et al report a peak slip of around 60meters at the trench.

[jgalazm]

I'm moving to the USGS finitefault model:

https://earthquake.usgs.gov/earthquakes/eventpage/official20110311054624120_30#finite-fault

See #138 to see a comparison of different frames between models.

[jgalazm]

First gauge

I'm just picking dart buoy 52406 at 5.307 S 164.977 E: http://www.ndbc.noaa.gov/station_page.php?station=52406 The time range is +- 15 days. This gives [24 feb 2011, 26 mar 2011]. The exact date of the earthquake is

2011-03-11 05:46:24.120 UTC

[jgalazm]

Many gauges

{21401: (152.583, 42.617),
 21413: (152.132, 30.533),
 21414: (178.219, 48.968),
 21418: (148.645, 38.727),
 21419: (155.717, 44.435),
 51407: (-156.545, 19.57),
 51425: (-176.32, -9.517),
 52403: (145.52, 4.02),
 52406: (164.977, -5.307),
 55012: (158.453, -15.664)}

I developed a couple functions in jupyter notebooks to download them authomatically

[jgalazm]

Here I got very nice results for buoy 52406

https://github.com/Inria-Chile/tsunami-lab/blob/3e9360c7caec718df22169dccb737376c93a4ed7/validation/04_tohoku2011/comparison/pois.ipynb

[jgalazm]

With two buoys:

[jgalazm]

With a lot more:

[jgalazm]

Need to check what happens with 51425, 51407 and 55012

[jgalazm]

Confirmed. those points are outside the domain or even very very far away (150 W , aprox). The location of the others is here:

[jgalazm]

Current cell size is 4 min, I will decrease/increase it and see its effect, and then enlarge the domain to covert a lot of the earth

[jgalazm]

The new domain:

Now it includes missing pois.

[jgalazm]

With 0.5x0.5 resolution than before in the simulation:

[jgalazm]

now I'm trying to use more Dart buoys, DARTs = [21414, 21419, 21401, 21418, 21413, 52403, 52406, 55012, 51425, 51407, 46419, 46404, 46407, 46411, 46412, 43412, 43413, 32411, 32412,32413,32401, 32402] but 32402 and 46419 don't have data for the selected daterange

The list of pois goes as follows

    pois:{
        '21414': {location:[178.219,48.968]},
        '21419': {location:[155.717,44.435]},
        '21401': {location:[152.583,42.617]},
        '21418': {location:[148.645,38.727]},
        '21413': {location:[152.132,30.533]},
        '52403': {location:[145.52,4.02]},
        '52406': {location:[164.977,-5.307]},
        '55012': {location:[158.453,-15.664]},
        '51425': {location:[183.68,-9.517]},
        '51407': {location:[203.455,19.57]},
        '46404': {location:[231.267,45.853]},
        '46407': {location:[231.168,42.682]},
        '46411': {location:[232.933,39.333]},
        '46412': {location:[239.437,32.492]},
        '43412': {location:[253.03300000000002,16.045]},
        '43413': {location:[259.91700000000003,11.012]},
        '32411': {location:[269.12,4.953]},
        '32412': {location:[273.626,-17.984]},
        '32413': {location:[266.483,-7.406]},
        '32401': {location:[286.579,-20.474]},
    }

[jgalazm]

New bathymetry and pois locations

Original tseries

Spectrum

Filtered tseries

Low and high freq components (left,right resp.)

[jgalazm]

Results

Bathymetry was 3101x1401 and simulation runs at half on each direction.

Time series

Max heights and arrivals map