Accretionary Wedge #25 : Images

I’m late with last month’s accretionary wedge on images hosted at Highly Allochtonous (I’m writing this post between chairing exam boards).

As a quickie, I’m going to nick Clastic Detrius‘ idea and use my blog masthead image.

It is a seismogram from my PhD thesis. It comes from the pre-digital era where the original seismograms were recorded on light sensitive paper, a day at a time. These were then copied to microfiche and the copies sent round the world to a number of libraries. To get a record one had to travel to the library, find the set of microfiches for that day, then the station you wanted, then the fiche for the long period vertical component. You then had to load it into a microfiche reader and decode the time by eye. Each dot above the trace represents a minute, longer dashes represented hours (with some missing so you could work out what hour it was!). When you had found the right part of the record you then took a copy of the image. The system in the British Geological Survey in Edinburgh used a liquid petroleum based copier and after a day working in the microfiche library one reeked of petroleum vapour. It took about two years between the earthquake happening and the microfiches being lodged in the library. The contrast between 25 years ago when I had to travel from Cardiff to Edinburgh, look up at least two year old records by hand, copy them, bring them back to Cardiff and then digitise them by hand using software that I had written myself on an Apple IIe – and today when I can watch earthquake waves arrive in realtime from around the world sat at my computer screen – is staggering.

The long-period, vertical component seismogram in the image is from a mb 6.0 earthquake on September 21, 1981 274 km beneath the Kermadec Islands in the Pacific recorded at Adelaide, Australia. The direct P-wave arrives just before the second minute mark. A minute later a second downward kick indicates the arrival of pP, a wave that leaves the earthquake upwards as a P-wave, gets reflected back down from the surface near the epicentre and then travels back down through the mantle. The large arrival on the right-hand side of the image is sP. This leaves the source as an S-wave upwards and at the surface undergoes a phase conversion to a P-wave as well as a reflection and continues back down through the mantle as a P-wave. The interesting arrival here to me is the high frequency one between pP and sP which doesn’t appear on the standard travel-time tables. I tracked it down to being S670P, a wave that leaves the source downwards as an S-wave, hits the upper/lower mantle boundary at 670km depth and part converts to a P-wave that then travels on downwards through the mantle. A few years later I had a PhD student do her whole thesis on S670P, but it all started with this strange pulse on this seismogram.

I’m a great believer in observational seismology, actually looking at earthquake records rather than just pumping seismograms through big inversion programs. It is by looking at things closely and recognising when something strange is occurring that science advances.

One Reply to “Accretionary Wedge #25 : Images”

Comments are closed.