Clara Daniels is a student at Georgia Institute of Technology currently completing her research at Georgia Institute of Technology under Dr. Zhigang Peng.
The focus of this project’s research is to examine seismicity changes in intraplate earthquakes. A systematic search for undetected events around the 2013 Ms7.0 Lushan earthquake will be conducted using a matched filter technique. Newly detected events will be relocated and used to search for spatio-temporal evolutions around the mainshock. Upon completion of event detection, I will work on merging the Lushan data with an additional data set of Lushan events currently studied at the Chinese Academy of Sciences. For the first half of this program, I will be at the Institute of Geology and Geophysics, Chinese Academy of Sciences, in Beijing, China. The second half of research will be hosted at the Georgia Institute of Technology in Atlanta, Georgia. This will facilitate the combination of both data sets.
The last week or two, I have been primarily focused on searching for potential fault zone head waves (FZHW).
FZHW’s are characterized by a low amplitude and long period. These waves also have reverse polarity compared with the direct P arrivals, although this seems to apply more to strike-slip faults than the thrust fault being studied in the Lushan region. So I am not looking very closely at the polarity of the first arrival when doing phase picks in SAC. An example of a likely FZHW as given by my advisor:
Using the templates I cut for about 4100 events across 28 stations, I started by narrowing down the dataset. My advisor, Dr. Peng, had done some preliminary searching for FZHW with an alternate set of permanent and temporary stations in the same region. In short, I looked at waveforms from nine stations near stations in the other network that show promising characteristics of FZHW’s.
Narrowing down the dataset to scan for FZHW involved imposing additional restrictions on the dataset. For example, I choose template waveforms only on one day, 4/28/13. I also chose events with a magnitude less than 5. This is because a source effect from large earthquakes can masquerade as FZHW.
I then ran a P/S wave picking code developed by Zefeng Li, a graduate student at Georgia Tech. Some of the automatic P-wave picks helped me better define the onset of P waves in the SAC headers, as I was going through waveforms. Zefeng’s code also output the SNR of the windows used for searching for the P waves. I used this SNR data to limit the waveforms in my dataset further so it only includes waveforms with an SNR greater than 5.
After going through manual first arrival or potential FZHW picks in SAC, I found 137 waveforms with potential FZHW’s, out of the 1571 I started with. I ran Zefeng’s automatic picker for FZHW and about 20 events of these were “detected”. However, further analysis is needed to refine the picking method, as not all the automatically detected waveforms look promising.
Last week, I reached a stopping point in my work wherein I prepared data and code to be brought back to work on at Georgia Tech. I then gave a presentation to some researchers and students at the Chinese Academy of Sciences, recapitulating the work I had done while at the institute. Included in the presentation were a couple of maps I have made of the study region, which are attached below, as well as a group picture after presentations;
The blue triangles represent the stations I am using in the YA network. One map is zoomed out to show the topography of the area, with the Tibetan Plateau to the west and the Sichuan Basin to the east.
The other map is a zoomed-in version elaborating on the station positions relative to the Longmenshan Fault Zone, specifically the Pengguan Fault near which the 2013 Ms 7.0 Lushan earthquake occurred.
After we left Beijing, Gillian and I took a three day trip to Guilin, in the South of China. The highlight of the trip was a river boat tour in the heart of the karst landscape of Guilin. The mountains were truly beautiful. A few pictures; The first is of the Sun and Moon Pagodas. And yes, the last one is of a chicken in a bamboo tree..
After living in Beijing for a number of weeks now, I am becoming used to living here. Where to eat, how to cross the street, and how to get boiled water are pretty routine to me. However, there are one or two things that are still different to me over here. The most blatant difference in my environment is when people come up to Gillian and I and ask (or sometimes don’t ask) for a picture with us. It makes me feel like a tourist attraction, so it is not my favorite thing about being here.
Another difference about my environment is all the tea. Since I’ve arrived, I’ve had tea served with almost every meal. I also bought a tea bottle to make tea at work, as boiling water is offered instead of cold water. So far, some varieties of tea I’ve had include; green, black, jasmine, pu-er, and a wheat-based tea of which I do not remember the name. The tea is one thing I really love about being in Beijing. This is why I’m buying a few teas to bring back with me to the United States, lest I cannot find some varieties back home.
As for research, I’ve been modifying code to run templates of p and s picks for each component. Previously, code I was working with only used p-phase templates from vertical seismogram components, and only used s-phase templates from the two horizontal components. This way, I’ll be examining both the p and s-wave arrivals on each waveform, which should be more inclusive. At the end of running cross-correlation detections, I will be able to make a plot matching both the p and s-arrivals of the templates against the detected events. I will provide a plot of this when templates finish running. I was hoping to finish the matched filter detections before leaving Beijing. However, the internet had different plans, and I spent weekends exploring the city. That is alright, though, I will finish the code and run it upon my return to Georgia Tech.
Lastly, I went to some beautiful locations this week. The Great Wall at Jinshanling is one of the most astonishing places I have ever been. We walked a part of the wall for hours, and it just keeps going and going.. Below are some pictures of me at the Wall, as well as the Temple of Heavan, a panda, and the Lama Temple.
This week was packed with activities, so much to do around Beijing!
Wednesday, Gillian and I were invited to go with a group of researchers at the institute for an anti-corruption outing. The day started with us going to an apricot field on the outskirts of Beijing, where we picked apricots. They were so good and juicy! I still have about 2 kilograms on my desk left to snack on.
Then we went for lunch at a restaurant where we had a taste of a typical home-cooked meal of the region. Also very delicious. I loved the anis that was added to some of the dishes, this is at least one spice that I recognized from home.
Our next stop was an anti-corruption museum and tour. This included scenes from Ming Dynasty China and historical accounts of corruption. A different type of museum that I’ve been to, but was interesting.
Last visit of the day, we went to the Ming Tombs, where many of the emperors were buried during the Ming Dynasty.
Some pictures from the area;
Also visited Tiananmen Square and the Forbidden City. Covered a much larger area than I expected. The architecture of the palaces was rather striking.
A few pictures;
Also went to the Beijing Zoo and Aquarium, I just love zoos! Saw a few pandas, which was quite nice.
On the research side of things, to answer the questions in the comments on what the matched filter is and what question I am trying to answer; Essentially, there is a catalog of known events, but no other events are recorded in the sequence. A waveform matching technique is used to search for missing events within a specified time range. The known events are used as templates to scan the continuous data for events with similar waveforms. The goal is to fill in potential missing events in the catalog. This ideally results in a higher magnitude of completeness, which allows for a more in-depth study of the earthquake distributions of the sequence.
This week, I focused on plotting spectrograms of some waveforms to determine an appropriate filter to apply to the data. Then I filtered the data with a band pass filter (BPF) of 2 to 16 Hz, and changed the sampling period of the data to an appropriate value for the highest allowed frequency in the BPF range, 16 Hz. I also started a simple GMT map of the study region.
On the elevator speech assignment (based on [url=http://thepostdocway.com/content/elevator-pitches-scientists-what-when-where-and-how]http://thepostdocway.com/content/elevator-pitches-scientists-what-when-where-and-how[/url]) :
I’m not surprised it took the blogger three hours to come up with their elevator pitch. A good, concise statement is the result of much searching and reflection on the part of the author. As the blogger said, one of the most difficult parts is determining what parts need to be said, then adapting the scientific vocabulary to a more general audience. I’m not entirely satisfied with the pitch I wrote, I’d still need several hours to figure out exactly what I want to say about my research.
Besides the AGU, such a pitch would be helpful at conferences, when first meeting other scientists in my field, and for people in my life who are interested in what I am studying. I’ve observed that when talking others, I often do a poor job of explaining what it is I study in seismology. I’d like people who ask what I do to be interested in seismology and they deserve a clear response from me.
As the second week here is coming to an end, I've been making progress with preparing data for the matched filter process. I'm almost done running the code I wrote to cut waveforms for templates and write the phase picks into the sac headers.
The provenance of the data is Lushan, China. I'm analyzing the continuous waveforms from April 24 to May 19 2013. Events and phase picks have already been determined and written into a catalog by someone else. What I'm doing now is recutting the data so that all templates will have the same length, and longer than the current average length of the cut data.
All in all, it's been a productive week. The only thing slowing me down right now is the internet connection to work off the server, which is good until it decides to stop working, as with much of technology. It has put me behind by a day or two because of when I have to restart running code. My next step will be to confirm the templates were correctly cut . Then I'll move on to plotting spectrograms to decide what filtering to implement on the data.
Also went to a couple lectures since I've been here. A couple were on seismology topics, but the last was on the climate history of Mars, which was fun to learn about. Here's what the office looks like;
And I went to a nearby park for a run the other day, and enjoyed the scenery. The water level was pretty much at zero, but it was a nice setting all the same;
Lastly, as for one of the skills I aim to develop; to explain how your research project will contribute something new to existing knowledge. This goal is particularly important to me because it is why I want to partake in science. I would like to add to current scientific knowledge in my field. If I didn't have a solid reason for working on a specific research project, I would not be doing it. One of my biggest oversights when doing research, I believe, is that I tend to focus more on the details than the big picture. That is, I get lost in the technicalities of coding and can lose sight of the specific goals of why I started the project in the first place. I need to take time to step back, reflect on where I am, and how to alter course so that I stay on track to addressing the main research question in the bigger context.
Nǐ hǎo from Beijing! Made it after an 11 hour flight from Seattle and at least a full day of travel.
It’s been a busy week waiting for luggage to catch up with us and setting up everything from the office to the dorm. Everyone at the Chinese Academy of Science has been very helpful, which is great because my Mandarin is very limited. Trying to learn some new words and phrases while I’m here.
While we’re here, I plan on visiting the Forbidden City and the Great Wall. I can’t wait! Will be posting pictures of where I’m headed.
Aside from the language barrier, what worried me the most coming here was the pollution. This is what the air was like when landing;
However, after a heavy rainstorm, the air cleared up and the sky was blue for the first time this week! It’s certainly different from home, but it looks like a nice city. Looking forward to venturing out and exploring away from the research institute over the weekends.
As for my goals for this summer’s research, in three part;
1’st third: I’d like a thorough review of Linux/SAC/GMT and what I learned last summer as a SCEC intern. I’m a bit out of practice specifically with shell scripting, so I’d like to improve on what I’ve learned already. Would also lie to become more efficient at writing code. I also don’t want to rush. Last summer I ended up doing some things twice over because I made mistakes by being impatient. I’ve learned since then how I need to code. Also, I’d like to be at least half way done on the matched filter technique work.
2’nd third: Complete matched filter technique analysis and be working on relocation/ merging the two datasets. Write abstract for AGU.
3’rd part: Finish relocation and merging datasets. Complete poster. I’d also like to have my results published, if possible. As for more personal goals, I’d like to be talking to graduate students and professors to explore my options for graduate school and to figure out the direction I’d like take.
Getting ready to go to China! Looking forward to working in a different country, despite the distance from home.