Sunday, July 8, 2012

Deep Earthquakes and Magnitude – Again!


The recent large earthquake in the Taranaki Bight was an excellent example of the kind of event I discussed in a previous blog – deep and widely felt. It was felt strongly in places far away from where it occurred, and demonstrates the usual confusion between felt intensity, local magnitude and more modern measures of earthquake size.

The Local (Richter) magnitude 7.0 earthquake occurred at 10:36 pm on Tuesday 4 July (New Zealand Standard Time) about  60 km south of Opunake (out to sea) at a depth of about 230 km. The main details of the earthquake and its relationship to other events are covered by a newsstory on the GeoNet website. This earthquake was very widely felt in New Zealand - from almost the top of the North Island to the bottom of the South Island. But if you look at the distribution of felt reports (see Figure 1) it was most strongly felt along and up the subducted plate from the epicentre. As I explained in a previous blog this is because the energy from the earthquake travels up the plate rather than directly to the Earth's surface. It is also interesting how many people in the Canterbury region reported feeling the earthquake quite strongly. This is again because the tectonic plate "guides" the earthquake energy down the East Coast of New Zealand - it is not unusual to have deep earthquakes under the North Island felt in Christchurch but not directly above where they occur!

Figure 1: The pattern of the more than 6000 felt reports received on www.geonet.org.nz (the new GeoNet Rapid Beta site recorded a similar number) for the deep M7.0 earthquake of 3 July 2012
This earthquake again demonstrates how this "guiding" of earthquake energy causes our New Zealand Local Magnitude (used by the current GeoNet website) to overestimate the magnitude compared to international estimates. The earthquake "feels" larger to most New Zealanders than it actually is! For example, the United States Geological Survey (USGS) recorded this as a M6.2 earthquake, the GeoNet Rapid (Beta) site estimated the magnitude at M6.5, while our own estimate for the Moment Magnitude was M6.3. Who is right? The last three numbers quoted (6.2, 6.5, 6.3) are all basically estimates of Moment Magnitude which is based on the actual source characteristics of the earthquake. In an ideal world all these would give that same value, but this range of values is reasonably normal. The Local Magnitude is known to be about 0.5 magnitude units high for deep earthquakes (because of the effect described above), so all these values are about what we would expect. If we were currently using GeoNet Rapid as the official site (we will be from early September) the earthquake would have been reported as a M6.5.

This large overestimation of magnitude does not occur for shallower earthquakes, although estimates of magnitude from different systems will show some variation. For example, a recent earthquake west of Christchurch (Friday, July 6 2012 at 3:29 pm NZST) was assigned a magnitude of M4.8 by both our current GeoNet systems and the USGS and M4.9 by GeoNet Rapid (Beta). However, another deep earthquake on Saturday, July 7 2012 at 12:50 pm NZST again shows the deep earthquake effect, with GeoNet Rapid (Beta) giving it a magnitude of M5.2 compared to M5.7 for our current system.

This event also demonstrates the speed of GeoNet Rapid. The first location appeared in 1 minute 27 seconds (although at that stage the magnitude estimate was less than M6), and the magnitude "stabilised" at M6.5 in less than 2 minutes 30 seconds. Knowing the depth and location for a large offshore earthquake that quickly is very useful - it immediately tells us that no tsunami will be generated (the earthquake is too far below the sea bed). In fact many people will have had access to the location before the shaking was over!