It’s the Tuesday of week 12… Just kidding, I finished the internship officially last Friday. I am still keeping in touch with my mentor and his student about our project as we approach the academic year. I feel bad about not making the time to blog as I entered the last quarter or so of the internship, but I was working on what have been the most difficult steps in the project yet.
Having found about a dozen or so potential lava bombs signals out of the almost hundred that I sampled, broader automation of signal detection was my next goal. It wasn’t to detect new events in the seismic data by running a detection method through continuous waveforms, it was to verify the events already cataloged in the hydrophone data by running detection methods on these events’ corresponding seismic data.
The SSBW came in very handy here, as it gave us a first look at the short-term average/long-term average (STA/LTA) and cross-correlation/match-filtering methods, which I’d be using on my seismic data. Overall, the SSBW was well worth the added effort this summer (on top of the internship); it wasn’t only directly helpful with introducing me to these methods, but also indirectly in that I’m probably going to use most of the code/methods and skills I’ve picked up somewhere down the line as I move forward in my academic career.
The rest of my time in the internship was spent crafting a filtered STA/LTA template that was robust enough to encompass the diverse variations I saw in my potential signals, with next steps to take in the future to be applying this template to the entire lava bombs catalog using cross-correlation/match-filtering to see about how many lava bombs out of the ones detected by hydrophones could also be detected by seismic stations. This is not only the main goal of the project, to see if lava bombs are detectable by seismic stations, but it would also help to verify the lava bombs catalog and hopefully make its measurements more precise.
The rest of my time overall was spent reflecting on my work this summer. Though the internship wrapped up officially last week, it felt a little lackluster because of the virtual format; I didn’t get to celebrate with my cohort or my mentor, but I was also encouraged by the prospect of continuing this project with Prof Shen and Puja in the future. At the very least, we have AGU to look forward to! I’m glad to have recorded everything on a Google Site (https://sites.google.com/ucsd.edu/iris-internship/home?authuser=1), and I am grateful to everyone around me for an enriching and productive summer.
I want to give a warm thank you to Michael for facilitating and guiding us through the summer term; Em for their support and excellent advice; and the cohort, for being there as much as possible given our severely limited, virtual-only interactive spaces. Provide circumstances allow for it, I really hope to meet everyone in the future in person.
It’s Tuesday of week 9, but I forgot to write up an entry for week 8. Last week, I finalized and submitted my AGU abstract, finalized a list of positive cases that I’m confident in (both seismic and acoustic signals), and prepared to take the next step in the project: automate detection of lava bombs, based on the preliminary/reference catalog. To automate signal analysis will require correlation calculations and usage of the short-term average/long-term average comparison method that Puja used with the hydrophone data, and both were covered by our work in the SSBW!
I found that a lot of things in the SSBW, especially in the last few weeks, have been extremely relevant to my work (spectrograms, automated detection, etc.), and I really wish I knew about the workshop last year because I had to learn so much of what’s covered in the SSBW on my own, from scratch, during the school year when I began research. It’s still really nice to have this high-quality refresher/second look at concepts and methods I already use.
Here is a test spectrogram, the code from which I used to verify my confident cases, which included both potential seismic and acoustic signals. The top row is land data (from the HV PHOD station, if I recall correctly) and the bottom row is reference hydrophone data (from the Z6 KSFK station). Note the color bar as a power scale, and the different arrival times – we theorize that we’re seeing a seismic signal, based on theoretical signal propagation velocities (seismic, by land; acoustic, by air).
Today is Monday of week 8! I finished week 7 just before this weekend. I found that the PHOD station, out of about 40-80 sampled events from the full 4000+, shows the most correlated signals, with both theoretical propagations as sound through the air (slower travel and later arrival times) and as seismic waves (faster travel and earlier arrival times). This week, I’ll be recording exactly which events have what kind of correlation, as well as determining noise levels of the background and of the signals by displaying color bars to accompany my spectrograms to show power (dB/Hz). I’ll be updating my final AGU abstract draft with my findings today, then submitting to a volcanology session along with my mentor’s grad student!
I think the internship program is a great primer to what things might be like outside of traditional undergraduate education, specifically how academic research works and possibly what being a graduate student is like (more specifically, what research in seismology is like). However, everybody in this internship had a different and unique experience and will have different experiences in the future, so I can’t say for sure that my internship accurately represents exactly what I’ll do in the near future.
I think that my project got me more and more excited the longer I spent on it. A lot of the work might have been a little tedious, but because I had a sense of ownership around the project, I still wanted to do my best (in contrast to wanting good grades or other mundane reasons). The fact that I got more engaged with my work rather than less as the summer progressed tells me that I am capable of committing to something if I like it and make the decision by and for myself to work at it.
Before this summer, I wasn’t considering graduate school seriously as an option after I finish my undergraduate degree. Now, I think I’m considering it as a very viable path to take; I found that I really enjoy research, especially in a more independent setting where I am left to solve my own problems as a self-motivated scientist (with appropriate guidance, of course!), and it seems like graduate school is a great way to continue doing research.
This was excellent practice in managing my own stress and burnout. I didn’t give myself any break in between my school year and my summer commitments (since my school years start and end so late, it would have been inconvenient for my mentor anyways), so it was a daily struggle with keeping productive… and sometimes, with staying awake. I ended up shifting my own work schedule forward by two hours (start work at 7; I proved to myself I’m actually a morning person!) and defining a very good work-life balance, something that is new to me because I’m still an undergraduate student.
I found myself wanting to continue work outside my workspace out of fear of not making enough progress – and I guess ideally it should be because I love the work – but I didn’t really work on things too much outside of my scheduled hours, either because I was committed to a balance and/or because I was too tired at the end of the day! Overall, I’m tired and skeptical of counting my life by the hour, but I can see possibilities in my future where I’m enjoying what I do without putting such time constraints/pressure on myself.
Given all this, I wholeheartedly think I’m burnt out, and a break is long overdue. Especially in the communities I came up in, the whole thing about constantly hustling, that nothing is ever enough, is a hard lesson to unlearn. Academia, especially, is far from the best place to find balance in that respect; this is why I never considered grad school seriously before this summer. Now that I’m getting closer to a big transition in my life, I’ll want to reflect more before committing to it.
Last week was week 6, and progress has been slow and steady. I’m still working on essentially the same tasks that I have been since week 4, which is looking directly at waveform data from land seismometers for lava bomb signals. Since I have frequent contact with both Dr. Shen and Puja (at least weekly), I am learning a lot about the lava bombs as a type of sudden event, and am trying my best to balance my time between searching for lava bombs as the primary task in my IRIS internship, and other things that are also happening this summer.
The challenges I’m facing right now might be a little different from the challenges I was facing, say, in weeks 1 and 2 of the internship. The biggest hurdle for me to clear is deadline anxiety, especially as conferences are coming up and abstract submission deadlines are approaching. Still, this is a specific concern and speaks to, as I’ve discussed before, my anxiety about my productivity and making good progress.
I think that a skill that I underestimated was being able to break tasks down into smaller chunks, which ties into giving myself direction and seeking guidance from others. Of course, general competency helps with the project, in the sense that developing tools to use for tasks is useful and such, but I don’t think I could have gotten nearly as far as I am now without being able to step back, take some time, and think about how I want to go about doing my work before actually doing it. This way, when I do dive in, I have a clear direction and I am focused.
Time management has been a big deal for me this summer, because I wanted to define clear work-life boundaries for myself as I tackled my research, both at my home university and at URI virtually. So far, it’s been going quite well in terms of occupying my workspace and having very clear boundaries, at least in time spent working. However, I am almost always in that work mode mentally, and I’m not sure it’s motivated purely by positive emotions relating to the research. Again, going back to that productivity anxiety!
We’re basically at the halfway point of the internship. I have six more weeks to go, and I feel really good about that. I think I’m finally moving on from the “what’s going on” phase to the “let’s get some results” phase of the project, which is both exciting and daunting. I sometimes wonder if I’m looking for something that isn’t there, or if I missed something along the way and have to backtrack, etc.
The figure shown below is a spectrogram taken from one of the closest land seismometers to the Kilauea south flank (southwest ocean entry), in a four minute timeframe centered around a lava bomb origin time. Taking into account the time it takes for waves to propagate from the lava bomb to the seismometer, this spectrogram is promising because it shows several broadband signals, one of which is a few seconds after the center of the time window (which is the lava bomb’s origin time); whether or not this is actually the signal of the referenced lava bomb will take more analysis to determine. There are still imperfections in how I am processing and analyzing my data, and I go more into detail on the Google Site: https://sites.google.com/ucsd.edu/iris-internship/home?authuser=1
The paper that I’m drawing the most information (and even some data) from is a paper that my mentor, Dr. Shen, co-authored and published very recently just earlier this year, titled “An OBS Array to Investigate Offshore Seismicity during the 2018 Kilauea Eruption.” The paper covers the 2018 Kilauea eruption as a premise for deploying a dozen ocean-bottom seismometers (OBS) off the southeast shore of the big island of Hawai’i, some preliminary findings from the OBS data, and the high number of events, both seismic and acoustic, that have been recorded by the OBS network. The last point indicates there is data to observe lava-water interactions, which directly includes my internship project! As such, this paper essentially provides the complete foundation for a lot of Dr. Shen’s current work, his graduate student’s work, and by extension, my work. What we’re doing now is directly based off of the work done to produce this paper.
This entry is a bit late, as I’m actually starting week 5 today! I plan to post blog entries at the end of every week from now on.
I think week 4 was my most anxious so far. I was trying my best to push through certain tasks in my work, but the speed at which I was making progress was still quite slow compared to what I wanted and expected from myself, even though I felt like I was becoming more efficient as I got into a rhythm. After some reflection, I’ve determined a few reasons why my expectations and my actual progress were misaligned.
Why my progress didn’t meet my expectations
Overall, the week got more nerve-wracking as I progressed, but by Friday afternoon, after I wrapped up all my meetings for the week and talked to who I needed to talk to, I felt a bit more relaxed and reassured that I was, in fact, making “enough” progress. This really speaks to a larger insecurity that I’m sure many of my peers can relate to - that I want to excel/don’t want to fail/however one wants to phrase it, so I think I need to continue openly communicating with the people around me and hopefully reframe my expectations such that they are more realistic and rewarding rather than punishing. I look forward to week 5!
Oops, I almost forgot, I’m maintaining a Google Site to keep track of my progress in more detail: https://sites.google.com/ucsd.edu/iris-internship/home?authuser=1
We’re starting week 3 off strong! Monday (yesterday) was my university’s observation of Juneteenth, so I had a three-day weekend to rest before getting back at it today. I feel like I’m becoming more time-efficient every day; I’m putting a lot of effort into planning out my tasks, which surprisingly helps a lot with discussing my work with colleagues, not to mention actually getting the tasks done on time.
I think that my biggest challenge at this point is just knowing what to do for my tasks. I find myself unable to make meaningful, timely progress unless I’ve broken down a task into specific, bite-sized details that I can tackle one at a time over the course of the day. Otherwise, I just sit there staring at the task until I either switch to working on something else or I make some small, almost insignificant step. I think that, while breaking things down into digestible chunks is a good practice, I’ll also have to find ways to “forget” other tasks to help myself focus and finish the one at hand before moving onto the next one.
Something I was really hoping to be able to do was to get to know other people in the natural sciences. While I haven’t been able to interact as much as I’d like with my fellow IRIS interns (realistically, what can I do in a virtual-only setting?), I’ve been so fortunate to get to work on campus at IGPP/SIO where I have met another cohort of undergraduate summer interns, who are also working on geophysics-related projects. I have also met graduate students and faculty who are interested in my work, some of whom I had interacted previously as a student in their virtual classes during the past academic year.
Here is a map I built using the Cartopy package in Python. As I progress, I plan to add more and more to this map, but for now it just lists the locations of the land stations I’m working with. I’ll add the ocean-bottom-seismometer network after I finish analyzing the land data.
My project is located on the east/southeast flank of the Big Island, and listed are some of the easternmost stations in the Hawaiian Volcano Observatory (HV) network. Kilauea’s 2018 eruption saw large lava flows to the east, eventually reaching the ocean along the southeast edge of the island. These ocean entries, where lava benches form and sometimes collapse (hence underwater landslides), are the focus of my project.
I am almost done with week 2. It’s crazy to think that some of my fellow interns are already about halfway through their internships…
Overall, things are going great; I’m settling into a daily work schedule where I am making slow and steady progress. It helps a lot that my mentor at my university has set me up with a workspace in the Institute of Geophysics and Planetary Physics (IGPP) complex at the Scripps Institution of Oceanography, where I am meeting another cohort of summer research fellows in person, as well as graduate students and faculty who are more than happy to discuss the ins and outs of seismology with me. Between my IRIS project and continued in-house research at UCSD, I have my work cut out for me the rest of this summer, but I am feeling more confident each day that I will be able to push through. The more time I spend on my work, the more familiar I am with my tools and the quicker I can finish my tasks. That isn’t to say I’m the most efficient, just that I’m improving!
The beach isn’t too distracting, I don’t think…
Elevator pitch assignment: a reflection
One of the tasks we were given to do this week was to prepare a quick, minute(ish)-long dialogue to explain and pitch our research. I thought this was a very useful assignment to be given, as it taught me several things on different levels. On the surface, it helped me prep for presenting my work on my lava bombs/submarine landslides project to any non-academic/unfamiliar audience, and here is where the AGU comes to mind. However, on a deeper level, it gave me practice in summarizing specific scientific topics so that they are digestible to a wider audience, which is a very useful applicable skill to have not just as a scientist or researcher, but as someone engaged in something that just isn’t widely known.
Overall, it was quite difficult for me to draft my elevator speech. Part of it was because I’m just now transitioning out of the settling in phase, but it was mostly because I’ve never had to present my work formally before to people outside of my immediate circles. Even then, if I had to present for a class, the topics were class related rather than research-related, so drafting a minute-long pitch was a new experience for me. I can’t say it was enjoyable at first, but after I finished my draft, I figured it’s a lot simpler, faster, and easier to revise what I had than to rebuild it from scratch; I’m glad that I’ve taken that first step, actually composing the pitch, as I believe the first step will probably turn out to be the hardest. As AGU approaches, I will revise and refine the speech.
Hi all, my name's Cameron Wang and I'll be working on identifying signals of specific types of sudden events in seismic and acoustic data, specifically lava bombs and submarine landslides during the 2018 Kilauea eruption in Hawai’i using various land and ocean-bottom seismometer datasets.
What do I want to accomplish?
I want to accomplish several things with my placement. Starting from the most specific to the most broad, I want to:
Since I first began engaging with research in seismology just this past academic year, I found I was finally able to put what I’ve learned in my studies to good use. Working on this project under Dr. Shen’s tutelage at URI will help me learn more about not only seismology, but also oceanography, volcanology, and how they all connect.
My experience so far in the natural sciences has been positive and inclusive, as has my limited time spent with the IRIS cohort. I look forward to getting to know my fellow interns as we progress through the summer!
This summer internship will be a great opportunity to see if this research environment is right for me, and if I would possibly like to look into graduate programs in the future as well. As I just finished my major during my third year of undergrad, now is as good a time as any to start thinking seriously about what I want to do in the future.
I can start by discussing with my mentor how the Graduate School of Oceanography at URI operates, as well as looking into other institutes that host graduate programs.
I am starting my project off by working with land-based seismometers on the island of Hawai’i in the Hawaiian Volcano Observatory Network (HVO Network), managed by the USGS with data being accessible through IRIS’ Data Management Center. The HVO data will be the first dataset I access, so that we may begin our investigation with consistent land seismometer data to compare against as a baseline.
We will then follow up with in-house data from a recently deployed network of ocean-bottom seismometers to approach the question of identifying specific event signals in seismic and acoustic data, which leads to interesting implications in recording submarine landslides, and more broadly, seismographic data analysis.
My skill in recognizing and understanding research problems: a reflection
This skill, boiled down to understanding a problem, is so widely applicable to almost any endeavor. For example – writing this blog post itself was an exercise in understanding my project so I could adequately summarize it. In the context of this internship, understanding my research thesis is the key to starting and finishing my project. I think that it will be important for me to not only develop but also maintain my understanding of what my research is trying to answer/accomplish as I continue forward.
As for specifics such as formulating research questions, contributing to an existing framework of knowledge, and figuring out my next step, I will also have my mentor to rely on for guidance and assistance where needed.
To be able to fully recognize and completely understand a research problem would, by traditional metrics, make me an expert in the field; I think it means being a dedicated and passionate researcher. For me, there’s no substitute for consistent time spent on the project, though using my time efficiently will also be an area I plan to improve on. All in all, I’ll have to settle in first and see how things go for a little while before I can identify what exactly the bottlenecks are in my work.
Take care and be safe!