Allante Harrison is a student at Fort Valley State University currently completing his research at Georgia Institute of Technology under Dr. Zhigang Peng.
The USArray stations are now deployed in the relatively stable Central and Eastern US (CEUS). Several recent studies have shown that small earthquakes can be identified from the USArray data, and many of them were not listed in the standard USGS catalog. A recent study by Frohlich [2012] has identified many events in Eastern Texas that happen to be associated with injection of waste fluid (hydrofracking). Because they are using standard STA/LTA algorithm, it is possible that some smaller-sized events were not detected. My research project will consist of conducting systematic detections of missing earthquakes in Eastern Texas based on a recently developed matched filter technique. Using the events identified by Frohlich [2012] as templates, I will scan through the continuous waveforms to identify more events. The newly obtained events would help to better understand the seismic hazard in CEUS and the relationship with waste water injection and other human activities in this region.
Please forgive my title (lol), just bringing a little humor to the unfortunate realization that my blog has been pretty much neglected for the second half of the summer. I assure you that my blogging activity is not a reflection of my work; I have a few presentations and a research paper in the works that can attest to that!
Picking up where I've left off, we were able to run the cross correlation on a full month of data. In the process, I've experimented with different filters for data, tested different thresholds for event inclusion, and completely dismantled and re-vamped a map of my study region. As far as result go, for the month of data that I've dealt with, I've generated catalogs with 9 times, 6 times, and 5 times more events than listed in previous catalogs. The number of detections vary with the mean absolute deviation threshold chosen. As you increase the threshold that events must have to be included into catalogs, the accuracy of detecting true earthquake events increases, which in turn, decreases the actual number of events. Enhancing the earthquake record of the region in such a way brings light to a trend of increasing earthquake occurence rate that could not be seen with the initial catalog!
My next step was supposed to be to correlate the sudden trend with instances of hydrofracking in the area... That didn't go well. We ran into difficulties acquiring daily injection data that would have tremendously helped the cause. Given the scope of my research, most injection records were too broad for my use, covering months and years, instead of days in a single month. Just a slight setback. We are currently looking for other ways and sources to access this information from.
I've read and discussed quite a few papers with the seismology group here at Georgia Tech, but by actually conducting research myself, I've gained a whole new level of respect and admiration for the individuals whose time and hard work goes into projects such as these. The interns and grad students I've gotten the chance to work with have been amazing, and my mentor has been great source of knowledge and advice. Though I'm physically leaving Georgia Tech today, I will be sure to keep in contact with the people I've met. While I'm sure that my research will extend into the upcoming semester, I think there's a beach in my nearer (excuse my poor grammar) future!
The past couple of days have been filled with data. I've spent the majority of my time downloading, extracting, filtering, and organizing in preparation for the waveform cross-correlation, which will scan a template event against a period of continuous data in order to find the similarity of waveforms. By running the cross correlation on my first period of continuous data, I got the favorable result, which is a seismograph with a single spike at 1.0. As in statistics, correaltions can run from -1.0 to 1.0, with 1.0 being an identical match "self detection". This was reassurance that the correlation was working as expected. (I'm not sure if I'm able to upload pictures using this editor, but believe me, it looks "pretty".)
During our first period of continuous data, which spanned 10 days, we were able to successfully detect a number of events not listed in our earthquake catalog. (Which is pretty cool!) I am now in the process of determining the magnitude of the new events, and expecting more good news very soon.
Short and sweet.
Last week (6/3 - 6/7) was one filled with mixed emotions. On Monday I was given my first insight into what the focus of my research would be. Honestly, I had no idea what my mentor was describing to me, but after reading a few papers it began to make more sense.
My project goals are:
To use Earthscope USArray data to detect more events in Texas from 2009-2011.
To use the newly found events to study possible interaction with hydrofracturing.
Some accomplishments from last week:
Read related research papers by Frohlich and Horton.
Downloaded continuous station data for a 10-day period from the IRIS DMC (16 total stations).
Downloaded, extracted, and organized data for template events (67 total template events).
Made a Makefile! Correction: I made what I thought was a Makefile.. I actually wrote a bash script, which was still very helpful and a lot quicker!
Began picking P and S wave arrival times in SAC for each template event.
And I guess I learned the difference between a Makefile a bash script if you want to count that lol.
My goals for the summer aren't too convoluted. This is great opportunity for me to get some quality research under my belt. By the end of the summer I hope to become a more efficient and independent researcher, know my true feelings towards conducting research, understand more about seismology and geophysics, and ultimately, experience enough to decide whether I have a passion for a particular field of the geosciences.
I probably just lost all two of my viewers somewhere during my rant lol.
Until next week!
Orientation Week was great!
Over the past week, I've had the opportunity to engage in many special, uncharacteristic activities. Coming from the small town of Blackville, South Carolina, (buried in the low-country somewhere near the Savannah River) it isn't often that I experience the climate and geography that Socorro, New Mexico provides. Visiting the Magdalena Ridge Observatory and the ensuing seven-mile hike down the mountain to Hop Canyon are definitely memories I will never forget. (My knees will especially remember the hike!) I've met a lot of great people, including my fellow interns, knowledgeble discussion leaders, and a very hospitable host in Michael Hubenthal. This was a great opportunity to expand my network. Before reading any more of this or any other subsequent post, I feel it is very important that you read the following message.
DISCLAIMER: I AM A MATHEMATICS MAJOR!!
That was my disclaimer, and your warning! Lol.
The past couple of days have been quite the learning experience. While I've taken a couple of Geology courses at school, it isn't quite a match for actually seeing the geology you're discussing. Coming into this summer I understand that my knowledge of Geophysics is very limited, but the programs and mathematical methods that we were introduced to this week provide avenues for vast growth. Using the Unix, Matlab, and GMT tutorials were especially helpful and I'm sure they willbe revisited many times! One topic that generated the most interest in me over the orientation week was Rick Aster's segment over Linear Algebra and the Geophysical Inverse Theory. It was refreshing to hear after having most of the geologic jargon of the week fly past my ears.
If there's anyone out there who endured my first blog post, thanks for reading!
Stay tuned, there may be hope for this lowly Math major after all!