Snell-Pym

Last Tuesday I went to the Meet the Post Doc's semaniar at UCL there were two speakers the first one was Rosie Smith with Evidence for Seismogenic fracture of Erupting Silicic magma. This was interesting becuase it was looking at the fractures that occure like in the mouth of volcanos where the magma (hot rock) is pushing upwards. This means that the rocks are hotter than normal rocks that are being deformed by say mountian building events and so the fualts break and fracture and then some of them basically anneal shut again - the fracture heals its self or the rock sort of glues itself back together again.

Again these faults and their behaviours will affect the volcano-tectic earthquakes. It is also interesting to me from a structures point of view and the effects all this fracture and annealing will have on say the minerallogy and the structural integrity of the resulting rock.

However the bit that really intrested me was her talking about the apparatus she uses to simulate what is going on - they put the rocks in machines that squash them and heat them and there are obviously alot of constraints and alot of issues as to weather the sample has been treated to get rid of water and all the rest of it.

I like these types of machines and would love to use them - they can also test the aucostics of the sample so I assume they can basically hear the fractures occuring and then propergating through their samples. However you do not seem to be able to control all three axis of stress even though it is called a Tri-axial deformation apparatus - there is also an issue with thermal gradants through the samples if you are say heating it from just the bottom etc.... they are designing new equipement which sound fun too

Last Monday I had a lecture on Fualting and Earthquake mechanics with Gerald Roberts - I'm not going to say I enjoyed this lecture becuase to be honest I found it mostly depressing.

I also both agreed and disagreed with the Gerald. Basically I will not be convinced that the apparent data skew as you back through Italy is a real data skew and not just people not recording or being in the right place to record the earthquakes - until I have seen a distrabution map of the monastrys and seen some sort of formulisation of how complete each monestries records is. Then I'd want to know what rock the monstries where standing on and therefore what sort of ground motion they would be expecting for each magnitude which I feel could dangerously biased the data - ie if they are on more ridged rock and modern towns aren't (I dont know if this is the case this is a 'suppose') then they may not have recorded the right sort of level of damage and there would be an underestermate for devastating earthquakes.

Having said all of this I do actually think he is right in the main points that he is putting across and as for those bits I doubt - I want to see the evidence and not just take his word for it - the same goes for his earthquake intensity diagram I think I understood how it was made but I don't think it was explain in a way that is going to get through to the poeple who need to listern to him.

Now the bit that I found very depressing is that earthquakes are assumed in all predictive models to have 475 yr reccurance - earthquakes can be produced by movement on a fualt now the strain builds up and up until the fualt shifts. But they are assuming that this happens at a steady rate but if you just think about snapping sticks in half this doesn't seem logical. They will break with a sort of concentrated judder. If this is happening with earthquakes then it means that you would expect a cluster or group of earthquakes to happen along a fualt one after another and then have a period of no activity whilst the strain builds up again.

Now Gerald has been measured slip rate which is I would of thought one of the first things they would have looked at with earthquake prediction - I remember doing the maths behind working the slip rates during my undergraduate so why has it been ignored? Plastic deformation ie mountian building, folding etc... is generally a precursor to the breaking of the rocks that is the fualting and thus the earthquake inducement and you measure this in strain rates which shows that for the middle of the 'crumple' zone or mountian ridge like the apanines in Italy the rate of deformation (how fast its folding up) is faster than that at the edges. So surely this means that the middle will have more earthquakes, larger and closer together.

But this is not mirrored exactly in the earthquake hazard maps produced currently as they assume this 475 yr reoccurance. This is for everywhere on the planet but they will be subject to different forces - why on earth did they get this sort of age from anyway?

Now his work shows that some fualts have longer re-occurence ages than the writen records we have - now Italy had the Romans so they have 3000yrs of records and they will still miss things - Gerald pointed this out himself. If you preject what he is saying to his risk maps then what you have is away of telling if a fault has shifted all its going to for this lot of earthquakes - this is where this become counter intuitive to poeple who haven't been looking at it in detial - areas that haven't had earthquakes for about 400 yrs are safe for thousands or at least hundreds of years rather than being about to blow (which the used maps suggest) and faults that shifted within the last century is going to go again anytime soon.

Now he showed us fualts that are behind where they should be - great you can warn those areas but then showed on the same map that there are fualts that are ahead for there earthquake quota so should be fine for ages.

This last point concerned me - if this is to be a useful tool it needs not to have faults that are ahead of themselves in the number of earthquakes they have had - the number you;ve predicted they should have had in that period of time - though over all the earthquake is expected as while the zone is active there will be strain and then slip but if there where unexpected earthquakes there then there will be when you actually project the technique into the future.

So this means you can say this area is going to have an earthquake soon but no where can actually be ruled out - now you could argue that they live in a tectonically active country they need to expect earthquakes and that is great but where should they prioritize their money for building reinforncement and the such like. Gerald thinks they should do all of it all at once but from a government perspective I doubt this will look feasible and would infact look unreasonable and there is no point in ramming facts about affects on economy of such big disastors as they are going to be getting that sort of statement from everyone about everything.

Yes I feel that as many of the buildings should be a safe as possible starting with the obvious infastructure buildings - anything new should be built properlly to begin with. Again this is heavily affected by the ground the buildings are on.

This was the bit that always makes me so depressed - he said that in Italy about 3 yrs ago there was a small mag 5 earthquake that hit a town and the only building that collapsed was the school, during the day killing about 50 kids - this sort of thing hurts me in physical scense. I don't know if it is becuase I was brought up with the nightmare shadow of the result of Aberfan where a slag heap berried a school in Wales - this happened to be where half my family where from (they were from Merthyr Tydfil to be precise though my nan was from somewhere even smaller). It was talked about with heart broken whispers when I was a child. My second cousin was involved with the rescue attempt.

I always get the guilty feeling that all this disastor management stuff or climate change stuff should be what I use my geology for but I would not be good at it - I would be doing it out of obligation and not passion. I feel I would be better off and more lickely to find something that will actually help people by researching something I am interested in and just talking to other geologist and scientists about their fields. The earthquake stuff is still depressing me though.

Last Wednesday I had an interesting lecture on the Linear dunes of the Namibia from Charlie Bristow he specialises in modern sedimentary environments. I have always liked sand dunes I really enjoyed all the Applied Sedimentology course I did at Imperial except the petroleum bits and even quiet alot of that I found interesting.

The premise people had been working on was that these huge dunes where left over from the last glacial maxium but this turns out not to be so. There is cool dating techniques you can use on sand that has been sitting in the dark - you get like an exposure date for the last time it saw sunlight as the sunlight 'bleaches' the sand grains and sets the exposure age to zero.

I remembered from before that there where three types of motion on the sand dunes - there are little sort of parasitic dunes moving across the huge great big stonking dunes for a start. It is/was thought that these types of dunes do not appear in the geological record so they where of interests becuase of that but in cutting into the dunes it is now thought that they might look simialar to other types of dunes - dune morphology and how they look inside due to migration (the wind blowing them sideways or back to front or whatever) is actually a bit of a geometric nightmare to get your head around especially when you start looking at how a paloe (old turned into rock) dune bed can be exposed - it could be tilted, cut at a strange angle and all sorts.

Anyway the dunes are younger and would appear to be able to tell us about climate changes that may have occured in the area - this includes looking for 'fossilise' hirax poo but these are rare. There are also fulgurites which are the formation you get when lightning hits sandy substrates which supports the idea of climate change which would have produced vegitation ie it got wetter - it started to rain and plants grew their root sort of stoped the dune sand from being blown about.

I suggested they look for root material which is apparently on the cards.

Mobile Earthquake Recording in Marine Areas by Independent Divers

This was last week it was the second one of the Departmental Research semiars which UCL's Earth Sciences department are having. This turned out to be about the project I had heard about when doing the MRes before becuase back in those days Frederik Simons was at UCL and thus took some of the Birkbeck classes - he's now at Princeton. I remembered his stuff from before so well as it involved fourier transforms which at the time Alaric was having issues in using himself. It was intersting to see how the project had evolved and to also to here yet agian the run down on the gaps in our knowledge of the earths structure due to the uneven distrabution of earthquake detectors - ie on land!

The interesting things I picked out of this talk was that the presance of mantle plumes below certian volcanics - ie the stuff that occures in the middle of the oceans away from either subduction zones or mid ocean rifts. I hadn't realised this and thought it was all sorted out!

In fact it turns out we are still asking questions like:

What is the convection style of the mantle

What is the nature of mantle plumes and what is the heat flow like

Then another interesting thing is are the siesmic data consistant with mineral physics.

Again Tomography came into it and he showed some really groovey pictures of like the who earth though there are lots of gaps still!

It also turns out that not only do you have to deal with the fact that the earthquakes are not evenly distributed but that they are often of the wronge size - now it's obvious that the majority of quakes will be too small to get any useful data out but what I hadn't appreciated was that they can also be too big! They fail as they are too big to actually be used as a point source!

He ran through things like Argo which have floating devices but these are tethered and there fore pick up alot of noise and stuff. If they get the funding for their floatation devices it seems that we would get much better data.

I remembered that there was a low velociety channel in the oceans but I hadn't known that the whales use it as obviosly the sound waves travel further. This can cuase problems as there are sound waves that can initially look like the p-waves from earthquakes - you seem to need to put the data through a few processes and then they look different this is important as it turns out you can use the initial 2 seconds of p-wave to predict the size/magnitude of the earthquake - p-waves are the first ones to arrive but they are what are known as body waves and they dont do any damage it is the surface waves that lag behind them that destroy buildings and if there is going to be a Tidal wave then there is even more warning for that.

This is still a highly contested idea but it would be cool if it worked - Japan seems to be the most lickly place to find out!

I went to the drinks afterwards but didn't get a chance to actually ask my questions about what type of batteries they are using on the devices as they look like one of the main issues - I also wanted to ask about how the positioning system worked but never mind.

Last Monday I had another really cool lecture with Andy Carter this time it was looking at the interconnections between things like the East Asian Monsoon, local warm pools of water in the sea, continental uplift in this case the Himalayas.

This for me was a fun mind teasing, problem solving lecture in which I guessed why there where extention faults at pretty much right angles to compressional features. I actually cheated with this and thought about where the sigma 1,2 and 3 where - this was structural stuff from back in the day and those Cosgrove lectures came flooding back - this was bizar as I swear I hadn't understood them at the time and suddenely I am using them to work out problems. I still avoided actually doing maths though.

I then ended up puzzling over the best way to do simulations and the like and again once more I feel this desire to make better simulations better models but they say that it takes months of processesing time. I thinking there must be a better way but want to know more about what makes models models and simulations simulations and what is good and what is bad. I want to use triangles for some reason and have them as colums which I don't see in all the stuff so far.

I'm thinking the gaming and graphics industry would know stuff that would be useful to this - this is just a strange urg I have :/

We disguessed ways of actually constraining stuff with dates from the field and this was fun too and it was pulling together many things from previous lectures and my undergraduate. This was cool becuase we looked at how upper crust interaction ie mountain building events are affected by the underlaying interaction of the crust and upper mantle - he actually mentioned the word asthenosphere which I haven't heard since like A'level geology I swear!

Then there is the interactions with oceanic currents and where the land is and weather curculation with the mountains in the way and that this cuases rain to actually fall and things and tempurature which can be affected by errosion rates (due to the chemical reactions involved affecting the amount of CO2 in the atmosphere) and the albedo affect which is like how much sunlight gets reflected back into space form say bare soil verses vegitation so if more rain get plants growing but then there is also rain clouds and how they act both as a green house gass warming the planet and how they are affecting the albedo affect etc.... I love this sort of stuff I really do.

Though Andy (I think) finds the structural, traditional, geology bit more interesting I found the whole concept of the actual interaction interesting - he said this was studying earth systems and I think I have found a name for some of the Jake of All Trades stuff within geology.

It was buetiful especially when you start looking at how say the evolution of certain things say lignin in plants can affect the climate and things like that - everything is interconnected and understanding how and why is something I can really get my teeth into - I am building up shortly on how I think geology as a subject will be tuaght in future but it still needs work!