Twitter | Search | |
Stephen Hicks
& what they reveal about our planet. imaging tectonic plates & ruptures . email: s.hicks@soton.ac.uk 🚴🧗🏃
7,221
Tweets
1,555
Following
7,055
Followers
Tweets
Stephen Hicks 7h
Thank you very much, Jamie, for telling us early on about the surface wave observations and for convincing me(us) that this needs to be looked into in more detail!
Reply Retweet Like
Stephen Hicks 8h
So inflation would be the opposite of collapse. But let's see how this evolves, will hope to share some figures tomorrow. Would be great to get data from Mayotte stations and add them into the mix!
Reply Retweet Like
Stephen Hicks 8h
If we interpret this mechanism in terms of magma chamber/caldera inflation / deflation (a la Bardabunga 1996/2014), then a "positive-T" CLVD mechanism would indicate magma chamber inflation. Seems a bit too deep for this type of mechanism though (depth is not so well constrained)
Reply Retweet Like
Stephen Hicks 8h
The best-fitting moment tensor wants to be highly non-double-couple (%DC< 20%) and with a high proportion of compensated linear vector dipole moment (%CLVD>80%). It has a "positive" sense - a bit like the schematic focal mechanism below:
Reply Retweet Like
Stephen Hicks 8h
I also found a good-fitting solution much deeper (~48 km), but that didn't seem sensible to me.
Reply Retweet Like
Stephen Hicks 8h
Lots of caveats on what follows as very preliminary and will need some more tests but ... the best-fitting centroid depth was 12-16 km, which places it in the lower crust (assuming a Moho depth of 17 km - from CRUST1.0 model).
Reply Retweet Like
Stephen Hicks 8h
The good news is that even with the low frequencies & relatively sparse station distribution, and with no prior constraints on location, the best-fitting centroid was able to "find" the location of the swarm. So, it seems that the low-freq event & swarm are at the same location.
Reply Retweet Like
Stephen Hicks 8h
Was going to share some figures from my preliminary CMT solution during my journey home, but turns out I left my USB drive at the desk. Doh! But I will summarise here ...
Reply Retweet Like
Stephen Hicks 10h
Ha yes I was thinking someone might pick up on that! The problem is that the stations are so far away (ave. 1000 km), I am worried that velocity heterogeneity will affect the inversion a lot at higher freqs, so I'm staying as low as possible right now. Using a basic 1-D structure
Reply Retweet Like
Stephen Hicks 10h
Possibly, at the moment I'm using a delta source, which is usually fine for *normal* earthquakes where you are below the source corner frequency. What the source corner frequency for this event is, I have no idea! Can try an extend duration later (parameterised as a triangle).
Reply Retweet Like
Stephen Hicks 10h
Depends if we're talking about Yellowstone, Campi Flegeri, etc or not ;)
Reply Retweet Like
Stephen Hicks 10h
Well my centroid moment tensor inversion using regional stations seems to be working with a better-than-expected waveform fit at low frequencies 👇 (apart from closest+furthest station) - not used. Hopefully, I get a stable solution before the day is out. Will keep y'all updated.
Reply Retweet Like
Stephen Hicks 10h
Interesting stuff - thanks, Pablo
Reply Retweet Like
Stephen Hicks 11h
Ahhh, I removed the noise spectra earlier to make the images clearer! Anyway, here it is for the closest station (since the "onset" is easier to define). Red is signal; black is noise. These are in velocity. Maybe some smaller peaks at higher frequencies, too?
Reply Retweet Like
Stephen Hicks 12h
Yes, thanks - that is where the above location table above comes from :)
Reply Retweet Like
Stephen Hicks 12h
Okay, so taking a look at these waveforms from the long-period signal a bit more ... these plots show signal power as a function of frequency. At all distances and azimuths, there is a weird, clear peak at around 16 s (0.0625 Hz) 😕😕😕
Reply Retweet Like
Stephen Hicks 12h
Replying to @VolColMac
Awesome, Colin! Very poetic.
Reply Retweet Like
Stephen Hicks 16h
do I recall that you were interested in running the Vienna (half-) marathon on the day before EGU kicks off? I was very tempted to do the half, but 80 EUR cost seems a bit pricey to me ... I'm torn ... need peer pressure ...
Reply Retweet Like
Stephen Hicks 17h
Wow I didn’t know that. In fact, whenever I travel with from London to Bristol these days, I often might as well put my watch forward by at least 11 minutes. Cc
Reply Retweet Like
Stephen Hicks 18h
More confirmation of the weird Mayotte island seismic signal. Thanks for the response, !
Reply Retweet Like