Dulcie Head

August 24th, 2012

August 20 to August 24
Last weekend I went with Marija to Questacon, the science center here. It was so much fun! We stayed until they kicked us out because it was closing. We had lots of fun with optical illusions and puzzles and simple machines and even a giant slide that had a small vertical section that allowed us to be in free fall for a very short time. Here’s a picture of me in a kaleidoscope!

This weekend I’ll sleep in, do my laundry, and finish packing on Saturday, and then leave for home Sunday morning! I’m so excited to finally be heading home! But I’m also really going to miss everyone here. We had a goodbye dinner together with other people from the Earth Physics group for Gerry (who leaves on Tuesday) and me today, which was really nice. My adviser has been really fantastic and especially great at organizing events like the dinner which are fun and good for trying new things and meeting new people.
I finished my work this week, we chose a checkerboard resolution and I plotted our tomography results and Yoshizawa’s tomography maps on the same color scale for comparison. It was good to see that we see consistently lower velocities for my values than Yoshizawa found both for the tomography results and the dispersion results. It would have been bad if the dispersion had all been lower velocity and the tomography had all been higher velocity. I can’t say anything too quantitative about the velocities I found but I can say that they are consistently lower velocities than the longer path solutions found by Yoshizawa and Kennett. It’s possible that this is an effect of using paths only within the continent, or that it is an effect of the receivers, or the particular way in which Yoshizawa and Kennett processed their data. It is exciting that I got a consistent result though!
I have started working on my poster and I have most of the figures and text already I think but there will be many visual things to play with and lots of formatting and editing to do before AGU. I will also present a poster at my school summer research poster conference in early September so I need to finish up the working draft by then and then I’ll update it before December.
This internship has really been a phenomenal experience for me, I’ve learned so much! And I actually managed to finish a project in this short time which was a major accomplishment. I think the thing I am most proud of is how much I learned about working in a Linux/Unix environment, coding and GMT. I have really enjoyed learning these very valuable skills and I’m sure I will use them a lot in the future.
Thanks for a great summer IRIS and ANU! 

August 21st, 2012

The 11th and final version of the goals – I need to get these done by Friday!!
• Use surface wave group velocity dispersion for tomography to investigate the subsurface in Australia
o Choose a resolution for the checkerboard test, I’ve checked a whole bunch but I haven’t really found anything I’m satisfied with yet.
• Be able to produce a poster which describes my project that is up to professional standards for AGU
o Gather all figures and write text for poster
o Make a preliminary poster on illustrator!
I think this about sums up what I have left to do! I’m also working on plotting Yoshizawa’s tomography maps on the same size and colorscale for comparison. Once I have a checkerboard resolution I’m satisfied with then we can compare those results and see if the comparison between the tomography maps and the dispersion curves yield a similar result.
I finished gathering, processing and plotting data from an event that Yoshizawa used to create his dispersion curves and found that my resulting velocities were almost always within the standard deviation of his velocity values, which is an excellent indication that the deviations we see in the rest of my results may actually be due to the difference between using long and short wave paths.
Here is one of the figures used for comparison from the event Yoshizawa used.

And here is one of my checkerboard tests.

The top figure is my tomography results. The bottom left is the input checkerboard model and the bottom right is the resulting inversion from the checkerboard model, which you can see is an okay approximation of the checkerboard but not excellent.
Tomorrow morning we will decide on an appropriate resolution and then compare that to Yoshizawa’s tomography maps. I’m currently having a bit of trouble plotting them but I think I just have the latitude and longitude switched. I will also have a lot of cleaning up to do for the tomography plots, right now they’re just a very rough version to help us choose the right resolution.
Anyways, cheers to a great internship and I really feel I’m quite close to finishing up, I hope I do actually finish this project in the next three days! I’ve really enjoyed my time here and my adviser has been really great to work with! This summer (winter) has been really great and I’ve learned a lot, especially about things I knew nothing about such as coding and keeping a blog! I’ll make a final post on Friday about the results, bye for now! 

August 17th, 2012

August 13 to August 17
This week I’ve spent most of my time modifying my dispersion curve figures and fighting with this tomography code. It’s actually been going well and it’s been a very productive week, I finally managed to figure out how to set up the inputs properly. It was a complicated problem because the code only wants to consider arrival times for each separate period separately, and it needs to have all the same stations for each event in the same order. This would have been easy to accomplish if I had arrival times for every station from every event for every period but instead I had to find a way to make files that had the proper order of the stations and events and indicated properly the places in which I didn’t have data. Eventually I figured this out though and since I got that working I’ve been trying to run checkerboard tests.
The idea behind the checkerboard test is that you make a model of a checkerboard of high and low velocity and calculate synthetic arrival times for your stations and events from that model. Then you feed those arrival times into the tomography code and see how good a checkerboard you get back out. When I started I was doing it incorrectly because I wasn’t actually creating the synthetic arrival times, I was just making a checkerboard pattern. Then I ran into several other errors but eventually worked around them and now it seems I am very close to being able to see the checkerboard pre and post inversion. I am not expecting to get very good inversions for most periods but hopefully some of them will have enough arrivals to get good coverage. The ones with better coverage we will be able to compare to Yoshizawa’s velocity structure maps. I’m excited about this step, but a little worried that to get any resolution from my checkerboard tests I will need to use very large squares which are low resolution and then comparing my maps to Yoshizawa’s will be uninformative.
As of right now I seem to be running the checkerboard tests correctly but I’m not really getting any useful results yet, so I just have to keep trying different grid resolutions until I can get a good result. I’ll wait to post a picture of one until it actually looks like a real result.
I wish I had more data, and it would probably be possible to get more data but it would take time that I just don’t have. Going into the last week here I’m still feeling optimistic about getting at least one good map to compare to Yoshizawa’s data and I feel that my dispersion curve figures are good and ready for analysis. To do that analysis I am going to attempt to process data from one of the earthquakes that Yoshizawa used to create his dispersion information. If I can get an arrival from one of his earthquakes to a station that I used and process that in the same way I processed my other data then I can compare the two dispersion curves and see if I’m measuring his values properly and what the effects of including the offshore values are.
This weekend I plan on going to Questacon, the science center and back to the Old Bus Depot markets which I liked a lot, this time I’ll make sure to bring my camera! 

August 14th, 2012

Goals 10! Already! Weird, right? I go home in just 12 days. I need to add packing to the goals soon.
• Use surface wave group velocity dispersion for tomography to investigate the subsurface in Australia
o I have the code working now, which is a major accomplishment because it means that all my inputs are functional. Most of the tomography code is written in FORTRAN so getting the inputs exactly right is important.
o I am currently able to plot the outputs but I have realized that I will not be able to get a good inversion unless I include data from all the events together, instead of separately. That makes sense and I knew this would happen eventually but I’ve been avoiding it because it’s fairly complicated to make the inputs for more than one event. I had everything working for only one event. So getting it working for multiple events is the next step.
• UNIX
o Learn about Illustrator, start on poster soon!
o Add comments!
• Be able to produce a poster which describes my project that is up to professional standards for AGU
• Explore Canberra
o Go to the National Zoo and Aquarium or the Cockington Green Gardens
o Go to Questacon – the science center
I have also (once again) updated my figures to make analysis easier. I changed the way I plot the seismograms, before I was normalizing the raw data, plotting it, then bandpassing the normalized trace and plotting that, now I normalize the raw data and the bandpassed data separately which works much better for the scales on both. I also mentioned that I added the Rayleigh wave group velocity dispersion data from Yoshizawa last time. Now I’ve changed the region of the plot so it only shows the periods and velocities of interest, 20s to 150s and 2km/s to 5 km/s. Unfortunately, this means that many of the group velocity values I obtain aren’t plotted. It makes me sad to feel I am wasting them but they are probably relatively unreliable so it’s better not to plot or analyze them. I just added error bars to the Yoshizawa dispersion curves (The group velocity at each point is an average value and the error bar is the standard deviation). So now my figures are really awesome! I like them a lot. Hopefully we’ll actually get some good interpretation out of them, it looks like my values are almost consistently below Yoshizawa’s and the PREM predictions which is interesting and possibly indicates that offshore structures contribute to an artificially higher group velocity.
A beautiful new figure:

I’m actually re-processing these now to make the tick marks on top of the error bars a little smaller.
We did some good exploring of Canberra on the weekend, we went to the Mint, the Parliament House and the War Memorial. I also volunteered at CSIRO which was fun, so I think the exploring Canberra goal is going well.
At the Mint:

At the Parliament House:

I’m feeling pretty good going into my last 2 weeks here! It’s a short time but it feels like it will be long enough to accomplish something, we’ll find out soon! 

August 10th, 2012

August 6 to August 10
I submitted my abstract and re-processed all my data and made new figures this week. It’s been very busy and productive! We decided not to include any results or conclusions in the abstract so we could focus on getting it in on time, which also gave me more time to try to figure out the tomography code.
My latest figures include the dispersion from my data, the PREM predicted dispersion and the dispersion calculated by Yoshizawa. We have files from Yoshizawa that give Rayleigh wave group velocity at grid points. To get group velocity values to plot against mine I took the path between the event and the station, calculated all the points at 2 degree spacing along the path, took the integer value of those coordinates, extracted the velocity for that location from Yoshizawa’s data files and then averaged all the values along the path. This gave dispersion curves similar to PREM but demonstrate clear azimuthal variation that we can also see in our results. In general it looks like the results I get are more similar to Yoshizawa’s model than to PREM but also drastically different in some locations. One disadvantage to comparing to Yoshizawa’s data is that he does not have group velocity values for below 20 seconds period. It seems that the results I get below 10s are scattered enough that they may not hold any significance though, and I may change the plots to only show data between 20s and 120s.
Here’s one figure that shows my results fitting more nicely with Yoshizawa’s than the PREM.

And here is a different station from the same event showing a somewhat different trend.

I’m looking forward to working on my poster once I (finally) get the tomography code working. I am making progress on that, it’s just slow going. So far I seem to have most of the inputs set up correctly, I’m just not sure what the outputs are or how to plot them. Once I get a better idea of what the outputs are I can more effectively change the inputs to suit my data. Effectively I don’t have much left to do, just get the tomography code working (I already have all the data necessary ready for this), plot my tomography results on the same plot and hopefully at the same resolution as Yoshizawa’s results (which I already have the data in the appropriate format for, I just need to plot) and then analyze the comparisons of my results to the PREM and Yoshizawa’s results. Then poster time! The question is, will I actually be able to get the tomography code running? So far I’ve only been working on it while waiting for other codes to run and now I’ll be able to devote my full energy to it. Hopefully I’ll figure it out soon.
Last weekend I went with Archana, another intern, to the Old Bus Depot Markets and we ate some excellent pastries and had fun wandering around and looking at everything. I forgot my camera unfortunately so I don’t have any pictures from that trip. I’ll be sure to bring my camera for where we explore tomorrow, which is tentatively the Mint, the Parliament House, and Questacon – the science center. It will be a very packed day and I’m not sure we’ll make it to all three as they will be closing at 5pm. I’m also going to volunteer at a science demonstration for kids at CSIRO (the Commonwealth Scientific and Industrial Research Organisation) on Sunday so that will be fun!
I’m enjoying my work and I’m feeling fairly positive about being able to finish the project and if not finish the poster at least get all my figures ready before I go. I still hope to comment my codes to make them understandable to other people and to myself when I’ve forgotten everything in three months but that will have to wait until after I’m done with everything else. I’m getting really excited about going home and also to go back to school! It’s been too long now, I’m grateful I took such a long internship because it has allowed me to actually get a lot of work done but I’m missing my family and friends. At the same time I’m also sad to leave my new friends here, but I’m really starting to itch to get back.  

August 7th, 2012

This is the ninth edition of the goals (so many weeks!) I’ve noticed recently that the goals list is getting shorter every week instead of longer. I hope this is a good sign about what I’ve accomplished, not just a sign that I’m being more concise.
• Use surface wave group velocity dispersion for tomography to investigate the subsurface in Australia
o Make a crude map of the velocity structure in Australia
o Compare that map to Yoshizawa’s velocity structure map created using longer paths
• UNIX
o Learn about Illustrator, use it to create poster
o Add comments!
• Be able to produce a poster which describes my project that is up to professional standards for AGU
o Register for AGU, sort out hotels and stuff as well
• Maintain this blog about what I have accomplished and my goals
• Explore Canberra
o Go to the National Zoo and Aquarium
o Go to the Cockington Green Gardens
Unfortunately I don’t have any new figures for this post. I just finished re-processing my data because we found that cutting the traces to try to isolate the surface waves and then selecting more periods to try to get dispersion values for gave better results. I am much more efficient at processing it now! Hopefully this will be the last time I process it and then we will just look at the results for analysis and conclusions. Right now I do have comparisons to the PREM but not to the results from Yoshizawa, so finding and plotting those dispersion results on my figures is the next step for analysis. I now have all the results I need to set up the tomography, and once I can perform the tomography inversion on my data then I can compare it to Yoshizawa’s results in that form as well. So after I add the Yoshizawa dispersion results to my figures (if that is possible) then I will perform the checkerboard test to see the expected resolution for the tomography using my paths and then actually do the tomography. I’m excited about my progress!
In other news, I went to the Old Bus Depot Markets last weekend which I really enjoyed! I ate a nice cinnamon pear danish and bought some vegetables. I’m hoping to go to either the Zoo and Aquarium or the Cockington Green gardens this weekend. Exploring Canberra is fun and there is no shortage of places to visit here.
Till next time! 

August 3rd, 2012

July 30 to August 3
It’s getting closer to the end here and very close to the abstract submission deadline! I’m currently on my second draft of the abstract, and I’m going to process my data and try to get that tomography code working before the final draft and submission. I just joined AGU yesterday so I’ll have to wait a couple more days before registering for the meeting. I’m quite excited to go! Hopefully in my last week here I’ll spend some time learning Illustrator so I can make, or at least start on, my poster. If I can’t finish the poster here I’ll need to think about how to transfer the necessary figures to my laptop before going. I’m going to miss everyone here and I’ve made some good friends but I’m also very excited to be heading home and then back to school.
I plan to submit my abstract to the Seismology General Contributions session and request a poster so I’ll need to make sure everything for my AGU registration is working on Monday. For now I’m trying to see if there is a way I can cut or filter the data differently in order to extract more results.
Today was a very unproductive day, which is always disappointing but I have high hopes for next week. There were many interruptions today because they were rearranging the room where I work, there was a going away party at tea in the morning for the director of RSES, I had a meeting, and there was an hour of lecture and an hour of tutorial for my adviser’s Physics of the Earth class. Also, my Ethernet cable went missing somewhere in the process of them moving everybody’s desks around so I need to get that back. Thrilling stuff, I know. Even though today was fairly underproductive, I think this week was actually fairly successful overall and I now know which data I can use and how to use it and all the processing and plotting codes are working. I’m just going to try to improve the results next week and get the tomography code working. If I can do that by Tuesday then I think I’ll be in really good shape for submitting the abstract - its fine now but I’d really like to be able to add some real results instead of just aims and hopes.
On the home front I made banana bread twice. And forgot to take a picture both times, sorry! Here’s a cool one of one of the animatronics at the dinosaur museum instead, though!

No particular plans for this weekend yet but hopefully next week’s blog will have some exciting pictures of wherever we go! And also some good (and maybe finalized) figures! I’m looking forward to my last three weeks here; I know it’s going to go by quickly. Perhaps too quickly for all the work I want to get done, but in some ways not quickly enough because I’m really feeling ready to go home. I think that’s always the trouble with time though. It always goes too fast and not fast enough, all at once. I’m sure
Cheers! 

July 31st, 2012

While I haven’t done an actual analysis yet, it is now possible! I’m very excited about my new figures. Here is the eighth update of the goal list, crazy how time flies!
• Use surface wave group velocity dispersion for tomography to investigate the subsurface in Australia
o Use the good data for tomography to make a crude map of the velocity structure in Australia
 Learn the tomography code!
 I’ve started work on this and I think I’m making good progress so far, I’m working with examples and following through the instructions.
• UNIX
o Learn about Illustrator
o Add comments!
• Understand what group velocity dispersion is used for and be able to communicate knowledgably about it
o Read papers and thesis from someone at ANU concerning group velocity dispersion.
• Be able to produce a poster which describes my project that is up to professional standards for AGU
o Select the good data from the new figures
o Register for AGU
o Submit abstract to AGU
• Maintain this blog about what I have accomplished and my goals
• Explore Canberra
o Go to the National Zoo and Aquarium
o Go to the Old Bus Depot Markets

Here is one of my new figures, I’m pleased with them!

Now I just need to run the code for all my data and look at all the outputs. I’m just waiting on my adviser to see if the layout is ok and if there’s anything else we want to add to the plots before printing them. After they’re all printed we’ll sit down with them and go through and only keep the ones with visibly good data. Then it will be time to use the tomography code to invert the travel times and make my first map! I’m really hoping to do this before August 8th, when we will submit the abstract for AGU.
We went to the National Dinosaur Museum on Saturday! It was really fun; there were fossils, sculptures, replica skeletons and animatronics.

His teeth were actually pretty sharp! 

July 27th, 2012

July 23 to July 27
Well hello again all, the time seems to just slip by! That AGU abstract deadline is just around the corner; I hope I actually have some results by then. Work is going well here and I feel very productive most days. I recently discovered that for some of my data there was another, smaller event recorded near the end of the time that I was investigating for some of the stations, but it shouldn’t take too long to go back at cut the traces a little shorter. By the end of Monday I hope to have produced the new figures of the data so we can go through and pick what is good enough to keep and what has to go. Then after we’ve selected the good data we can invert for velocity structure and create the first rough tomography map! I’m very excited about this and the progress I’ve been making, I feel that the project is chugging along.
I’ll share some images of the data on Tuesday, and that’s about it for this week! In unrelated news I’ve been practicing making my own bread. It’s a good wayt to spend some time and I enjoy it. This is my third loaf:

This one has a much nicer shape than the first two because I actually bought a loaf pan. I plan on making some banana bread soon as well, if it turns out well I might even post a picture! I’m very spoiled at home and in my dorm in California where I have lots of cooking tools and pans and ingredients stocked up, but I’m very pleased with my new loaf pan and there’s a muffin tin at the house here so I’m slowly expanding my baking enterprises.
This weekend I will go and see the National Dinosaur Museum with Marija and Archana, we will miss Debjani though because she left early this morning. Hopefully there will be some cool pictures of the dinosaurs for next week’s post; apparently there are skeletons, sculptures and animatronics! I’m very excited.
Last weekend we went on a good adventure to see some kangaroos, and took some great photos. Here is one that Debjani took.

Cheers! 

July 24th, 2012

I'm finally closing in on some figures that we might be able to analyze and get some interpretations and results.
• Use surface wave group velocity dispersion for tomography to investigate the subsurface in Australia
o Go through the data and see what is good enough to use (sadly I think a large portion of the data will be rejected)
o Use the good data for tomography to make a crude map of the velocity structure in Australia
• UNIX
o Learn about Illustrator
o Comment
o Learn the tomography code
• GMT
o Make new figures for analyzing whether the data is good enough to be used. New figures will include:
 A map showing the path between the event and the station
 The dispersion for that path
 The raw seismogram from that path
 A filtered version of that seismogram
 The earthquake information (magnitude, location, date)
• CPiS
o Learn about sacmat96 so I can add useful commentary to the code
o Learn how to make a dispersion curve for the PREM model
• SAC
o Make a nice bandpass filter to show the filtered seismogram
• Understand what group velocity dispersion is used for and be able to communicate knowledgably about it
o Read papers and thesis from someone at ANU concerning group velocity dispersion.
• Be able to produce a poster which describes my project that is up to professional standards for AGU.
o Select the good data from the new figures
o Find a meaningful way to arrange the figures for the poster
• Maintain this blog about what I have accomplished and my goals
• Explore Canberra
o I don’t think I’ll be taking and more weekend trips while I’m here but there seems to be quite a bit to do here in Canberra
o Go see the National Dinosaur Museum and other attractions in Gold Creek Village
o Go to the National Zoo and Aquarium
o Go to the Old Bus Depot Markets 

July 19th, 2012

July 16 to July 20
On Wednesday at tea one of my friends asked me how long I had left here, and I said 5 and half weeks to which he responded, “Good! Six more Cake Fridays!” I think that’s a good way to summarize how I feel about my time left here, it’s short but there’s a lot left to look forward to! There’s also a lot left to do. I’ll be both happy to go home and sad to leave, but right now my main worry is not accomplishing enough before I go. Today was a good Cake Friday, a nice chocolate cake topped with Smarties (they’re like m+ms, we have them in Canada but in the States you can usually only find them in the foreign foods section). So now I have only five Cake Fridays left!
In my work right now I have processed the data from my six events and I have all the dispersion curves from that, ones for each station and ones showing all the stations. I updated the codes to plot the two different filters on the same page for each separate station and to have different colors and a legend for all the stations together. I’m learning a lot about GMT, and soon I hope to post a map that has all my events and lines connecting to all the stations where data was gathered. I probably don’t have enough data to do tomography and make a map, but we’ll see. I still need to find out if the way I’m processing the data is appropriate though and I’m assuming I’m going to have to change some things in the code before I can really produce any final pictures.
The next big thing for me to do is to find or make the dispersion curves for the PREM model and add those to my plots. It will be good to have something to compare my data to. I think I may also add plots of the seismograms to the individual dispersion curves, so I can see which curves we can clearly see the surface waves and which ones look mostly like noise.
I’ve been catching up on reading people’s blogs, it’s good to finally catch up a bit on what people are working on, and see that frustrations with coding are not uncommon. Fortunately for the most part I find the coding to be more interesting than frustrating, even when it isn’t working. I definitely surprised myself with how much I like it!
Here are some of my plots for the March 23, 2012 event.

On a side note, it’s my mom’s birthday today! Happy Birthday Mom!

Cheers!
 

July 19th, 2012

I actually can’t believe that I’m more than half way done now. I had to count the days to double check! The time has flown by so far and I think it’s only going to go faster. Here’s the updated goals list.
• Use surface wave group velocity dispersion for tomography to investigate the subsurface in Australia
o Create all the dispersion curves for all six events!
o Figure out if the differentiation step in the code is appropriate for my data.
• UNIX
o Learn about Illustrator?
o Comment my codes! I want them to be comprehensible to other people and to myself in six months when I’ve forgotten all the specifics.
• GMT
o I’ve been doing a lot of GMT practice recently, making my dispersion curve figures. I do have a GMT helper code, and I should add more information to that.
• CPiS
o Learn about sacmat96 so I can add useful commentary to the code.
• SAC
o Comment in the code what SAC is doing and how we use it.
• Understand what group velocity dispersion is used for and be able to communicate knowledgably about it.
o Read papers and thesis from someone at ANU concerning group velocity dispersion.
• Be able to produce a poster which describes my project that is up to professional standards for AGU.
o Find PREM model to compare my dispersion curves to.
o Find other maps to compare my data to.
o Find a nice way to arrange all my figures so they can be interpreted in a meaningful way.
• Maintain this blog about what I have accomplished and my goals.
o Get back on schedule with blog posts – almost there!
• Explore Australia!
o Consider going to Melbourne for a weekend? 

July 17th, 2012

July 9 to July 13
I somehow barely noticed that week going by. I got back Monday night (the 9th) from Sydney and my brother was still here with me for another week. So between a 4 day work week and having my brother around the week just went by so fast! In fact it went so fast that the homework deadline really snuck up on me. Fortunately I keep a master to do list on my computer so I didn’t forget entirely. I kept meaning to work on the homework while I was at work and then forgetting because there was more code and debugging to do. I spent some time modifying the big code so that the naming conventions and variables are more suited to my data. It was written for ambient noise data and I’m using earthquake data so there were quite a few changes to make.
Everything seems to be up and running now, the next step is modifying my scripts that make images of the dispersion curves so they are nicer images and possibly include the calculated dispersion for the PREM or other models. The dispersion curves will be most useful if we can also compare them to dispersions that other people creating maps of Australia computed. My adviser is working on getting those maps and dispersion curves for us.
The time in Sydney was wonderful and I’m going to miss having my brother here. I can’t believe I have less than six weeks left, it’s weird how fast the time has gone already and it’s only going to get faster. I hope I have something to write an abstract about soon! I do have lots of figures now though! Here is a map of an event I use and the stations that I took data from. The event occurred on March 23, 2012 and was magnitude 5.7.

You can see a place where there are a lot of stations all in a small area. I think looking at those curves could provide a good estimate for error or at least some sort of maximum difference calculation. Here are some of the dispersion curves from that event. This is the curve from QLP, one of the stations closer to the event. 

And this is a plot of the dispersion curves from all the stations that I gathered data from.

I think I’d also like to look at dispersion as a function of azimuth if there is time. Hopefully next week I’ll have some more figures to post, especially these figures with the PREM model dispersion curve added. One other thing I do need to do more is read everyone else’s blogs, and send some post-cards. I stocked up on postcards in Sydney, now I just need to write them and get some stamps.

Till next time!  

July 14th, 2012

I’m running late on my blog posts, the trip to Sydney kind of threw everything off for a bit but I’m slowly getting back to my usual schedule. This week was short but fairly productive.
• Use surface wave group velocity dispersion for tomography to investigate the subsurface in Australia
o So far I have gathered data from six earthquakes, magnitude five and above, that occurred on the land in Australia
o I need to go through that data and create the dispersion curves, maps of the event and seismometers (hopefully adding lines between the event and each of the seismometers), and then see what the next steps are
• UNIX
o I finished two different codes to plot the seismograms, one that plots each one separately and one that plots them all on the same plot in different colors.
o I want to learn how to make those figures nicer, adding legends and making a nice layout for all the separate ones.
o Keep learning about the codes I’m using – avoid the black box!
o Comment the codes I use frequently.
• GMT
o I would like to update the code I have that plots the earthquake and the seismometers used to create dispersion curves so that it plots the great circle path between the earthquake and each of the seismometers. I would also like it to automatically adjust the window of the graph to include all stations I used.
• CPiS
o One of the codes I use calls sacmat96 and I would like to know what that program actually does.
• SAC
o I don’t have any specific goals concerning SAC now but I would like to keep learning about it.
• Understand what group velocity dispersion is used for and be able to communicate knowledgably about it.
o Read papers and thesis from someone at ANU concerning the group velocity dispersion.
• Be able to produce a poster which describes my project that is up to professional standards for AGU.
o Make the dispersion curves for all six events, then start contemplating the abstract.
o Find PREM model to compare my dispersion curves to.
o Find other maps to compare my data to.
• Maintain this blog about what I have accomplished and my goals.
o Get back on schedule with blog posts.
o Include more figures from my work.
• Explore Australia!
o Consider going to Melbourne for a weekend?
 

July 10th, 2012

July 2 to July 6
This past week was a successful and short one. I felt really good about my progress because I got the big code to work. The big code is how I think of the code Mallory gave me that creates the dispersion curves. There is still quite a bit to do with the code because I need to tailor it to give useful results for what I’m working on, but it now does exactly what it was written to do with no errors on my computer which was a major accomplishment. So now I’m at a place in my research where I might be able to write a more complete and accurate project description, which is exciting. Stay tuned for that.
The final output of the big code for this week was two different dispersion curves, one not using the phase matched filter and one using the phase matched filter. These dispersion curves aren’t particularly helpful right now because they are an average of all the stations response from one event. I’m now going to work on making figures that have all the dispersion curves on them without averaging.

I concluded this week by giving a short presentation to the group I work with. The group now includes three PhD students and three other interns, it has grown since I arrived. We had another PhD student, Maria, and another intern, Jerry, arrive this week. Maria is from Croatia and Jerry goes to Princeton, I’m looking forward to getting to know them better. I really like working in this group, everyone is very friendly and helpful.
I missed the webinar because it occurred at 4am my time while I was in Sydney, I tried to wake up but it didn’t go so well. I’m looking forward to seeing it though, and catching up on that homework.
My brother and I spent Thursday night to Monday night in Sydney and we had a great time. (That’s why this post is late as well.) We went to the aquarium, the wax museum, the wildlife park and the Sydney tower. We walked across the bridge and went to the Opera house and the Royal Botanical Gardens. Here’s the part where I make up for not having enough pictures in my other blogs.

Till next time! I hope everyone is having as great a summer (winter) as I am!  

July 3rd, 2012

I always find it a little disconcerting to update this file once a week because everything changes so much in a week. But that’s definitely a good problem to have!
• Develop a specific project and write a set of goals for that project.
o Use surface wave group velocity dispersion for tomography to investigate the subsurface in Australia
o The dispersion allows us to look deeper than just the crust, hopefully get good data for longer wavelengths so we can have resolution at larger depths
o Find data from recent (past couple of years) large (greater than M 5.0) earthquakes
o Use the data to create dispersion curves
o Use the dispersion curves to do tomography
o Compare tomography from my results to other maps of Australia
 This will be interesting because most tomography in Australia is done using data from earthquakes that occur around Australia but not actually on the continent
• UNIX
o I have quite a few new codes I’m using to do the frequency-time analysis that develops the dispersion curves. I understand (I think) all of those scripts and what they’re doing, but they call other scripts which I don’t understand. In total I run four scripts which call nine other scripts, two of which are written in FORTRAN, and two CPiS programs. Then I wrote three other scripts I use in conjunction, one to prepare the data, one to clean up and organize all the output files from the main scripts and one to (hypothetically) extract and plot the final dispersion curves. That last one is what I’m working on right now; I just need to fiddle with GMT a bit more.
o I’m learning a lot about UNIX and shell scripting, my goals for now are keep working on my codes, improving and writing new ones as is necessary, and keep learning about the codes I’m using – avoid the black box!
o Comment! When I started out I was so good about commenting every step! That’s not efficient but if I want to use this script in the future I will probably have forgotten everything about it and so I should spend some time closer to the end of this projecting commenting on scripts I use regularly. I think that if I used it to make a figure for my poster at AGU I should have a script that is well documented and explained to back it up. We’ll see if I have time to meet that goal.
• GMT
o I made a nice code that follows on the code that prepares the data which reads the headers and extracts all the station names and locations and the event location and puts all that on a map. So far it’s not a smart code in that it doesn’t figure out the region and the scale by itself, but I have plans for that.
o I’ve experimented with some code to plot seismograms and now I’m going to put that to use in making dispersion curves. More news (and hopefully images) from this soon.
o I’m really excited to finally be making some figure that actually might mean something! Still a long way to go but it looks like I have a project that I can make some good progress on now and I’m feeling very optimistic about this.
• CPiS
o The scripts that Mallory gave me call several different CPiS programs and use SAC. I’m glad I spent time learning both those programs. For now I don’t have any new specific goals regarding CPiS, other than attempt to understand exactly what they’re doing when I use them in other scripts.
• SAC
o The code that prepares the data does most of the work in SAC, changing the headers and updating the origin time. I mostly understand what the code Mallory gave me uses SAC to do and my goal for that is fully understand those parts of that code.
o I don’t really do any data filtering myself, it’s all integrated into the CPiS programs, so I’d like to learn more about that too.
• Understand what group velocity dispersion is used for and be able to communicate knowledgably about it.
o I have some papers and a thesis to read about this topic and its uses in tomography so I’m looking forward to being more informed about that. Hopefully soon I’ll have some more specific questions and better ideas for exactly what I’ll be working on and looking for.
• Be able to produce a poster which describes my project that is up to professional standards for AGU.
o Once I have some figures I can start thinking about writing an abstract. That deadline for AGU is definitely going to sneak up really fast.
o I’m making an informal presentation to my group on Thursday about what I have been and will be working on. This will be helpful for defining what I hope to accomplish and getting an outline of what I’m working on. I think this will help with abstract writing and general clarity about my project, for myself and my peers.
• Maintain this blog about what I have accomplished and my goals.
o Keep posting goals on Tuesdays and general blog on Fridays. Keep including pictures. Start including pictures from my work.
• Explore Australia!
o My brother and I will be in Sydney from Thursday to Monday so that will be nice. I’m looking forward to having lots of pictures!
 

June 29th, 2012

June 25 to June 29
This week I made some good progress. I now have a code that prepares all the data from the seed file so that all the SAC headers are updated and the origin is the earthquake time and all the traces are the same length. Mallory, a PhD student here, gave me a code that extracts the group velocity curves. It mainly uses a frequency-time analysis which has multiple filters to determine the arrival times and group velocities at a variety of periods.
The code I was given requires many steps and many other codes and programs to run so I first spent time going through it and getting it to work for some of Mallory’s data on my machine. This took quite a while because all the different file paths and other programs called by the script had to be set up or modified. It was also written in bash and I had been learning about writing in tcsh, which isn’t completely different but there were several new syntax things to learn too. We did manage to get everything working for the original code and data on my computer.
Then the next task is to get it working for my data. Fortunately the program I wrote for preparing the SAC files did everything we needed to do to prepare the data except for change the delta so there weren’t quite so many data points. That was an easy fix though and I felt really good that I’d actually created a helpful script that did what we needed it to. So my SAC files are ready for analysis but I need to change some things in her script because she was doing noise analysis and I’m looking at earthquakes. Most of what I’ve had to change so far is file paths and naming conventions but I also need to make sure this is providing a helpful output. That will be what I work on next week.
On a totally different topic, one of the traditions here at RSES is teatime. It’s a great way to meet the other students and take a break. The students also have a teatime tradition that every Friday is Cake Friday and someone brings cake for the group. Most of the students here are gone right now, taking a winter holiday or doing fieldwork or at a conference. So it’s a small tea group and it seemed that everyone had taken their turn to make cake recently, so today it was my turn to make the cake! I did not have any baking ingredients but I checked in the house and there was a muffin tin so I decided I would make muffins. It was fun to bake! Another really good thing about this is that it forced me to buy the basic ingredients for baking. Cake Friday is fun. I like the students here and I’m grateful they’re so friendly!
On other unrelated topics, my brother arrives tomorrow and I’m really looking forward to seeing him! Also, the campus looks really beautiful when it’s foggy and cold in the mornings.
 

June 26th, 2012

Hello again! I think this week the goals have changed quite a bit because I’m making good progress on learning coding for all the programs.
• Develop a specific project and write a set of goals for that project.
• UNIX
   o I haven’t made any progress with looking at the script that converts sac files to ascii files but I am making progress on UNIX and shell scripting in general.
   o Dr. Tkalcic gave me two of his scripts that take some sort of table input and make nice GMT plots. I’m currently working through these and learning a lot from them but I’m still pretty far from being able to fully understand them.
   o Every day I learn something new about writing code and I was especially pleased today and yesterday to get some for and while loops working. I am also slowly learning about awk. It’s rewarding to see my codes become more efficient and effective.
• GMT
   o I managed to create the code I wanted that converts the SEED file to SAC filez, converts the SAC files to ASCII files, reads the min and max values from that file and then sets those to the window for the plot and plots all the seismograms using psxy. But after doing that I found out there wasn’t much use for a bunch of separate images of raw data.
   o So now my goal is to make a map of Australia that shows the location of the seismometers and the earthquake I’m looking at and then create nice figures of the traces. I’m not yet sure what I have to do to the traces before I can make nice figures though.
• CPiS
   o In order to use the CPiS program do_mft, which calculates the group velocity dispersion from a user defined fit, I need the vertical, radial and transverse components to be the same length with the same number of data points. I need to prepare and filter the traces before I can actually calculate anything meaningful from them. This is where SAC comes in.
   o I also want to learn a new CPiS program sacmat96 because apparently it does the same thing as do_mft but with automated picking instead of user defined so that should be more efficient and equally accurate.
• SAC
   o I want to be able to use this program to filter and transform my data so that it can be used in CPiS and GMT.
   o I have learned how to change all the headers so that the event time and location are recorded and then change the origin time to the time of the earthquake. Then I can move the origin back to 0 and cut the traces so I’m only looking from the time of the earthquake to 1600 seconds later which is long enough for the surface waves to have died off. Then I can rotate the north and east traces to radial and transverse.
   o The next thing to learn is how to choose an appropriate window to cut the data to for each trace to best observe the surface waves. If I can find a way to automate that process then the traces will be ready for use in do_mft or sacmat96.
• Understand what group velocity is used for and be able to communicate knowledgably about it.
   o I’m going to start reading papers about group velocity dispersion tomography now which is exciting.
• Be able to produce a poster which describes my project that is up to professional standards for AGU.
   o I’m looking forward to seeing all the other interns and also some people working in my group and on my floor at AGU. As of right now I don’t have anything to put on a poster other than the IRIS and ANU RSES logos but I think that soon I’ll be making some figures worth looking at.
   o I will need to start thinking about writing an abstract in the not-too-distant future because I think that deadline for AGU is going to sneak up really fast.
• Maintain this blog about what I have accomplished and my goals.
   o Seems to be going well, I need to take some more pictures for this week though!
• Explore Australia
   o My brother arrives on Saturday and I’m really excited to spend some time in Sydney!

June 22nd, 2012

June 18 to June 22
I’ve sort of gotten used to it but every now and then it really hits me exactly how much it is winter here. Despite how much I remind myself that it’s going to get colder, not warmer, I’m still surprised to see the frost in the mornings.

I took a picture of this site in central downtown here in Canberra because I’m excited about it, but also because it’s a good reminder that if we can skate outdoors in the city from June 29 to July 29 that means it must be (literally) freezing outside!

Sometimes it’s truly miserable and rainy in the mornings, but we still get beautiful sunsets every night.

The picture doesn’t really do it justice but it’s nice to watch from ouw window at work. I’m fortunate that the building is nice and warm though, and I get to spend my days at the computer looking at seismograms and code!
I spent some time this week trying to get data from IRIS and eventually managed to request the right seismometers at the right time. I’m attempting to look at an earthquake that occurred in central Australia on March 23 this year. It was a 5.3 (or 5.7, depending on where you get the data) and should provide enough surface wave action to look at group velocity dispersion.
I’ve wrote a script which I’m quite pleased with this week. It unpacks a seed file from IRIS that has the North, East and Vertical components of several stations in it and uses GMT to plot those traces, and it uses minmax to determine the window size of the plot and titles each plot with what trace it is using. While it doesn’t help with analysis I’m sure it will be a useful basis for a code that unpacks seed data, does some basic manipulation of the data and updates the headers in SAC and then plots the traces together for comparison.
My next step is to write a code that filters the data using SAC. I learned today that to prepare my raw data I need to change the SAC headers so they know about the earthquake event location and depth, and also so that the origin time is the same on all the traces. I think it will be worthwhile to have a script that does this because with just 31 stations that all have three traces it took me several hours to manually change all the headers.
After I have the traces properly set up I will need to rotate the North and East components to Radial and Transverse, cut down the number of data points, and then figure out which filters to apply. I’m not entirely sure about data filtering and how it works and how to write scripts for SAC yet so it’s time for me to sit down with the SAC manual and also learn more theory about filtering.
Once I have all the traces properly prepared then I can start using the CPiS program do_mft which does an interactive multiple filter analysis to determine the group velocity dispersion. I’m not entirely sure what I’ll be doing with the dispersion curves once I have them so for now my script concerning that ends with the comment #Now What?
I am on a very steep learning curve here and really enjoying it. I’m so grateful that there are many people here who help me with all sorts of coding and conceptual questions. There’s plenty to do here and I’m enjoying the work. ‘Till next time!  

June 19th, 2012

Greetings! I aim to update and repost this goals list once a week, reflecting my progress on old goals and showing new goals.
• Develop a specific project and write a set of goals for that project.
     o Learn how to get specific data from IRIS DMC
     o Unpack that data using rdseed
     o Learn what to do after unpacking the data
• UNIX – basic competence at navigation and shell scripting.
     o I am using a script written by one of Dr. Tkalcic’s other students in 2007 that takes the SEED data from IRIS and puts it into a two column time-amplitude ASCII format that psxy in GMT can use. I would really like to understand this script as more than just a black box because GMT’s version of sactoasc doesn’t produce anything even remotely similar. Also, understanding this particular script will help with my general understanding of c scripting, SEED data, and Linux.
• GMT – basic competence in map and figure construction.
     o Be able to take a seismogram from IRIS and plot it using psxy. Now that I have a file I can use in psxy and I have figured out how to plot it I am trying to streamline the process by creating a code that converts the SEED file to a SAC file, converts that SAC file to an ASCII file psxy can understand, reads the min and max values from that file and then sets those to the window for the plot. I can do all these things right now but not at the same time so that’s the next goal.
• CPiS – basic competence at general use and group velocity calculations.
     o Be able to take a seismogram from IRIS and calculate the group velocity dispersion. The tutorial script that allowed me to calculate group velocity dispersion before relied on the fact that the script was also creating the synthetic seismogram, so it created the Eigenfunctions for the Rayleigh and Love waves. I would like to learn what those Eigenfunctions mean and also figure out how to calculate the group velocity dispersion without that input.
• Understand what group velocity is used for and be able to communicate knowledgably about it.
• Be able to produce a poster which describes my project that is up to professional standards for AGU.
     o I’m already starting to get excited for AGU, I’m really looking forward to seeing the other interns (miss you guys!) and it looks like some of the people working here might go, not from my group other than Dr. Tkalcic but some working on the same floor as me.
• Maintain this blog about what I have accomplished and my goals.
     o Include a picture every week, which implies that I will have to be taking pictures! I feel I have been remiss on this front.
     o If I make a particularly helpful code, include that as well. I think if I polish up my pscoast code and elaborate on it, and make something similar for psxy for making seismograms then that would be worth sharing.
• Explore Australia!
     o My brother is visiting for two weeks and we are going to spend 4 days in Sydney. Dr. Tkalcic has graciously agreed to let me take off the Friday and the Monday as long as I can keep up with the work. This arrangement suits me because I’ve been coming in for a few hours on the weekend, to do a little work, do my shopping in town and use the internet connection at the university which is much better than the one at my house. I’m very excited to go to Sydney for 4 days and I will be posting lots of pictures then! 

June 15th, 2012

June 11 to June 15
G’day!
I have to admit I’m a little disappointed that nobody here has actually said that to me yet. It turns out the Research School of Earth Sciences (RSES) is a very international community. I’ve met many more people from Germany and Russia than from Australia. On my floor there are people from India, Indonesia, and New Zealand, and my adviser is from Croatia. I met “our token Australian” at tea last week and found out that native Australians are very much in the minority here.
This week has been another good one, I came in for a few hours on Monday to install Computer Programs in Seismology but the building was quite empty because it was a national holiday here, for “the Queen’s Birthday.” It wasn’t actually specified which Queen’s birthday it was. After getting CPiS up and running I went to the Parliament house with some of the people I work with. It’s a very beautiful building and we took a lot of pictures.

For the rest of the week I spent my time reading the CPiS manual, doing the sample runs and using the Surface-Wave Synthetics and Group Velocity Determination tutorial. On Wednesday I read through the code for the tutorial and commented after each step exactly what the code was doing. It really helped me understand which parts I needed to focus on in the manual and what concepts I needed to ask more questions about.
On Thursday I met with my adviser and we talked about tomography and group velocity dispersion. We discussed how we can use the signal from a seismogram, filter it at different frequencies, look at the group velocity and how it changes with different frequencies, and use that dispersion in an inverse model of the earth. So effectively we can use the group velocity dispersion to make different tomography maps, or different layers of the same tomography map. My understanding of this is very rudimentary now but I’m excited to get a better grasp of the idea.
For my project I might be using a dataset from Iceland or Australia depending on what the other people working this summer are going to be using and doing, but it looks like I will definitely be attaining seismograms, filtering them, analyzing their group velocity dispersions and using that information as a type of tomography to better resolve the subsurface.
My task for the coming week is to get some seismograms from the IRIS DMT, plot them using GMT, use SAC to view and filter them, and then run the group velocity dispersion analysis on them. I’ve requested the data from the IRIS NetDC request service and I’m just waiting to hear back. I’ve asked for data from a specific earthquake that occurred in central Australia on March 23 and soon, hopefully, I’ll be able to analyze it. I’ve also started going through the tutorial script again and trying to figure out which parts I need and which parts I don’t when I don’t have to artificially create the seismogram before analyzing it.
While I’m waiting to hear back from IRIS I will probably make some synthetic seismograms and practice on those, and hope that soon I’ll start to have a better conceptual understanding of what it is I’m actually doing.  

June 12th, 2012

Hello again!
It is June 13 over here and I managed to finish up a couple things from last week. ITS did install the gcc and g77 compilers I needed for using Computer Programs in Seismology on Friday afternoon and I managed to install CPiS and get everything running soon after. I still don’t have a particularly specific project description but I do have some more specific goals to concentrate on for now.
For my first week I had fairly simple goals such as figure out how to take the bus to work and pay my rent, find a grocery store and feed myself. My only tasks at work for the first week were to familiarize with UNIX and GMT by creating a map of Australia with all the IRIS seismometer locations marked (see previous post) and to install CPiS. Having accomplished those my task for this week is to use the CPiS tutorials and manual to understand the basic workings of the program and to figure out how to use the programs to calculate the group velocity of a surface wave. There is a specific tutorial for that but after working through the tutorial I realized I didn’t understand the program enough to really understand what I’d just done. So I’m working my way through the manual now and am trying more basic tasks.
Here are some of my more specific short and long term goals for this summer.
• Develop a specific project and write a set of goals for that project.
• UNIX – basic competence at navigation and shell scripting.
• GMT – basic competence in map and figure construction.
• CPiS – basic competence at general use and group velocity calculations.
• Understand what exactly I am doing/calculating with the group velocity.
• Understand what group velocity is used for and be able to communicate knowledgably about it.
• Be able to produce a poster which describes my project that is up to professional standards.
• Maintain this blog about what I have accomplished and my goals. 
• Explore Australia!

On an unrelated note I have seen quite a few exciting birds here, and I took some pictures of a cockatto and two different types of parrots.
 

June 12th, 2012

June 4 to June 8
Hello there! My name is Dulcie and I am an IRIS intern working at the Australian National University with Dr. Hrvoje Tkalcic. I will be posting to this blog weekly about my work and my progress and life down in OZ.
I left New Mexico after orientation week at 11am on Saturday and arrived in Canberra at 1:30 pm on Monday. The flight took longer than usual because the entire southern coast was covered in fog that morning so we had to stop in Fiji to refuel and wait. Unfortunately it was about 4am in Fiji and all I saw was the runway. My advisor, Hrvoje Tkalcic, patiently waited as they kept changing my arrival time to pick me up at the airport. He showed me the Australian National University campus where I’m working this summer (winter down here) before dropping me off at the house where I’m renting a room.
On Tuesday morning I started work and was introduced to all sorts of people working here, other interns, Masters and PhD students and Post-Docs. Working in the same computer lab as me are Archana, Debjani, and Surya. Archana is working on a combined Bachelors and Masters, Debjani is a Masters student and Surya is working on his PhD. We are all working on separate projects but it is nice to have other people to interact with and to ask for help. Surya has been able to answer many of my programming questions and show me the way to ITS.
My first task was to familiarize myself with UNIX and GMT, so I spent Tuesday, Wednesday and Thursday recreating maps from the GMT manual, and creating maps of Oregon and Australia. I have learned a lot about GMT and UNIX, but I know my skill level is still very rudimentary. My next task is to download, install and learn how to use Computer Programs in Seismology. The computer I’m using now lacks certain compilers required to do this, so I have to wait for ITS to come fix it. In the mean time I will be attempting to place dots on my map of Australia that indicate seismometer locations.
As of right now I don’t have a clearly defined project description, after I learn more about Computer Programs in Seismology I will discuss with Dr. Tkalcic what specifically I should start working on. We have talked briefly about using CPiS to look at surface waves and follow the group velocity of the main energy peak for a dataset from Iceland. When I have a better defined project I’ll post it to the project description section on my profile. And once my maps are looking beautiful I’ll find a way to post those as well.
I’m enjoying my time here so far, everyone is very friendly and I’m enjoying the novelty of frost in June, for now! I also saw some parrots which I found very exciting and I’m learning to navigate the bus system. All in all everything is going well so far!

Update as of June 12
My map of Australia, with IRIS seismometers marked by green dots.