 It's one o'clock on Tuesday, April 12th, so you must be watching Science at Soast. Soast is the School of Ocean, Earth Science, and Technology at the University of Hawaii at Manoa. And every week we bring in an exciting students project to try and get some idea of the active research which is being done here at the Manoa campus. And today I'm really pleased to introduce Kay Takazawa, who is a graduate student in Earth Science. So welcome, Kay. Really appreciate you coming on the show today. I wonder if we can start with you just saying a little bit about yourself. You're a grad student, but we're talking about infosound today. So why are we interested in sound? Thanks for the introduction, Kay. Yes, so sound we use, or we actually use it as our senses quite often. We hear it all over the place, but there's a lot of unique applications we can use with sound by monitoring. And so I'm mainly on task to look at sound that's in the wild and try to figure out what kind of signatures we can pick up maybe relating to rockets or explosions or anything. So what's your background? How does a graduate student or someone doing a bachelor's degree become interested in sound? You know, are you a physicist? Are you an engineer, a psychologist, or what's your background? Yeah, so my undergraduate, I did, I started as a physics major. I was interested in the physical science and generally, and I was, I think sophomore year towards the middle of it. I realized that I wanted to do more real life applications. And then so I added another major applied math. So I was using physics and math to do some projects. I think I did some hurricane tracking in my undergraduate years. Then I was also working as a audio engineer for my campus. So I was a stage person for a little bit, doing gaffing and so on. Then I worked my way up to doing cameras. Then at the end, I was at the sound board mixing for concerts and so on. Where did you do your undergraduate degree? I was in Wheaton College. It's in Illinois, not the Boston College too. Oh, all right, fascinating. So with that background, if you were a sound manager at a nightclub or you were doing other productions, it's probably a labor of love for you, isn't it? But sound has both the pleasure as well as the research side. Yeah, no, it's great. When I first applied to University of Hawaii, Milton asked me like, you might be interested in what I'm doing. You look at sound just kind of like you, but more on the lower frequencies and we use it to detect a bunch of natural signatures. All right, and that's Milton Gar says your thesis advice. Okay, well sound is something everybody has heard of, no pun intended. But maybe we can go to the first slide so you can explain a little bit more about your kind of research. Sure, yes, so I guess it's probably good to kind of start with what is sound. But yeah, so sound is pretty much there's vibrations in the air or it doesn't just have to be air, but in some kind of medium. I have an example here with a speaker. I think most people are familiar with what a speaker is. Maybe they're not quite sure how it works. So you see those little discs on there, those things vibrate, and then that vibration then moves the air particles around and then that different pressures reaches our eardrum and then that eardrum shakes and then it goes into our inner ear. This little snail looking like structure, that's the cochlea, but that kind of works as an inverse piano or more like maybe a xylophone would make more sense. And so you have the different lengths and different strings that correspond to different frequencies that get moved and then that information goes to our brains and tells us what frequencies those sounds are. Okay, and the diagram which we're looking at there presumably is just one frequency because the rare fraction waves look as if they're all evenly spaced. Yes. But your kind of research involves much more complex literally a spectrum of sound waves. So maybe the second slide will help us better understand what it is we hear or even don't hear, right? Yes. So what I said, frequency as it shows on the very bottom here, it just means like how many times that waves compression and rare fractions happen per second. So when we have the very left, those long periods, when we see at one hertz, that means there's a contraction and then smaller and then larger one second. And then at the higher frequencies, that were like a hundred where more vowel sounds are around or maybe 150 hertz. That means there's 150 basically oscillations or backs and forths that are happening every second. So we generally hear in like the 22, maybe 20K hertz. If you've ever seen the backside of earphones, a lot of times it will write a frequency range on it and that's what you would see like 22, 20K hertz. And that's what we speak in, but you might have heard of like ultrasonic, which maybe like bats using ultrasound to hear when they make really high frequency noise, get that echo back to them and then try to figure out what's around them or maybe medical ultrasound, which is even higher that's used to look what's inside, you know. I suspect people are familiar with the fact that your bats can hear at frequencies that we cannot. And I see from the slide that whales can hear a little bit lower frequency than humans do. So what is infrasound? What wavelengths would you describe as infrasound? So generally infrasound will be 20 hertz and below. So it's like the opposite of the ultrasound. It's not too high that we can't hear. It's that it's too low frequencies that we can't hear. And we say 20 hertz, but in reality, we can hear some infrasound, probably even in the 16 hertz range, if we crank up the volume extremely. It's just that our ears aren't tuned to it. So generally we speak of infrasound as less than 20 hertz. So less than 20 oscillations per second and that's an infrasound range. And at those frequencies, that's where your research interests lie. So am I correct in thinking that all of the applications which we're gonna hear about for the rest of the show happens at frequencies less than roughly 20 hertz? Yes, that is where the main, I guess, energy of the signal is. There are still higher frequency information even in like the explosion sounds or one of the greater natural sources of infrasound that we've all are familiar with is like ocean waves. And we can hear them as well. However, the more energy that the louder sound that we can't hear is actually in the infrasound range. And most of our signals that we look at are most of their energy or the loudness is actually in the infrasound. Okay, and it's a complete mixture of different frequencies. I don't think anything in real life is just a single frequency unless it's been human produced. Yes, yeah, okay. Well, we're talking about infrasound today. So let's take a look and you can explain to us some of the environments in which infrasound might be produced that you can do scientific research on. So let's go to the third slide. I think that will describe. All right, so nice diagram. What are we looking at? Here are some, yeah, sources of infrasound that we might be interested in or familiar with. So one of the, like my advisor, Milton's original research was looking at volcanoes. So he used infrasound from volcano eruptions and also just the based sound of volcanoes to try to determine what are their activity levels at and so on. So that's one way of using infrasound. Another is earthquakes. Most people use seismometers, which measures like the ground motions of ground motions caused by earthquakes. However, like earthquakes or ground motions, sound waves are compression waves. Well, not all ground motions are compression waves, but sound waves are compression waves, which means that earthquakes also produce infrasound. So we can use infrasound as a secondary detector for earthquakes, large rockets, airplanes also cause them. So we might use multiple phones to try to track where rockets are flying or where they're headed towards. We can also use it to detect meteorites entering in the atmosphere when it kind of gets into earth. It creates a large boom sound, which can travel very far. And then explosion storms, all these types of things. One of the benefits of infrasound that I forgot to mention is that it travels very far distances. So kind of similar to how our sunsets are more red because the red lights have a lower wavelength or a lower, or as a larger wavelength and lower frequency, sound waves don't get absorbed as much as it's lower in sound. So then it travels through walls, it travels through long distances, and that helps find and use multiple arrays of microphones to figure out where these infrasound sources are from. Now, would it be correct to sort of say that these different sources of infrasound, we saw medias and rockets and volcanic eruptions? Can we think of it as like a fingerprint? How would you perhaps distinguish between a meteor and a piece of space debris coming into the atmosphere? Can you tell that they are different phenomena? Yes, so in a sense, these large infrasound, especially for sparse, so for something that's just entering and not a continuous sound, but like a brief moment in sound, a lot of the infrasound generated corresponds to how much energy it was carrying. And so for the maybe smaller space debris compared to a maybe larger meteorite, we could probably start seeing a difference in what might be the mass of the entry or was there more burning or was there more of a disruption in the atmosphere? Yeah, I know your advisor, Milton, worked on the breakup of the Space Shuttle Columbia, for example. It is clearly quite a large piece of debris and volcanic eruptions at Kilauea. Those signatures are very different. Yes, yeah, I'd say they're quite different. Also the propagation, so we could also tell from which direction things are coming in. So for like space debris would be from above the sky whereas volcanoes would be from the ground. So by having multiple microphones around the world, we can see where it's coming from and what's likely the source. So it'd be similar to like a seismologist would have different seismometers around the surface of the globe and you can triangulate to where the source of the sound is coming. Yeah, exactly. We'll do the exact same thing. It's just instead of the Earth, we use the atmosphere. Right, right, right. And you've got some great examples. I'm really keen to let the viewers take a look at. So let's move on to slide four and this is quite a timely example. So we're looking here at a volcanic eruption, okay? Just talk us through some of the slide here. Yes, so this was earlier this year, we had the Hanga-Tanga eruption. And so that was a, you might have heard it in the news, I assume. So it was one of the largest eruptions that have happened in the, yeah, it's probably a 100 year event. And are we looking, we're looking at the top of the eruption plume in the left-hand image? Yes. Okay, so this was taken from the space station or a satellite, right? So, and that's pretty big. I understand it went up about 40 kilometers or like 35 months, yeah, okay. And then the diagram on the right-hand side, what's that shown? So that is pressure data from one of the stations, well, one of the infrasound stations or maybe barometer stations more active. So sound, like I said, with the vibration and there's these different pressures, if we're looking at more long-term changes rather than the short-term, all we could look at are these barometers that you might have heard of. You'd see these different pressures like overall, but for a large explosion, these periods or the main frequency of the way is so low that you'd see the major changes in the, even just in a barometer. And so what we see here is three peaks from the first arrival from the, so I guess, you know, first is the sphere. And so we have an explosion happening at one location and then from there sound travels, you know, like a sphere outwards, so then it travels both ways. So we have the closest path from the eruption to this signal that you see the first spike there. And then we have the farther path coming around from the other side then causing another speak. And then that first signal went around the earth one more time and then got back to the signal. So we were seeing the sound wave for this pressure, major pressure differential going around the earth multiple times, which is incredible. And those pressure data, were they recorded in Hawaii? Or they were, so that's part of your group in other words. And that's quite remarkable. You said earlier that these low frequency sound waves can travel great distances, but going around the entire planet three times is quite impressive. So do you know how long that would take? How long for sound to go one way around the planet? I think it was, I think the last slide showed it too, was the 35 hours for it to travel once around. So yeah. Okay, that's still quite remarkable that a single eruption. Let's move on to slide five then, which I think is a smaller scale acoustic event. And again, we've got two slides. So tell us what the left-hand slide is showing. So this was the 2013 Russian Meteor, which might be in the back of people's memories by now, but at the time when people were still excited about the end of the world in 2012 or in the company, you know, it didn't come happen, but in 2013 there was a Russian Meteor and they're like, oh, maybe it's the next year. So it's just a picture of, I think it was a structure that captured the meteorite entering into the Earth's atmosphere. And then to the right, we see on top a more time series of the wave form. You know, having like a large spike and going on and so on. One of the colors, what's purple, what's red, and what's... Yes, so blue is something called like a spectrogram, which shows not just the time progression, but then the frequency content of the wave form. And then these colors show the signal-to-noise ratio or just the loudness compared to the ambient pressure. So the yellow means much louder and then the purples less and then black, meaning... That looks like the voice recognition display on some of these TV detective series, right? Trying to identify, is it the same principle but lower frequency? Yes, that's exactly. We can extend this as long as there's the frequency rate of the captured microphone is high enough, we can go all the way up to a human vocal recognition range. Okay, so the technology would be having a microphone that operates at a fast enough repetition rate to record some of this information. Yeah, for infrared sound, because the frequencies are so low, we can actually get away with recording at very low frequencies. For example, the Nyquist frequency or what you can capture with the microphone sample rate is... No one knows what Nyquist frequency is. Come on, let's mic up, help the viewers. Yeah, I was gonna explain onto it, but if we're capturing a sound at 100 hertz, so the microphone is trying to record making note of what pressure is 100 times every second, we can recreate signals up to half, so about 50 hertz. And so that 50 hertz would be the Nyquist, which is just the half of the capture rate of the microphone sensor. I think in your former career as a disc jockey, would you get involved with the technology needed to make some of those measurements so are you doing that as part of your graduate work now? Yeah, I guess yes, so a lot of the things that I did was like mastering tracks. So you'd record as flat as possible, hopefully some kind of instrument or some music, and then now you want to balance that so it sounds nicer. So there's certain frequencies that might not sound well or there's sort of frequency that was a little bit low to capture, so you want to increase that frequency. So a lot of sound boards that you might have seen with all the knobs or you see DJs touching stuff, those control different frequency bands and how loud those frequencies are. So similarly, I guess with my research currently, we look at all these different infrasound signatures and then try to extract the most useful parts for our research. And again, it's kind of like you've got a fingerprint for each type of sound source, right? So do you have a library of what say, a media sounds like or what a volcanic eruption sounds like or is it more detailed and unique one off? It's both, I would say. There are like the theoretical or like the ideal, like if there was no disruption we're in like a nice controlled environment and then there was a meteorite dropped in there. We'll see a clear meteorite signal. Similarly with explosions, if it's in the nice environment, we'll see this ideal shape of a explosion signature. However, you might know that the atmosphere is a lot of times messy. It might be windy, it might be raining, there's different freshers and so on and that jumbles up the signal quite a bit. And so what we end up with is like a core signature. And if it's quiet enough, we could figure out that that's what it is, but sometimes if it's too loud. Now you mentioned the explosion. So let's move on to slide six. And I guess one of the areas of research at least your advisor, Milton Gases is involved in. It's a really useful application of infosound, right? We've got a global map here. Tell us a bit about what that map is showing. Sure. So there's something called comprehensive nuclear test ban treaty. It's just an agreement with a lot of countries said that, okay, let's stop nuclear testing, at least on surface levels. And so that we could limit the environmental impact and also so that we won't cause a terrible war with nuclear weapons. And so part of the agreement is to make sure that no countries are actually doing nuclear explosions. So in part of that, there is the multiple sensor arrays across the world. And one of the ones is infrasound. So these little red dots that you showed nicely highlighted on the screen, these are where some of the infrasound microphones are located. And Milton, my advisor, is in charge of the one in Hawaii and then also on in Macau. So we look at those as well here. Wait, the viewers may have heard, I mean, this is the sort of thing, presumably that's monitoring potential tests in North Korea. You don't have to go to North Korea because the sound waves that an explosion test would generate as we've seen with the volcanic eruption would go all the way around the planet. So you can actually do quite a lot of monitoring. And again, it's the triangulation part. Yes, so it is the triangulation part. It's also different explosion yields. So the size of explosions cause different signature of the waveform. So we could look at the frequencies of the cause signal and then determine how large of an explosion they were testing. Okay. And I was surprised in slide seven, I'm not sure if this is showing a cartoon or an actual facility. So this is an actual, yeah. This is an actual, actual away, huh? Yeah. I was confused when I first looked at this too. This is actually just one microphone that's hosted in the middle and all these round things around are supposed to dampen the local effect. So not so much wind noise or the surrounding information. Like a lot of the noises would be canceled out by these little air tubes that come into. And as a good scientist, there's no scale bar on this. Is that like an inch across or a mile across? Or what is it? It's a pretty big field. So like, yeah, there's a door to go in, a hatch. So it's a nice, it's a nice size. A little blank dot on the left side is a door. Yeah. Well, we're running out of a bit of time. Okay, so let's move on to slide nine because I think the group of UH Minos developing some innovative things. Yes. We've got an explosion and we've got what looks like an iPhone to me. Is that correct? Yeah, this is, I think it's a Samsung Galaxy S20, I think. But yes, any smartphone, we have an application called RedVox, that's R-E-D-V-O-X which you just converts your cell phone into an infrasound capturing device. We have collaborations with national labs to do some explosion testing. And we send them our phones, like about 10 of them. Have them sit at various locations and then we collect the explosion data on these phones. So I could download this app and provide you guys with some useful information, right? Yes. And we have also a nice website that helps you look at the collected information as well. Yeah, and we provided that website is the boy.edu as here we go. So that's their website. If you need to download the app, the viewer, go to that particular website. So, Kay, where do you see your career going after you've got your PhD? Are you gonna stay in infrasound or are you going back through a rock band? I think I'll say into the research field. My funding comes from the Department of Energy. We have a lot of connections with the national labs and connection with this, especially the non-nuclear proliferation things. I'm pretty interested in keeping people not working on nuclear weapons as a way to keep everything safe. So I think I'll stay in the field, maybe work at a national lab or a private company or academia. Is there any potential that you could stay in Hawaii? It seems you're doing some really beneficial kinds of research. What are the career opportunities here in Hawaii, do you think? Yes, I think, well, Milton as well as being a professor, he has his own startup, which is the red box. So he usually offers his students a position there. So that's also a possibility that I stay here working with my advisor. And if not, there's not just even with the explosions, but with tsunami warning things or with volcanoes. There's a lot of applications for infrasounds that Hawaii is a good place to record. Right, right. Well, I believe you're like three years into your PhD, so writing the thesis right now. So good luck to you. Thank you so much. Yeah, and thank you again for being on the show. Let me just remind the audience, you have been watching Science at Soast. I've been your host, Pete McGinnis-Mark, and my guest today has been Kay Takazawa. Kay, thank you very much for being on the show. Thank you very much. Really interesting. Good luck with the thesis research as well. So thank you everybody for watching and please join us again next week when we'll have hopefully another exciting guest appearing on Science at Soast. Until then, goodbye for now. Thank you so much for watching Think Tech Hawaii. If you like what we do, please like us and click the subscribe button on YouTube and the follow button on Vimeo. You can also follow us on Facebook, Instagram, Twitter, and LinkedIn, and donate to us at thinktecawaii.com. Mahalo.