kareina

I have been aware of the existence of "spider" diagrams to display the concentration of Rare Earth Elements (REE) within a rock sample since I was an undergrad. However, I have never really used them because A) my data from my previous research projects didn't include that sort of data, and B) I didn't really understand how to "normalize" the data in order to create the diagrams.

However, for my current research project it turns out that I do have REE data for most of my samples. I hadn't really thought about that fact yet, but today, when putting together an outline for a paper based on my research, inspired by a paper I recently read by some people who are doing a very similar project to mine, looking at geochemical data in 3D, but for a VERY different type of ore deposit. In their paper they used spider diagrams to look at their REE data and drew some conclusions as to how the REEs differ based on different patterns of alteration in their rocks. This inspired me to look at my own REE data.

Therefore I spent the day playing with it. That part about "don't know how to normalize the data" is no longer an issue, because my favorite program for plotting geochemical data knows how to do that, and does it automatically. (Have I mentioned recently how much I love this program--there is nothing like using a tool that was developed to do EXACTLY what I want to do, by people who need to do the exact same things with it?)

When doing these sorts of diagrams one organizes the data into groups, and all of the samples in one group contribute to a single line on the diagram. One possible way to group them is by rock type. Another is by concentration of one (or more) ingredient within the rock (oxides or elements). Because I have no idea which ones would be the best I decided to take a systematic approach, and, after playing with it all day, I now have a pdf of some 60 different diagrams, each one focusing on grouping the samples by the concentration of a single ingredient. I also have a spreadsheet summarizing the results--for some of them the resultant patterns are smooth regular curves, for others they are more of a saw-tooth shaped. For some of them the pattern is smooth for the groups with not much of that ingredient, but saw tooth for the group with the highest concentration of that ingredient. For others it is the exact opposite--saw toothed for the low concentrations, smooth for high.

It was a fun, and reasonably productive day at work, followed by a nice, reasonably productive evening. We bought some second hand scythe blades (because some of what is growing in the field will be easier dealt with by hand with them, than trying to use the ride on mower we got second hand from his brother), we managed to stack half of what was left of the pile of wood in the yard--one more session should see that done, which is good, because it will be nice to have it gone for the midsummer SCA event we are hosting here, and I made time to play hammer dulcimer, too. I have nearly worked out the one tune I never quite learned on the nyckleharpa--I can play most of it, but there is a couple of spots that call for a very quick sequence of four notes, and which four notes it is changes from one spot to the next, and I never quite managed to memorize that bit, and I do not yet have any other option for playing other than memorizing what I should do.

But tonight I think I managed to finally learn them on the dulcimer. hopefully I can find the time to try it again tomorrow and see if it stuck, and, perhaps, even check the nyckleharpa and see if learning the sequences on the hammer dulcimer translates to also being able to push the correct keys on the nyckleharpa.

Oh--yes, I nearly forgot--yesterday's adventure. The department had a "meeting", which, in this case, is to say, a boat trip to the local archipelago. The archipelago of the northern Swedish coastline is, our guide (the park ranger for this area) told us, unlike any other in the world. It consists of islands made of sand and rounded stones which are growing rather quickly, as this part of the world continues to rise in rebound after the glacier melted. The sand and stones (ranging from gravel to boulders in size, with more on the small end than on the large), was all deposited here by the glacier, and gradually pushed southwards as the glaciers advanced. As a result the Bothnian bay is kind of shallow, and shallowest in the area where these deposits were thickest. One might wonder why these islands exist at all, if they are built from loose sand, gravel, and rocks--wouldn't storms and the resultant waves wash them away?

Nope-the storms are from the south, which is the direction the glaciers pushed the sand in the first place, so there is more of it out there, and during the storms the waves push the sand and rocks back to the north, and each storm builds the islands up, just that little bit. Add to that the rebound, and the islands are growing noticeably. Our guide says that the island he brought us to has grown 18 centimeters taller in the 21 years since he first started going there. What does this mean in terms of how it looks? Well, the old fishing village, which was very active there from the late 1800's to the early 1900's, has a cluster of houses all facing what was once the harbour. They houses are all offset from one another, because it was forbidden to build your house somewhere where it could block another man's view of his boat from his house. Today there is a wide, grassy field with a bit of forest where that old harbour was, and there is a much newer harbour, built by manual labour by folk who would have otherwise been unemployed during the 1930's.

Yes, I did take some photos, and there remains a chance that I might yet upload some to share, if anyone actually wants me to (e.g. the links aren't good enough), but not tonight--tonight I need to do yoga and get to bed. This is a short work week, and I have lots to do...

(Alas, LJ wasn't working when I typed this, so it is being posted the next morning instead…)

I wound up staying up kind of late on Monday; I went home at a reasonable hour, but then did a nice long yoga session, followed by going out jogging in the rain (only around a largish block, but still, I got out and did something!) followed by some stitching on my underdress in progress, all of which combined meant that I didn't get to bed till after 02:00. Since sleep is one of the things I keep high on my priority list, I, of course, slept in till nearly 11:00 (have I mentioned recently how much I love being able to set my own schedule?)

However, this translated into a late start to the work day as it was necessary to first catch up on reading LJ etc. and then head to the bank. I have been running low on cash in my Alaska bank account, since that is where my student loan payments come from. At my current rate of loan repayment the money there would have run out as of the January payment. Since it can take up to a month for a transfer from my Italian bank to Alaska to actually finish processing I decided that it was time to actually do the transfer. When I arrived at the bank and went to the desks in back where complicated transactions occur the lady who offered to help me said "yes" when I asked "Parla inglese?", but the expression that crossed her face when she said it looked very pained—like she was foreseeing an unpleasant experience trying to actually communicate details on whatever transaction brought me to this part of the building. When I said (slowly and clearly) that I needed to do a transfer to the US, here is the paperwork from last time (and handed her the paperwork) the look of relief that crossed her face was amazing. She happily went and got the correct paperwork (which form, oddly, looks different now than when last I did this months ago), and the transaction was complete in record time. Now I wait for the money to actually show up there, and try not to worry about it going astray in the meanwhile.

Despite the morning distractions, I actually did settle into doing uni work by about 14:40, and was happily entering data from the geologic literature into spread sheets to be compared with my experimental data later, when my boss came in to speak with me. He had a question about one of the phases in one of the experiments, which plotted in an unexpected location on the AFM diagrams (page down) I'd sent him last week.

It turns out that the sample in question had exactly *one* chloritoid analysis that hadn't been rejected due to problems (and 15 which had been)—this sample's chloritoid grains are really small—the microprobe can be focused to a beam size of 1 micron, so any grains that are smaller than that (or thinner than that) the analysis will be partly that grain, and partially whatever is adjacent to it. For this sample the long axis of the grains is normally about 1 micron, and as a result I had problems getting even the one "good" analysis. But was it good?

He suggested plotting all of the bad data for the chloritoid for this sample to see if they fall on a trend line connecting an ideal chloritoid and one of the other mineral phases present. Since my experiments make use of a simplified composition (only the elements Ca-K-Fe-Mg-Al-Si-H2O are present in the starting powder), I was able to get away with only a page worth of diagrams plotting CaO, K2O, FeO, and MgO against SiO2 and Al2O3. As it turns out there are two distinct trends—some of these analyses fall nicely between chloritoid and kyanite, and the rest fall nicely between chloritoid and muscovite. In addition, the one "good" chloritoid plots in the group that is mixed kyanite and chloritoid, and needs to be rejected after all. That nicely solves the problem of it plotting in an odd place on the AFM diagram, but, alas, we now have no data at all for that sample's chloritoid. In the short term I can extrapolate what it might be (same as I did for two other samples which also have chloritoid too small to analyse and use those numbers in the graphs with a huge footnote that they are based on a guess, and why). However, I am now more keen than ever to re-run this particular experiment. He says we can do one more run starting later this week, and I think that this particular P-T conditions get my vote in hopes of getting better results next time (this is one of the experiments I blogged about ages ago that had poor textures due to a failure to seal them properly). I am confident that the capsules waiting to be run are properly sealed, so it is reasonable to hope for a better texture if I run them at the same conditions again.

Much to my surprise, when I had finally finished the last of the graphs and sent the e-mail to my boss about them it was midnight! 9.5 hours work did nice things to my graph of average number of hours worked per week for this month. However, after working there were still things to be accomplished with my night (like telling Quicken about the bank transfer and checking mail), so it will be another late night, given that it is pushing 03:00 and I still need to do my yoga. (Fortunately, my 1000 is done for the day). Time to head home, and hope that I can actually post this tomorrow.

Starting in May of 2008 I begun keeping a log of the number of hours a day I spent doing uni work, sleep, exercise, reading (email, fiction, blogs whatever—movies, plays and other forms of passive entertainment fall into this one), social stuff, or "useful tasks". I started this because I was genuinely curious as to how my time was being broken down, and in hopes that it would be a tool to help me focus on actually doing my uni work and getting my PhD project finished and written up. I think it helped, in that I did get it done, and I watched the graphs for each day to see how my hours of uni work were doing, trying to bump the totals up higher and higher. The achieved a record high during the final push to finish my thesis, and then I boarded a plane and commenced a short vacation travelling to visit friends and family before heading on to start my first post-doc position. Because I like records, I continued to dutifully keep records on how I was spending my time, but fell out of the habit of looking at the graphs and totals.

The first couple of weeks I was in town my days were largely spent doing "useful tasks" in the form of necessary paperwork so that I could get paid, so that my visa would become official, so that I’d have a place to live and a bank account. I learned where the supermarket was, and where the natural food store which carries the things I consider essential but aren’t available in a supermarket is located. After helping me with the portion of these tasks which required his assistance my boss gave me a stack of papers to read and went on holiday, telling me that if I want to do any travel, while he’s gone would be a good time. Great thought, but the budget hasn’t really recovered from the move. So instead the plan was to spend that time continuing to get to know this area and focus on doing that reading and working on tasks left over from my PhD project in terms of preparing papers for publications and talks/posters to present at conferences. However, while I continued to keep track of how I was spending my time, I didn’t look at the graphs and paid no attention to the totals. And it shows!

The category of "uni work" is down under 20 hours a week (ok, I confess, it is under 18 hours a week) for the months of July and August. 30 to 40 hours a week was typical for most of the record period, and it broke 63 hours a week during my final push to finish before boarding that plane. The only other time it got this low was December of last year, when my scholarship ran out and we moved house—I did very little work while packing boxes and moving, and then not much more doing family stuff with the holidays. Clearly it is time to actually watch these graphs, and see how high I can push August, given the slow start to for the first 10 days. Feel free to enter your guesses as to what the August total average hours/week winds up being now that I’ve realized that improvement is in order. I’ll come up with something nice for the person who guesses closest to correct!

There is good news in the log, however. The category of "exercise" is up to more than 13 hours a week (and just broke 14 hours a week in August), as I walk everywhere because I’d rather do that than take a bus and am often doing more than my "minimum" number of minutes of yoga in a day. Previously "exercise" ranged from a low of 10 to a high of 12.5 hours/week, back when I was doing the 2 hour+ walks with baronsnorri most days.