15 March 2011

New GPS vectors

Posted by Callan Bentley

Just wanted to call your attention to two new maps showing GPS displacement vectors from Japan. (Barry left links to these images in a comment yesterday.) These images are hosted on the website of the Geospatial Information Authority of Japan, and though I can’t read the Japanese to verify their authorship, I presume that agency produced them as well. They are easier to read than the one I posted on Saturday morning, and they include additional information about the vertical component of displacement.

Horizontal motion between March 1 at 11pm and March 11 at 3pm:

The biggest one shows a maximum displacement of 4.4 m! And this, keep in mind, is on the land-based receivers on Honshu. (Of course, there would be more displacement on the seafloor closer to the epicenter, but we don’t have GPS receivers mounted there. If this is your first time looking at one of these maps, let me point out that the arrows are not drawn at the same scale as the underlying map. A scale for the base map is in the upper left, and a separate scale for the GPS-measured displacement vectors is in the lower right.

Here’s vertical displacement between March 1 at 11pm and March 11 at 3pm:

At first, I found it surprising that the land dropped in elevation, in some places by as much as 75 cm (~2.5 feet) as a result of the quake. Initially, I would have guessed it would increase in elevation instead, since northern Honshu is climbing up relative to the subducting Pacific Plate beneath it. Upon reflection, however, what I think was happening is that the crust beneath Honshu was flexing upward as stress built up on the fault (i.e., before the earthquake), then at the moment of slip, this “arched” crust of Honshu relaxed back into a wider, lower configuration. That’s also the interpretation presented by this video, which I highly recommend watching if you haven’t already.



That also would explain some of the lingering flood effects shown in these before/after satellite images of the coast of Sendai: the shore is lower than it was before the quake!

The Sendai Supersite shows a plot of displacement as measured in the vertical dimension, latitude, and longitude, for a GPS measurement station at Mizusawa, a city north of Sendai on Honshu. According to the description on that web page, this is work done by a graduate student named Simon Banville at the University of New Brunswick.  Please consult the Supersite for a full list of credits for this work, including information sources. I’ve redrawn his figure (to make it display better in this space) here:

The vertical axis of the plot shows displacement in meters (1 m = 3.28 feet). The horizontal axis shows the passage of time. For Mizusawa, the action starts at 5:48 and ~20 seconds GPS time. Overall, the plot shows the station moving 3.5 meters (11.5 feet) to the east-southeast during the first minute of shaking, and then shifting back a bit, ultimately coming to rest at a final position 2.3 meters (7.6 feet) east-southeast from its pre-earthquake location. This adjustment takes about two minutes, during which waves with amplitudes approaching a quarter of a meter (~9 inches) roll through.

So Mizusawa ends up ~10 cm lower in elevation as a result of this event. I wonder what a similar plot would look like for Sendai itself, given those -75 cm vertical displacement vectors on the coast…

I’m grateful for scientific agencies and geology departments like the Japanese Geospatial Authority and the University of New Brunswick for sharing this information with the world via the Internet immediately after a big event like this. It helps the rest of us immeasurably in coming to terms with the magnitude of the event.

What are your thoughts on these displacements?