Chelsea Potier is a student at St. Norbert College currently completing her research at U. California, Riverside under Elizabeth Cochran.
In 2010, a seismic array was deployed near Cholame, CA. I will work with the data from that array, along with nearby stations, to create a database of continuous seismic data. That data will then be analyzed for earthquakes and tremor that occur on the San Andreas Fault and nearby fault strands. I will explore how those earthquakes and tremor are related, and where they occur relative to each other on the fault.
This summer has been quite an experience. Living halfway across the country from everyone I know helped me to grow as a person, while the research opened my eyes to what graduate school may be like. I learned that I'd rather stay in the Midwest when I go to grad school, closer to my family and friends. The large distance was hard on me, and I feel like even staying in the Midwest would give me some peace of mind, knowing that family or friends were only a few hours away. I also learned that, although I enjoyed seismology, I may want to explore a career path that involves more geology. I guess I'll have to see what I can figure out.
Although I was unable to get as far as Elizabeth had originally planned, I was very happy with how far I got. Elizabeth let me know that she had been very ambitious when planning our project for the summer, and that although I didn't finish the project, I actually got farther than she originally expected me to. I basically achieved every goal I set out to, except my last goal, which involved analyzing the tremor and earthquakes for spatial and temporal patterns. This was by far the most ambitious of the goals I had set, so although I was unable to achieve it, I'm still pleased with the other goals I accomplished. I think one reason we didn't completely finish the project was some of the issues we were having with the code I was writing. Little errors kept occurring, but all those little errors ended up taking a lot of time out of our schedule. These errors taught me that sometimes things don't go your way, but you have to keep pushing through to get as far as you can. I feel like graduate school is going to involve a lot of this - getting as much done as I can on my projects, while pushing through little errors. I also understand that graduate school will be more independent than undergraduate, something I got a taste of this summer.
This summer taught me more than I ever expected it to - in and out of the research room. It was invaluable experience, something that helped me grow, while giving me the opportunity to live in a region of the country I never had before. I appreciate it so much, and would say to anyone even considering it to give it a try.
This week was pretty productive. I ran the second cross correlation script I had written, and came up with relative arrival times for different tremor events. We ended up using a detection threshold, cutting down the results to the top 488 events for me to focus on. My next job was then to group similar events together into families. If a certain segment matched more than one time period, or a time period matched more than one segment, those were grouped into families. As a first pass, I came up with around 180 families for each station. Looking back at the data though, I realized that some events may have been counted more than once. The event may have been counted twice if both the segment and the time period were off by a half second compared to a different event. Taking this into consideration, I'm writing a script to eliminate these possible repeats. Elizabeth and I decided on three seconds as a cutoff for whether or not something is a repeat. If both the segment and the time period fall into that three second region, they're going to be considered a repeat and not counted. We can always modify it, but as a first pass this is what we thought would work.
Our field day went well, we got all the stations installed, but it was a little rough. Working in teams of three or four we installed ten stations (five for each team). The main problem was that the road to get there had been washed out, so we had to carry the batteries over ridges to where the stations were going to be located. It took us around six hours to complete all the stations, and afterwards we were all dead tired. It was nice to get to work in the field again though, and I enjoyed being able to set up stations, which I hadn't done since our orientation week.
Next week is the last week, and I feel like I still have so much to do! Hopefully everything gets done, and I have a respectable poster to present.
This week was full of troubleshooting. First I had to make a correction in my cross correlation script and rerun it. There was a problem with some of the stations and my index for pulling out the correct cross correlation was wrong. The stations that had the main problems were the recently deployed stations that I was supposed to be focusing on. I had been using "ceil" for rounding in matlab, but then found I needed to use "fix" instead. I switched it for all the stations and reran it. This second pass produced more problems, as it ends up matlab decided to round 20.0000 to 19 for all the stations that previously did not have problems. Not sure why, as it was supposed to round to the nearest integer, but then I had to rerun it again. This basically took all week because this is the program that takes two days to run. While I was waiting for it to finish running, I wrote a second cross correlation script, similar to what Brown et al. (2008) used in as their second pass. I will run this one on the top few hundred high cross correlation values. This will give me the relative arrival times of tremor at each station, which I will then use to locate the tremor. I'm hoping I have time to locate the tremor in these last two weeks, along with making the poster for the conference. It'll be a time crunch.
Next week I get to get out of the office! Besides the first day, I've been working at my computer. I love the computational aspect of seismology, but it'll be nice to get outside and help install ten seismometers near campus. Hopefully it's not too hot though!
This week went by really quickly! Rather than working on MacRay, we decided to wait on that and work on it after we had gone through the cross correlation data that my script had outputted. As a first pass, we searched for where the cross correlation value was the highest across the entire network. I then wrote a matlab script that pulled out a 20 second part of the seismogram around that high cross correlation value, and plotted that 20 second segment along with the six second segment that it was being correlated with and a 20 second window of the correlation values. I did this so we could see possible move-out of tremor across the network. I marked where the highest cross correlation in that 20 second window occurred in each station, and moved the six second segment to plot directly above that. This allowed us to compare the segment and the seismogram rather easily. It took me a while to work through writing the script, so we've only run it on the highest cross correlation value so far, but we will hopefully run it on more values next week. We will use the relative arrival times of the tremor in MacRay (if we can get it to work) to estimate where the tremor is located.
I cannot believe there are only three weeks left. I feel like I have so much more to do! We planned on leaving approximately the last week for making the poster, which means I will have to have most of the actual analyzing done these next two weeks. I still feel like we have a lot to do, but we'll take it one step at a time. I read back through my goals, and am very happy about the progress I have made. I'm much more comfortable with matlab and unix, and I've learned a lot more about earthquakes, tremor, and seismology in general. My last goal, the one where I look at earthquakes and tremor and how they are related, is the only one I feel like I've made little progress on. We are working towards it though, so I'm hoping in these last three weeks I will be able to cross that one off my list too.
Early this week I finished my cross correlation script, and ran it on one component of one station over around 1300 seconds. This took an hour, which was much longer than we expected. We had hoped to run it on all 3 components on 31 stations over longer periods of time. If we end up needing to use the script for that many stations and components, we may have to modify it to make it run faster. For now though, I just ran it over all the stations on one component over the selected 1300 second window. It ended up taking almost two days to run, so I haven't had a chance to work with the data it outputted yet. Next week I will work through the data, deciding where my cross correlation cut off will be, and determining at what times the high cross correlation occurs.
I also worked on locating the earthquakes and tremor. This is the most confusing part of the internship so far, so it's taking me a while to work through it all. I used the SCEC 3D community velocity model and a 2D grid of data points running across the fault to determine a grid of velocities. We're hoping to use MacRay to read in this 2D grid and determine relative arrival times at the surface. We would then use those relative arrival times to determine approximately where the tremor was located. MacRay is proving problematic, as although a manual and examples were promised with the program download, they were nonexistant. For a couple days I was trying to figure out how to work it without the manual, because Elizabeth had never worked with it either but it was impossible to even determine what file type my grid would need to be to be read in. Elizabeth then mentioned that a postdoc student of hers worked with MacRay on my computer, which led to me searching the entire computer for a MacRay manual, and after some digging, I was able to find it! Although it is not the same version of MacRay that we are using, it was still a start. Hopefully I can eventually get it working.
Sadly, the surfing trip got cancelled, because not enough people signed up to go. I should have known better than to sign up for a surf trip the same day as the world cup final. Hopefully I'll still get to go snorkeling later this month. I realized today that I am officially halfway done, which is crazy considering I still have a lot of work to do. Elizabeth and I sat down today and figured out the last four weeks together so I have a plan of attack. I'm hoping we get everything accomplished and a pretty good poster ready for the conference later this year.
This week was slightly frustrating, as I didn't end up getting as much done as I wanted to because Elizabeth was gone, and the graduate student I'm supposed to go to for help was also gone. After getting stuck on writing the script for cross correlating the tremor, I moved on to studying how earthquakes are located. I'll hopefully finish the correlation script next week. We are going to try to locate the tremor to see if there is a relationship between where the earthquakes are located and where the tremor is located. I found a few good papers on locating earthquakes, and am working on getting through all the crazy terminology and scary equations to actually understand what is going on. I also wrote up what I had done the first two weeks so that I actually remember what I did when it comes time to make the poster. If I got stuck/done doing all of those things, I worked on visually going through the envelopes like I did last week. I had only gotten through the first few days of our week of data when we moved on to correlation, but Elizabeth suggested continuing to go through it if I needed something to do.
Good news - I signed up for a surfing lesson! I'm kind of nervous about this, as I'm not the most amazing swimmer in the world and I don't know anyone else that is going, but I figured it would be a good way to meet people. As of yesterday I was the only one signed up though, so if no one else signs up by next Sunday, it probably won't happen. I'm keeping my fingers crossed it does. I'm also thinking about signing up for a snorkeling trip later this month. I've gone snorkeling before and loved it, so I think it'd be really fun to go.
This week I worked mainly on running the data through a matlab script written by Ken Creager that bandpass filters from 2-8 Hz (prime tremor range) and creates envelopes of data, which was then supposed to be written out as AH files. Well...this didn't go quite as planned. Elizabeth was gone Monday-Wednesday, and the grad student I'm supposed to go to for help hasn't been around because she was going home. Thus my main means of help was through email. Errors kept popping up, and it took a while for us to figure out what was wrong through email correspondence. It ended up being something with an updated version of matlab no longer being able to write out the AH files using the script we were(I think), so we had to change the output files to SAC files, which solved the problem nicely. While we were working through the errors, Elizabeth had me work through a GMT tutorial, and then modify an existing GMT map to display my stations and the catalogued earthquakes. I ended up with a lovely looking map, which I'm still working out a few kinks on. I tried to upload it just now, but couldn't figure out how to. Oh well.
I then began working through the data after it had been run through the script. I looked at only one component, searching for areas in time in which there was a similarity among the stations. This is somewhat time consuming, as I had to step through each day in 4000 second increments. I recorded any times that there was a similarity and then looked back at the original data during those times to see if I could identify tremor. This was a slightly frustrating process because the tremor was hard to pick out, but I did find some examples of it that Elizabeth agreed were interesting. The next step is to cross correlate during the tremor intervals in a method that involves taking a six second segment and cross correlating that with the rest of the interval we're looking at, setting a lag of 0.5 seconds between each six second segment. For now we're focused on just an interval when tremor is occurring, but we will eventually expand this to cross correlate for the entire day. Elizabeth is going to be gone again starting on Tuesday until the following week Wednesday, so hopefully I can get it working in that time. Otherwise she gave me some other things to work on, which includes writing up methods of what I have done so far for our poster. I'm sure I'll be able to keep plenty busy.
So far things are going pretty well. I previously told you about what we did Monday, but I did not tell you the frustrating fact that the seismometers we removed where in perfect position to catch the earthquake that occurred during my last blog. Sadly, we removed them just a few hours too early. Tuesday began the real work on our project. I was shown to a small office with three desks in it. There are two other undergrad interns that work in here with me, though we all have different things we're working on. My first task was to search through all Northern and Southern California seismic stations in a certain latitude/longitude range to see if they had continuous data during the week we are focusing on. I plotted the stations in google earth and downloaded the data from each station. After downloading the data, I compiled that with the information from the stations Elizabeth set up before I arrived. My next task was to look through the data using sac for catalogued earthquakes that occurred close to the stations during that week. If I found one of the earthquakes, I then created a postscript file of the seismograms showing that earthquake. This turned out to be a much longer task than anticipated because the server crashed, and I was unable to work on it for a large chunk of the day yesterday. Instead, Elizabeth gave me papers to read which outlined a similar way to how we are going to analyze our data for tremor. Today I finished looking for the catalogued earthquakes and am starting to analyze the data for tremor. As far as my goals go, Unix and I are starting to get along better. I haven't really used Matlab yet, but will hopefully be starting to today when the server comes back up after being temporarily shut down to be fixed. I am still very homesick, but am hoping that as time passes this gets better.
Oh, and on a completely different note, my new bike (which I bought less than a week ago) has a flat tire, which I discovered after I was done working. This meant I had to walk the whole way back, pushing my bike. In a strange way, the whole situation just made me laugh. I wasn't laughing this morning, however, when I then had to walk 20ish minutes to get to work. Hopefully I can get it fixed soon.
So I know I'm a little late getting this on here, as I started working today, but access to the internet was very sketchy for the past week, and the internet access was just fixed in my house today. After orientation week I had a week off. My parents flew out to LA and met me there, and we took a mini family vacation around the area. Saturday they dropped me off at my new home for the summer. Today was my first day of actual intern work. My advisor thought it would be good for me to see how they remove seismic stations. I went with two grad students, Corrie and Kayla, to the California/Mexico border to remove five seismic stations. They set them up a little differently than we did during orientation, so it was good to see. It was HOT though. 105 in the desert is not easily done without a ton of water. And I drank a ton. Tomorrow I actually start what was originally scheduled for me to be doing, which is working in the computer lab mostly (at least I think). I did get to meet Elizabeth though, and she seems really nice. Whoa. As I was writing this, there was an earthquake. Pretty sweet. It was little, but still, enough that we could feel it. The best part is that all my housemates immediately went online to see where it came from. It's pretty sweet living with earth scientists. Ok, back to business...my goals. Here goes:
1. To learn more about Matlab and Unix. I knew nothing about these programs before orientation week, and I'll be using them quite a bit from what I can gather in the outline. Matlab is what I especially want to focus on, as I can see many practical applications for it. Unix...well, right now we're not getting along. Hopefully we can find a common ground throughout this internship.
2. To be able to identify tremor and earthquakes. I have no idea how to pick out tremor, and before orientation week, I had never really picked out earthquakes either. Considering I'm going to have to do this, I think it's pretty important.
3. Get over my homesickness. I've been away from home before, it's just more the people from home I miss. Moving to college wasn't as bad because I had other people going through the same things as me. Here it's just me, with links to all the other interns only through blogs. Ten more weeks seems like an endlessly long time, so I'm hoping work will keep me busy and I'll get to see everyone I miss again before I know it.
4. I was going to put feel an earthquake, but now that's kind of been accomplished. I figured I should just put it on here anyway since it was one of my goals.
5. Learn more about seismology and geophysics in general. I haven't taken any previous courses in either, but have always been interested. I'm hoping this summer will inform me whether or not I'd be interested in doing this in grad school.
6. Learn more about analyzing earthquakes and tremor together. One of the goals in the outline of my internship was to analyze earthquakes and tremor for spatial and temporal patterns. Learning to do this may be the most challenging goal I have. Hopefully I can accomplish it.
That's all the goals for now. I'm sure I'll add more daily, so these won't be the only ones, but they're a start.
So far, orientation week has been pretty amazing. After flying out to New Mexico (first time in the state!) I got to meet all the other interns, and they're all really fun. It's nice to talk to other people with the same interests as me. Yesterday we installed seismometers and spent some time in the classroom learning the basics about the earth and seismology. I thought the install and the classroom time was really helpful because I don't have any seismology courses at my school. This morning was a hike, and it was beautiful! I'm hoping the rest of the week helps teach me more about seismology before I head out to Riverside for my internship!
Ok...continuing a couple days later. Yesterday was the seven mile hike. Overall it was beautiful, but I'm so sore today! We've been learning so much in the last few days. Today's the last day :( Tomorrow we all fly out. I really think all I've learned this week will help me in my internship. I'll keep you informed with how it goes!