one draft, sent to Tasmania
Feb. 2nd, 2012 08:00 pm![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
kareinaAs our two weeks in Glasgow draw to a close I report that working at "home" when "home" is a short term rental apartment in a city far from one's real home is rather like working at home as a PhD student. By this I mean that motivation to actually work varies drastically from day to day, ranging from totally lacking on some days to working long hours and not going to sleep till after 02:00 on others.
What have I accomplished during my time here?
My focus has been cleaning up the loose ends and finishing that paper from my PhD research. Sadly this involved needing to re-do some calculations. It is possible to work out the temperature and pressure at which zoned minerals must have first started growing by using Perple_X to calculate what minerals (and the specific compositions thereof) will be present in a rock of a specific composition at various temperatures and pressures. Once that is done for a sample one can then find the specific pressure and temperature at which the calculated composition of garnet matches that of the core of the zoned garnet in the real rock.
This is very useful if one is, as I was, trying to work out the full story of the metamorphism of a region. However, I did my calculations for my thesis back in 2007, and the program has been upgraded since then. My adviser, correctly, suggested that it would look odd to publish a paper in 2012 which uses the 2007 version of the program. Therefore I needed to re-run the calculations for the samples I wanted to feature in my paper. Sounds easy, doesn't it?
Sadly, there have been some minor changes to the solution models that resulted in several wasted days trying to figure out what I need to do different this time for one of the two samples. The other sample was even more complicated. These calculations tend to be done in "simplified model systems" using only the major elements which are present in the sample, neglecting things like boron, which are present only in trace amounts. However, for this sample one of the minerals present is tourmaline, which contains boron. Back in 2007 when first I did these calculations I happened upon a paper by someone else who had tourmaline in his samples, and he published a modified solution set to be used with Perple_X that includes boron and the mineral tourmaline. Even though I had only about 3% tourmaline in the sample I decided to try using his solution set, in part because I wasn't having any luck with the calculations I tried for that sample before seeing his models.
However, when I went to use his models this time I discovered that due to other changes in the program they no longer work as written. I tried emailing that author and asking if he happens to have an updated version to use with the newer version of the program, but sadly his research has moved on and he hasn't needed to do these sorts of calculations for long enough that he hasn't updated the models to the new required format. He did give me some information that would probably make it possible for me to do that updating, after many hours of fussing with the text files. When my erstwhile adviser heard that he suggested that instead we simply subtract the ingredients in the tourmaline from the whole-rock composition and run perplex for that.
Sounds easy, doesn't it? Especially as he did the work to set up the spreadsheet to do the subtraction. However, at my first try it didn't work at all--the program was convinced that the garnet cannot have as much iron in it as it does, at any temperature and pressure in the range of interest. This was not encouraging. Until I looked again at my notes from 2007, and remembered that the reason I had a problem with that sample back then was the many different attempts I did which said that the garnet could not possibly be as iron-rich as it was.
Luckily for me, there are many possible variables one can control in such a calculation. One of these is the presence or absence of water. The simplest assumption is that there is enough water present that the system is "saturated". When that doesn't work one tries the assumption that the amount of water available to participate in the reactions is limited. Other variables that can change is how much of the iron in the whole sample is Fe2+ and how much is Fe3+ (the technique used to determine what elements are present cannot distinguish between the two, so we have to make an educated guess).
This week, after trying a bunch of different possible starting conditions I finally determined that if I set everything exactly the same as I had for this sample back in 2007, with the exception of subtracting the tourmaline instead of considering it I wind up with calculations that say that the rock should start growing garnet of the exact composition as I have in my sample at the same temperature and pressure as I had calculated back then. This caused me much rejoicing!
So now that I have new calculations, using the revised version of the program, that give the same answer as I got years ago, I have finally done the rest of the work on the paper--it now is much shorter than the last version (down to about 3,500 words, as compared to the ~48,000 words of the thesis from which this has been distilled), says only what it should, and has conclusions explicitly spelled out. The figures have also been revised and cleaned up, and the lot have been emailed to Tasmania for comment from my erstwhile adviser. I hope he likes it, it would be nice to get it submitted for publication before the month is over.
In other news I am not amused by the UK postal service. When we arrived in Scotland two weeks ago we stopped by![[livejournal.com profile]](https://www.dreamwidth.org/img/external/lj-userinfo.gif)
clovis_t's home and picked up the stuff he has been storing for me for a year. It is mostly stuff that I am glad to have back (and in many cases have been wondering where they have been--fearing that I had lost them in the move), but there is a bit more than easily fits into the two suitcases we have as luggage allowance between us. (We brought one down, and the other suitcase was amongst the stuff we picked up.) No problem, we though, we can simply post the rest back--I haven't had it for a year, what matters if it takes another few weeks? So we boxed it up (he got a box from the company at which his class is being held), and today I carried it off to the post office. Where I found out that in this country one may only ship a package via "surface mail" if it is less than 2 kg, and my package is about 10 kg. This strikes me as odd--every where else I have been they prefer to ship the larger packages by surface, and save the air mail for the light stuff. Not here--here they want to send it by air, and want £68 to £108 for this package, depending on if we want it there next day delivery, or if we are willing to wait five days. No thank you! I am not going to spend that much to post it, not when the airlines only charges €20 for excess baggage. It will not be as convenient to have the third piece of luggage, but we can tie the box to the rolling suitcase with a leather belt and it will be fine.
Tomorrowlord_kjar should have only a half day of class (during which he will take the exam) and then we will take the train to Edinburgh, where we will stay with clovis_t and do some sightseeing on the weekend, before we fly home on Sunday. That will be nice, but we are also both really looking forward to going home.
What have I accomplished during my time here?
My focus has been cleaning up the loose ends and finishing that paper from my PhD research. Sadly this involved needing to re-do some calculations. It is possible to work out the temperature and pressure at which zoned minerals must have first started growing by using Perple_X to calculate what minerals (and the specific compositions thereof) will be present in a rock of a specific composition at various temperatures and pressures. Once that is done for a sample one can then find the specific pressure and temperature at which the calculated composition of garnet matches that of the core of the zoned garnet in the real rock.
This is very useful if one is, as I was, trying to work out the full story of the metamorphism of a region. However, I did my calculations for my thesis back in 2007, and the program has been upgraded since then. My adviser, correctly, suggested that it would look odd to publish a paper in 2012 which uses the 2007 version of the program. Therefore I needed to re-run the calculations for the samples I wanted to feature in my paper. Sounds easy, doesn't it?
Sadly, there have been some minor changes to the solution models that resulted in several wasted days trying to figure out what I need to do different this time for one of the two samples. The other sample was even more complicated. These calculations tend to be done in "simplified model systems" using only the major elements which are present in the sample, neglecting things like boron, which are present only in trace amounts. However, for this sample one of the minerals present is tourmaline, which contains boron. Back in 2007 when first I did these calculations I happened upon a paper by someone else who had tourmaline in his samples, and he published a modified solution set to be used with Perple_X that includes boron and the mineral tourmaline. Even though I had only about 3% tourmaline in the sample I decided to try using his solution set, in part because I wasn't having any luck with the calculations I tried for that sample before seeing his models.
However, when I went to use his models this time I discovered that due to other changes in the program they no longer work as written. I tried emailing that author and asking if he happens to have an updated version to use with the newer version of the program, but sadly his research has moved on and he hasn't needed to do these sorts of calculations for long enough that he hasn't updated the models to the new required format. He did give me some information that would probably make it possible for me to do that updating, after many hours of fussing with the text files. When my erstwhile adviser heard that he suggested that instead we simply subtract the ingredients in the tourmaline from the whole-rock composition and run perplex for that.
Sounds easy, doesn't it? Especially as he did the work to set up the spreadsheet to do the subtraction. However, at my first try it didn't work at all--the program was convinced that the garnet cannot have as much iron in it as it does, at any temperature and pressure in the range of interest. This was not encouraging. Until I looked again at my notes from 2007, and remembered that the reason I had a problem with that sample back then was the many different attempts I did which said that the garnet could not possibly be as iron-rich as it was.
Luckily for me, there are many possible variables one can control in such a calculation. One of these is the presence or absence of water. The simplest assumption is that there is enough water present that the system is "saturated". When that doesn't work one tries the assumption that the amount of water available to participate in the reactions is limited. Other variables that can change is how much of the iron in the whole sample is Fe2+ and how much is Fe3+ (the technique used to determine what elements are present cannot distinguish between the two, so we have to make an educated guess).
This week, after trying a bunch of different possible starting conditions I finally determined that if I set everything exactly the same as I had for this sample back in 2007, with the exception of subtracting the tourmaline instead of considering it I wind up with calculations that say that the rock should start growing garnet of the exact composition as I have in my sample at the same temperature and pressure as I had calculated back then. This caused me much rejoicing!
So now that I have new calculations, using the revised version of the program, that give the same answer as I got years ago, I have finally done the rest of the work on the paper--it now is much shorter than the last version (down to about 3,500 words, as compared to the ~48,000 words of the thesis from which this has been distilled), says only what it should, and has conclusions explicitly spelled out. The figures have also been revised and cleaned up, and the lot have been emailed to Tasmania for comment from my erstwhile adviser. I hope he likes it, it would be nice to get it submitted for publication before the month is over.
In other news I am not amused by the UK postal service. When we arrived in Scotland two weeks ago we stopped by
clovis_t's home and picked up the stuff he has been storing for me for a year. It is mostly stuff that I am glad to have back (and in many cases have been wondering where they have been--fearing that I had lost them in the move), but there is a bit more than easily fits into the two suitcases we have as luggage allowance between us. (We brought one down, and the other suitcase was amongst the stuff we picked up.) No problem, we though, we can simply post the rest back--I haven't had it for a year, what matters if it takes another few weeks? So we boxed it up (he got a box from the company at which his class is being held), and today I carried it off to the post office. Where I found out that in this country one may only ship a package via "surface mail" if it is less than 2 kg, and my package is about 10 kg. This strikes me as odd--every where else I have been they prefer to ship the larger packages by surface, and save the air mail for the light stuff. Not here--here they want to send it by air, and want £68 to £108 for this package, depending on if we want it there next day delivery, or if we are willing to wait five days. No thank you! I am not going to spend that much to post it, not when the airlines only charges €20 for excess baggage. It will not be as convenient to have the third piece of luggage, but we can tie the box to the rolling suitcase with a leather belt and it will be fine.Tomorrow
lord_kjar should have only a half day of class (during which he will take the exam) and then we will take the train to Edinburgh, where we will stay with clovis_t and do some sightseeing on the weekend, before we fly home on Sunday. That will be nice, but we are also both really looking forward to going home.