The primary object of the the project is improving understanding of subducting Tonga slab geometry by analyzing intermediate depth (70 - 300)km earthquakes. The project will specifically be focused on PS and SP waves converted at the slab surface. Precise location of the slab geometry will allow for better thermal models of the slab surface leading to better understanding intermediate-depth seismogenesis.
Here I am in the final week of the internship, and wow have I come a long way in my abilities! I am proud of the skills I have developed. What a wonderful and profound experience this summer has been. I have been so fortunate to spend time with geophysical PhD students, post-doc workers, data analysts, and professors. I have grown very fond of the group, they are my kind of people, and will undoubtedly be saddened to say good bye in a couple of days. However, I will be reunited with the IGPP people and the other IRIS interns that I miss so much at AGU, and hopefully at other professional meetings in the future.
The summer internship has helped me get a flavor of graduate level research and have strengthened my drive towards graduate school. I have decided to pursue a PhD as I have seen first-hand the amount of time needed to work on and finish a meaningful project. I have begun preliminary searches of grad programs and feel a tremendous level of support from so many involved, Shawn and Donna, the rest of the IGPP staff, advisors back at UCSC, and of course the IRIS staff. After this summer, I feel incredibly more confident in my application.
Last week I did a lot of coding, I split my PS – P vs hypocentral distance plots by station to try to tease apart the diffuse somewhat horizontal PS – P time, which we expect to be horizontal and more linear. We expect the travel time plot to have a zero derivative being that the converted phases should all be taking place at the slab interface same distance below the station. After splitting the plots up by station, we notice a trend of horizontal linear distances less than about 350km and at most stations; at greater distances the trend becomes more diffused and a positive slope. After meeting with Shawn and Donna we are still unsure of the exact cause of the trend but are developing some hypothesis. To hone in further I am developing some subplots for the stations that will show 4 separate plots, travel time vs: hypocentral distance, epicentral distance, depth, and azimuth respectively. The azimuth plot is giving me some trouble now but I hope to figure it out soon.
I also began to work with SAC to develop similar plots but showing the waveforms lined up. I could make the plots that I want in the span of a few days, but have been having trouble writing a script to perform a task to all sac files that will align all stations and rotate to the great circle path, to look at the transverse component of the seismogram. I also need to apply a similar filter to the one I used previously in Antelope which allow the slab converted phases to b observable in the seismograms.
So this week will be continuing to code and trouble shoot the problems addressed above as well as mining as much information about the inversion to come as possible from Shawn. I have procrastinated in one of my goals I set at the beginning of the summer, to meet with professors around IGPP to discuss their research. I have spoken with many graduate students and post-docs and a handful of professors, so I have a list of ones that I’d like to speak with but I need to buckle up at drop in with them.
This week has been much different from what I’ve been doing in the previous 6 weeks here. Shawn has been stuck in Houston due timing with Hurricane Harvey, and he has limited internet access now. I completed my picking task late last week, which had been my primary task since week two. Earlier this week I struggled with Antelope (not the most user-friendly software) to relocate the events where I had picked a slab phase (PS or SP) but didn’t already have the appropriate P or S phase already picked and event located. I went back in with the error list and either removed the picks that I input if the seismogram was noisy or the direct arrival couldn’t be found. In looking back at some of the my early picks, while searching for errors, I realized how much better and more confident I’ve become in filtering and looking at waveforms.
As I’ve said in my previous couple of blog posts, my programing aptitude continues to improve, and could rework my codes in the span of an afternoon to handle the final data set, and output plots with the same general trends as before but with data points from all 4 subsets (rather than the 2 I was working with before).
So, with those tasks completed early this week, and the delay in getting SAC installed on my computers, I’ve taken on learning seismic inversion, which has been difficult. The Inversion is the next and final phase of the project. I don’t have a strong background in linear algebra so progress is coming slowly, and it has been a biggest struggle to stay motivated that I’ve come across thus far. In the last 3 days I think I’ve developed at least an understanding of what the inversion method is, however, the math hasn’t clicked yet. Luckily there are many resources available in the IGPP reading room as well as online to continue to look at. Persevering through being stuck/totally lost/confused on a subject is likely the biggest difference between this and undergraduate work, and a skill I wish to continue to develop.
Life in San Diego has been great these last few weeks. I have made many good friendship and acquaintances that have shown me cool places and experiences. I have gotten back into surfing in the last couple of weeks. I surfed a bunch in my later teens, but ended up hanging up the board after a few scary mishaps in the water and deciding to strongly purse music and push our bands popularity. I have been somewhat apprehensive to get back in to surfing ever since quiting. Surf culture in San Diego is different than in Santa Cruz, and so many of my labmates and SIO friends surf, so I decided to give it another go. My roommate has an extra foam board that she has been letting me borrow, very generous of her. Surfing is kind of like riding a bike, I progressed quickly in the last 3 weeks. I’m apt at navigating the water and making it through the break (not as trivial as it may seem). This week I’ve been popping up on a bunch of waves and trying to get my balance back and working on turns. I’m pleasantly surprised in how quickly I built my skills back. I think all the Yoga and the Cycling helps the balance, flexibility, and fitness that surf requires! I may even brave the cold waters up north when I return to Santa Cruz.
This is where I’ve been going for last 6 weeks. You can see my desk complete with desktop Mac, my personal laptop, fuel, and room for my bike. I get to work in a large comunal office known as ‘the barnyard’ in Munk lab. A times there are about 10 of us working in the room, more than half of which are PhD students that are in separate community offices within the barnyard. Most of the time everyone is working on their own computers listening to headphones, but there are times where we hangout and talk, which is nice. My favorite things about being in Munk lab is the 70s-vibe complete with wood paneling and old photos, and the view at the end of the hall is probably the best in all of Scripps. Atop the Terrace looking down at the pier, La Jolla Shores, and La Jolla Cove, it’s where I go to take a few deep breaths, stretch, appreciate being here, and become present.
Being behind the desk and screen for 8+ hours each day took some getting used to it’s been a long time since having an office job. After I adapted I found ways to become productive, and look forward to coming in each day. I find the stress level in doing research much lower than in academic classes. There have been basically no deadlines set for me at any point thus far, which is vastly different from undergrad work. The lack of deadlines affords the opportunity to get side tracked and off on tangents easily, a point which to keep in mind the overlying goal of the project or the day. Without deadlines I don't know when to quit and just keep working!
One thing that took me by surprise was the very sparse directions given. I had tasks and with no idea where to begin. At times it has taken me a week or more to educate myself enough to get started. At first I would feel over whelmed given a task that I know nothing of. I have since become more comfortable with these feelings, and have learned to not put as much pressure on myself and that I will figure it out. Becoming self-aware in this way has maybe been one of the biggest overarching lessons I’ve learned thus far.
I found myself enjoying coding much more now that I have some familiarity with the language and how it. In the beginning, I would get confused by the instructions that were intended for people without experience now I understand how to interpret my google search results and implement them in my code to serve my purpose. There is a specific language about program that also needs to be learned in addition to the basics of the program
I’ve reached the halfway point in my internship. Today I reflected on the areas that I’ve made progress in with both Shawn and Donna and it was good experience to hear their feedback and to think about the skills I have developed and improved upon.
I am also celebrating my first successful python script. After working through 3 different python scripting tutorials over the last 4 weeks I tackled a problem in my project it took me a few days of tinkering to get it right! I feel a real sense of accomplishment with my script and have found that I can more quickly write scripts that analyze the data in different ways, it’s a good feeling.
Here is a GMT map that best reflects the works that I’ve done thus far. The map depicts the region of the study, Tonga Island Arc, Lau Basin, and Fiji Islands. The orange circles map the epic- central location of the pS picks that I have picked thus far, and the green diamonds are the locations of the sP locations. The Blue Triangles are the locations of the deployed array, along with the counts picked at each location.
An interesting bit that we discovered while examining the above figure is the large number of PS and SP arrivals being detected in the Fijian land-stations. We were initially not expecting to detect these (as they aren't slab converted), but upon further inspection the arrivals are due to conversion taking place at the moho rather than at the slab boundary. The crust near Fiji is ~30km thick vs ~7km in the tonga back-arc, so the moho converted phases travel enough distance to separate themselves from the direct p-wave in the seismograms to point where they're observable.
The above figure still needs a legend, which I'll work on at some point!
Today I submitted my abstract to AGU with the help of my advisors, Shawn and Donna. It was a good feeling to submit my abstract, it felt very official and like I was a part of something. Most of the people in the IGPP (Institute of Geophysics and Planetary Physics) department here at Scripps have been working on or talking about submitting abstracts, it helped me break into conversations with people around the building more than ever before, which was nice. I also got some nice positive feedback from Shawn today and he spent a good deal of time today answering questions about grad-school and how to select potential advisors and schools. I have something thinking to do but once again feel excited and positive about my future.
As promised here is a copy of the final submitted draft of my abstract:
Authors: Travis Alongi, S. Shawn Wei, and Donna Blackman
Title: Refining the Tonga Slab Geometry Using Slab Phases of Seismic Waves
Abstract: Although the Tonga subducting slab geometry has been previously mapped by earthquake distribution, its detailed morphology is poorly constrained. The uncertainties of the slab surface relative to earthquakes can be translated into large errors in predicted temperature of hypocenters that is considered as a chief control of intermediate-depth seismicity. Seismic waves converted at the interface between the slab crust and the overlying mantle wedge can provide additional constraints on the location of the slab surface. A PS phase converted at the slab interface is observable in the horizontal components, whereas an SP converted phase can be seen in the vertical component. In this study, we analyze PS and SP phases in the seismic dataset of the 2009–2010 Ridge2000 Lau Spreading Center project, which consisted of 50 ocean bottom seismographs (OBSs) and 17 island-based seismic stations deployed in Fiji, Tonga, and the Lau Basin for about one year. More than 1,000 PS arrivals from local events were manually picked, predominantly with a 1–3 Hz filter. Next, the PS–P differential travel times will be inverted to determine improved depths of the slab surface relative to the local earthquakes and the receiving stations. The refined slab geometry will allow us to assess the thermal structure and dehydration reactions of the Tonga slab, lending further insight into the mechanisms of intermediate-depth seismicity.
Presentation Type: Poster or Presentation
Session: DI016: Slabs in the Mantle and the Fate of Subducted Material
Also as promised are attached screenshots which are examples of the seismograms I’m picking:
What you're looking at here is 3-D componets of two different stations NMKA (a land station in the Tonga Arc) and B09W (An OBS station from the Lau Ridge 2000 deployment in close proximity to the spreading center). These seismogram contains no filtering. You can see how noisy and chaotic the OBS data are.
I apply a 1-3Hz Filter to smooth things out a bit to make it easier to read:
The P and S arrivals were previously picked by Shawn and other students that have volunteered in the passed. Generally these arrivals are fairly obvious in the land stations, however are more difficult to pickout in the OBS stations because they are atop soft ocean sediments and aren't as well coupled to solid rock. My job for the last 11 days of work has been to pick the pS phases in the horizontal components of the roughly 70 seismograms per event. The pS hypocenter are on the top of the slab surface, and I occasionally find sP which had their origin in the lower plane of the double seismic zone. Above is an obvious and easy to pick phase, but often times I need to mess with different frequency filters to convince myself of the phase arrival. I wasn't completely convinced with the arrival in B09W at 7:35:25:000 in the OBS data when trying different filters so I didn't pick it and moved on.
Some events occur in days where I find no slab converted recorded at any stations, this is probably due to much noise in the data probably stormy weather.
I am about 3/8ths of the way through picking the entire years worth of data, and I feel a sense of accomplishment in that. I think it takes a disciplined person to sit infront of a computer scrolling through squiggly lines for hours on end.
Please chime in with questions on the data or the science!
I am in the thick of writing drafts of an abstract for submission to AGU (American Geophysical Union) conference in December for which the deadline is in just a few days. Shawn, Donna, and I decided this past week that we will have made enough headway in our project to present at AGU, something they were very unsure of only at the outset of the project, a few short weeks ago. Needless to say, I am excited to be able to attend.
I’ve concluded that an abstract is like an elevator speech, but with a targeted audience of your peers as your aim, meaning their baseline understanding of the subject matter is assumed to be college level understanding of science. Anyhow, I’m on my third draft of the abstract, awaiting additional feedback from Shawn, and will have finalized feedback from Donna before submittal deadline of Wednesday at 9pm PST. I will post the finalized draft here later.
So, for the 60 second elevator project I am going to create this one targeted toward a broader audience. I have been asked many times about this summer internship project, mainly by adults, but as I found out have a wide array of backgrounds and it’s best to not assume prior knowledge. As part of our IRIS internship we took part in a science communications webinar, where I was turned on to the idea of asking questions to those interested in the project to develop a better understanding of what they know and what they’re interested to know about. The notion of following a question may seem rudimentary or some, but for me and my upbringing this was a surprisingly fresh approach and takes some of the assumption pressures from me. During the webinar one of my fellow interns proposed that a 60sec speech should actually be more of 60sec conversation, and I really like that approach.
But a skeleton 60s speech is as follows:
Audience’s prior knowledge:
Understanding of the Theory of Plate Tectonics
Knowledge of Earthquakes
No Geology, Physics, Chemistry background (no Jargon)
Wants to know what exactly I am doing
In a subduction zone earthquakes are produced where the two plates are forced into contact with one another, but earthquakes also occur in the region of the subducted plate that is no longer rubbing against the overriding plate. These deeper events known as intermediate depth earthquakes are occurring at depths where the surrounding pressure is greater than should allow for faulting as we see near surface and are still a bit of mystery to scientist. I am using distinct properties of earthquake generated seismic wave and computer software to better locate the plate at depth to better understand the probable causes for the intermediate depth earthquake.
In reflecting on the exercise, I think that each time I attempt to explain my project in different terms to different audiences and even to myself it becomes easier and clearer to myself and probably clearer to the audience. Given the timing of my internship compared to others, it may have been easier for me than others, as I have already written an abstract, and prepared and presented a half hour presentation on my topic. I generally think I am good speaking in laymen’s terms as that is how I think about things, and find it more difficult to go the reverse direction. I do take every opportunity to present and talk about geophysics as possible to become comfortable and fluent in the jargon and lingo of the science.
Today is my 10th day working on my internship project. I’m astonished to know that I’ve already been working at this for two weeks. The first day of my project Shawn asked me to make a GMT map and luckily I have a fair amount of experience with GMT and completed a beautiful map of the Tonga region, including marker locations of historic earthquakes as well as earthquakes marker locations from the data set that I will be using for the project. More importantly I plotted the locations of the seismographs used for the study, including a network of permanent and temporary land stations (placed on the Tonga volcanic arc and Fiji islands to the west) and locations of a temporary array of Ocean Bottom Seismograms, OBS. The OBS were at roughly the same latitude as Tonga but in the back-arc region to the west known as the Lau Basin, more specifically the East Lau Spreading Center ELSC. The temporary network was a joint effort deployed by Woods Hole Oceanographic Institute (WHOI) and Washington University in St. Louis (WUSTL) and was in place for 1 year 2009 – 2010. Deployment of this array was an incredibly expensive endeavor and will likely not be repeated in the future due to the cost.
The next day I was to compile a database from lists that Shawn already had been using in order to make a BREQ_FAST database request through IRIS DMC. I thought I knew how to get started with the request, but then quickly realized the limitations of my coding and programming knowledge in handling data tables. I began scanning python tutorials and have been working at that every day since for atleast 3 hours per day.
When I take a break from programming I read peer review journals pertaining to Tonga subduction zone, intermediate depth seismicity, and thermal structure of subduction zones. Many of the papers were provided to me by Shawn, but I am also doing my own outside research to develop a better understanding of the tectonics of the region. I am to give a half hour presentation to Shawn and Donna this coming Wednesday, so I have about 4 days to synthesis everything I’ve been learning and to prepare a PowerPoint and present their work and study to them. I’ll admit I’m a little nervous to present, but I’m confident that I’ll supersede their expectations.
At the beginning of this (2nd) week Shawn provided me a external hard-drive with the data sets that I’ll be working with, which if I understand correctly are mini seed files obtained from IRIS. Shawn helped me set up my computer with Antelope software and showed me how to view the seismograms with a combination of BASH and GUI interface. The data that I’m working with is the data of the temporary array I explained earlier. I am systematically going through the events in descending order with respect to magnitude. Many of the seismograms have already been picked for P and S arrivals either by Shawn or other students that have worked with him. I need to look at all 3 channels of each seismogram in search of PS arrivals, which show up on the horizontal channels, and are a little more obscure than the P or the S arrival, and I’ve found that they show up the best with a 1-3Hz filter and then confirming with no-filter or 0.5-1.5Hz filter. I am also searching for SP arrivals that are in the vertical component, but I haven’t found many arrivals yet.
Looking at the seismograms is tedious work, but I find that listening to podcasts really helps relieve some of the monotony. The data is difficult to work with due to the high levels of noise in the OBS data, which accounts for ~80% of the data. The OBS are dropped in the ocean and land on various soft sediments which likely shift in the presence of shaking, and move with the currents. Some of events all seismograms are so noisy that I get no picks in looking through all 69 seismograms. Additionally, some stations are either non-operational or some channels may be out, in which case I need to skip over them. I’m finding that I’m having a low chance of finding a PS arrival the deeper the event, which more than half are at around 600km depth!
Antelope isn’t the most user-friendly software to use, so a lot of time is chewed up just repeating the same clicks and commands for each event. In the past 4 days I have looked at 85 events, each with about 30 seismograms, and 3 channels per seismogram. I’ve seen a lot of wiggly lines and I’ve found 208 PS arrivals, and I have about an 8% chance of finding one!
Well, enough for now. I need to prepare my presentation, the clock is ticking!
I arrived in San Diego 4 days ago and have spent the passed couple of days settling in, getting myself set up, and finding where things are in the area. I have knew very little about the area before arriving, so there has been much time spent in google maps and feeding my curiousity with Wikipedia. Scripps Institute of Oceanography and my appartment are on in a community of San Diego called La Jolla which is in the northern part of of the City. La Jolla appears to be a very affluent area, the condos across the street from me are priced at 3.5 million! So far the weather has been incredible 80s durring the day and 60s in the evening, the hummidity is much higher than I'm used to, so these temperatures feel much warmer than they would back home. Shorts, T-shirt, and sandals seem to be appropriate at all times of day, and it's not uncommon to see people walking around town with no shirt, women dressed in bikini, this is the California I think many people imagine.
I have enjoyed getting to know the area by bike, to me the best way to explore a new town or city is to do it by bike. I have meet with some local riders that have shown me the spots to ride and where to go for food, everyone is very eager to help and very friendly. The local riders have warned me about 'zoni's' which I've come to find out means Arizonan who are notorious for being oblivious to cyclist, for some reason i found that to be funny.
I've been to the beach everyday since arriving, the sand is nearly white, due to an abundance of quartz and feldspar (probably from the nearby pennisular province granitics), very little to no organics, and what appear to be minor amounts of mafic minerals (probably hornblende). The cliffs surrounding La Jolla are sandstones interbedded with conglomerate. I also found what look like turbidtes amoung the layering, but I'll have to inspect them futher, and maybe send some pictures to the sedimentology peers. I look forward to further educating myself on the geology of the area, and am awaiting my roadside geology of southern CA book to arrive. The water is a nice temperature to swim in it is refreshing, but not painfully cold like it is in Northern California.
I get started tomorrow and I'm excited to learn and experience the next 10 week here.
I am excited to get started on my summer project at Scripps Institute. It was suggested to me that having research experience would be a big plus on graduate school applications and give a better sense of what graduate research will be like. A major objective for me this summer is to grow and evolve from an undergraduate to a student who is prepared for the rigors and challenges of graduate school. last week at orientation, I learned a lot about the differences and expectations placed on a graduate student versus an undergrad. There will be an adjustment from a student who just needs to take classes at get grades to a student who is much more independent and self-motivated. I have set the following preliminary goals for myself to begin prepping.
1st It has become apparent at just how important computer programming skills are in this discipline of geoscience and is probably my biggest weakness. I plan to work with Matlab and Python within my summer placement and will work through the Matlab examples in Turcotte and Schubert Geodynamics.
2nd In attempt to narrow my focus further and find potential graduate advisors I plan to read at least 3 journal articles per week. I will choose from the top cited journals articles and consider review articles on my favorite topics.
3rd It’s no surprise to me that networking was a topic discussed often during orientation, so my plan is to jump at every opportunity to meet with other students and staff and to probe for topics that peak my interest. I will do my best to come out of my shell, engage, and curb intimidation.
Today is Wednesday and this is my first blog post. I flew out to New Mexico Tech, arriving on Sunday evening. Our days have been book solidly with lectures, seminars, field work, and team building activities. We are collectively learning about geophysical application of seismology, how to use computers to interpret the large data sets, and how the recording devices work. There have been presentations from a variety of people of differing backgrounds currently employed in the geophysics, they have been a wealth of knowledge providing guidance and expertise.
After working together for the past several days, the interns have begun to coalesce and bond. We have spent a lot of time together and the group is very supportive of one another.