Alright, now comes the final push before heading back to Connecticut. To be honest, I'm feeling a little overwhelmed this week by everything that needs to get done and everything I need to catch up on. But, I started off the week in a good way: the day after I got back, I came into the lab just so I could submit my abstract for AGU, which felt so so good. It's crazy just seeing my name as the first author of something, and it's nice being able to produce something meaningful out of everything I've done the last few months. Now I'm trying to pick up on the momentum I left with so I can do some cool stuff with my data. So far I have stereonet plots of all the infrasound sources I've been able to determine throughout the five years of data. I can see distinct groups of signals, one of which clearly corresponds with the location of EMRTC (the Energetic Materials Research and Testing Center), which detonates regular explosions and is expected to be the most prominent source of infrasound received at the array.
This is just a screenshot of the three years where the infrasound sensors were running noise-free. Each plotted point represents a "celerity" vector that I've calculated using the arrival time differences amongst the three stations in the array. These incidence vectors show where the sound wave is coming from in relation to the array. In a stereonet plot, the azimuth angle is simply the angle (0 to 360) around the plot where the inclination goes from 0 (on the horizon, along the ground) at the edge of the circle to 90 (directly overhead) in the center. Plotting all three years over each other shows some serious correlation--there is one dominant direction (EMRTC), and a few other locations of common sources.
The distribution of impulsive events is best represented in a histogram. Here you can clearly see where the most dominant sources are in terms of azimuth angle with respect to the array.
With all of this information, I want to go through and make some connections--what do these groups of signals have in common? For the signals not coming from EMRTC--where are they being produced? Are their directions a product of refraction in the stratosphere or some other atmospheric effect? Even further, what do the signals themselves have in common? Jonathan has given me a cross-correlation code to play around with in the coming weeks where I can group signals that are physically similar (where the wiggles have the same shape). If I know the types of signals coming from an unknown source, this could help me narrow down where they're coming from.
Before I left, I started playing around with grouping signals. In R, I have utilized a program that lets me draw a polygon around a group and just view the information (date, time, etc.) of those signals.
I guess you could say I'm playing detective from here on out. Besides the cross-correlation program, I would also like to investigate atmospheric effects, so, with help from Jonathan, Danny, and Kyle Jones at Sandia, I plan on using a ray-tracing program on my data. With this I should be able to more accurately source the signals taking into account the structure of the atmosphere on a given day and my calculated "celerity" vectors. I'm really not sure yet how it's all going to come together, so the next few weeks are going to be all about learning, experimenting, and manipulating lots of code (once I understand the code, that is).
You must be logged in to post a comment.