Transcript
Speaker: Looking at Japan as a whole, it seems like over the past hundred years, it it looks like it's averaging around four-ish everywhere, right? Just for people listening on the Shindo. Four is startles most felt by nearly everyone walking wakes up most. Okay, so if you are buying in Japan, you're pretty much looking at experiencing an earthquake in the next hundred years that will at least wake you up.
Speaker: Welcome back everybody to another episode of buying a house in Japan podcast. My name is Joey joined by Taki and a special guest today. Keita, we actually went to university with Keita. The three of us were at a ICU at the same time and then kind of crazy story how we reconnected, but welcome Keita to the podcast.
Speaker: Keita, where are you calling in from today? Yakushima. It's a small island in the south of Japan off the coast of Kagoshima. Keita is a software engineer, specifically geospatial. Does a lot of geospatial work and we've been getting some help with him for various projects of ours. But yeah, Keita, I'll let you take it away.
Speaker: Living in Yakushima. It's a small island, but actually it's not that small. It's about 100 kilometers circumference. um There are a few small settlements around the island.
Speaker: But actually, a reason why I'm working with you guys at Akiyamart is because you guys actually met me on the island. i think that was a ah year ago, right?
Speaker: Take and I just planned a trip to Yakushima just for fun. Take loves the island I'd never been before. And then a mutual friend of ours was like, oh, you guys are on Yakushima. You guys should meet up with Keita. And I'm not sure we ever like met while we were at university. Potentially, the first time we really met up was on this Yakushima trip that we did there. I think it was like kind of destiny. Me and Joey talked about this. Like we needed to find someone who could read Japanese and was familiar with government data and like geospatial map stuff. And was like, we just met you for lunch. I think we had that like flying fish filet at that one restaurant. And we're like, holy shit, Joey. Yakushima is the Princess Mononoke Island for everybody. How did you end up there, Keita?
Speaker: Actually, I was living in Tokyo for about 10 or 12 years and my wife, her family lives here on the island. So we would visit here you know maybe once a year for New Year's and for summer vacation, stuff like that.
Speaker: So during COVID, you know, we would stay in this really small apartment in Tokyo um and we wouldn't even go outside for like a a week or so because, you know, um all of the schools are closed. um All of the nursery schools are closed. There's no reason to go outside. And we were like, why why are we even living here when we would be able to live in a big house in the nature in Yakushima? So that's kind of where motivation came from. And also another big thing was, know, I think maybe three or four years ago, we got fiber internet on the island as well.
Speaker: And so before that, um there's only like ADSL or everyone would be tethering from their cell phones. I believe it's like a temperate rainforest and like we're driving down the highway there and there's like massive monstera plants just kind of growing out of cracks in the side of the road. It it really is magical place.
Speaker: Yeah. Tell me about this avocado farm idea. Let me in on this too. Yeah. So I'm not sure like why I'm interested in avocado plants, but um I've always wanted to do avocado planting.
Speaker: And so I tried, you know, um everyone ah tries, you know, putting an avocado seed. in a cup and you know seeing yeah and so I tried that maybe like 10 times in Tokyo and it never worked it never worked out not even the the roots like first yeah the sprouts it I never like before the sprout came out it was just you know um kind of a rot away but The first time I tried here, it actually worked. It got pretty big. So I was like, you know, maybe I'll try working on getting some actual trees.
Speaker: So I ordered um three ah trees to put on... my you know land around my house. Just three trees, you know probably i won't get that much avocados. So recently I've been kind of passively looking for some land so I could you know plant, I don't know, maybe like 50 or 60 trees, see how that goes.
Speaker: Keita, I want to be part of this avocado project if there's an opportunity. Okay, let's talk about the earthquake filter. So Keita has been working really hard on doing like an earthquake overlay. This is probably our our most requested feature on AkiMart as it's it's you know top of mind with a lot of earthquakes, a lot of natural disasters happening in Japan.
Speaker: We needed an AkiMart overlay earthquake filter and Keita and Joey have delivered it. Tell us a little bit about this filter, Keita. This is based on actual earthquakes that have happened over a span of 100 years. there There is a little more data data. So we're getting data from a couple of government sources. But um there is a little more data. But especially the data that's really old data, it doesn't have like there aren't that many data points. The intensity scale is kind of different and stuff like that. But this data here is um
Speaker: ah for the past 100 years, basically. How many earthquakes in total in the past 100 years have there been? It's ah probably hundreds of thousands. Japan has like a Shindo scale. So for the Shindo one or higher, it's probably 50 100,000 probably, yeah.
Speaker: We had actually discussed this, I think in Tokyo together. Our big problem was we had seen a lot of these seismic risk maps and and maybe some other folks overlaying, but it was missing some of the big earthquakes.
Speaker: And, you know, I think this is a controversial take, you know, this is our take on why we built this, but big thing I wanted to know was like, where have there previously been earthquakes? Um, again, our, our data is limited to about a hundred years, but.
Speaker: yeah basically we we map this on like kate to correct me if i're wrong i like shakes that have been felt with the data we have in japan over the past hundred years right yeah to be a little bit more specific i believe in kate to jump in here if i go off the rails here but we took every earthquake in the past 100 years and filter that down to, I believe it was only earthquakes that registered a magnitude, was it one and above or five and above?
Speaker: Four and above. So take every earthquake in the past 100 years that has at least hit a four somewhere in all the different observations. So there's like these little observation points all over Japan and it sounds like they've been increasing. and And one of the technicalities or limitations on the model, we'll get into it, around the Hanschen earthquake in 94. I think they didn't have so many observation stations and now there's many more. so you have to understand that there were, for older earthquakes, there's less observations and for more recent observation earthquakes, there's more observations. But so we took every earthquake that had at least one observation point
Speaker: where it reached a Shindo of four Shindo is like, I guess, seismic intensity. And then what we did is for each of those earthquakes, we, we put on a map, um, those points, those observation points with that, um, Shindo observation. There's not a complete observation point in every square kilometer in Japan. So we had to interpolate the value of those values in between. then what we ended up with was if there was a hundred thousand earthquakes in those a hundred years, we had a hundred thousand overlays with interpolations of all those Shindo for and above earthquakes. And then, so to, this is, I think we're showing the, the Noto earthquake here, but to make the whole data set, we would take all of those earthquakes and then take the maximum value of each box. So we have a aggregated maximum at this point, this is the maximum it's ever shaken or ever experienced an earthquake. You see, it's broken into cute little squares here or pixels. We'll probably call them in our conversation going forward here, but for every pixel, we looked at all those, you know, hundred, let's just call them a hundred thousand earthquakes in the past hundred years. And then we figured out what's the maximum shaking that's ever happened across all those earthquakes in that pixel, in that little square. And then, so every little square got a number representing the maximum the ground has ever shaken there since recorded history or the past hundred years. And then we just colored it on a scale from, guess, green to kind of gray, yellow to red.
Speaker: So our GSOGI luck spot, maximum recorded intensity 4.1. That's not too bad, right? Not bad. Not bad. Hold on. Now let me take a look at our Tama place. Oh my gosh. Let's know what Noto is. So an actual bad earthquake. think it's going to be like 6.1-ish, I'm guessing.
Speaker: So we have these overlays, if you're looking at the the screen share right now, by the three big earthquakes that have happened probably in the past 30 years or so. So the Hanshin, the Tohoku, and then the Noto.
Speaker: Again, this this was this whole project was sparked because we were looking at some of the seismic risk data that just like showed Noto as a yellow. And there was just like a huge catastrophic earthquake that happened. Yeah. I met some someone at a party.
Speaker: some some earthquakeologist guy at a party and the the guy was like oh you need to look at the seismic risk map that's like the best way to you know assess seismic risk and maybe that's true and maybe that's like the safest scientific way to measure it but i think it doesn't reflect like hey noto had a big earthquake like fukushima had a big earthquake so we wanted a map where you know those places are are clearly in red so take just um changed the view here so before we were on the aggregate map where we took those hundred years of earthquakes and recorded the the max shaking at those points of land this is just one specific earthquakes data set now that we're looking at this is the noto earthquake in 2024 so there's just one of those earthquakes and what was the max shaking across all japan just from that one earthquake and target briefly flipped to the uh what was the other one the fukushima one in 2011. this was the the big tsunami and nuclear power plant meltdown disaster earthquake
Speaker: and then the han chine this was actually a really large earthquake but it doesn't look that bad compared to um noto especially they didn't have as many uh observation points online missing data i see yeah that that makes sense i was actually wondering about that because it was such a big disaster in kobe so like even a 4.1 or like 4.7 can cause ah quite a lot of damage as well as what i was thinking but maybe it is just a lack of data Yeah, that's that's a big a big reason. And also, well, I think another reason was because it was so um like local localized or you know it it was like in the middle of the city. So there's a lot of damage that ah occurred because of that as well. I know a lot of people are going to be like, you guys only have 100 years of data. And yes, we recognize that over a geological scale, 100 years is is a very minimal amount. But like,
Speaker: came to talk us through, is there like, this was all the data that was there, right? Like again, how are we finding this and like, where is it coming from generally? Yeah, so we're getting it from basically the Japanese government. They have this data available, ah but um I think even the maximum, it's like 120 years maybe. And even those first 20 years, most of that data, it's not like digitally created, obviously, because they didn't have computers back then. um So someone would... ah
Speaker: be like observ observing the earthquake and then they would write it down in a notebook or something like that. So especially for the earthquakes like 100 years ago, 120 years ago, it's like, ah where was the epicenter of this? It's like, we don't know.
Speaker: There are, I mean, again, this is only 100 years of data. So it's it's so hard for us to to know. Yeah, I think no to my understanding, like predicting earthquakes is impossible. I think like, right, even though there's supposed to be a big non-guy earthquake,
Speaker: Soon. So we we recognize that our hazard map and earthquake overlay is, is going to be controversial to a lot of folks. But again, this is kind of our. Our methodology, again, we we just have the data we have. We wanted to overlay it to give our customers the max knowledge, you know, hey, I'm buying this house and has shook in the past one hundred years at a 4.7 intensity. I think that's good knowledge. As an Akumar Pro subscriber, if you're not subscribed to Akumar Pro, give it a shot.
Speaker: Yeah, I mean, it's it's one view of earthquake data, and I think there's many different ways to view it. And I'm sure an earthquake scientist probably has ah strong opinions on methodologies and how they think the data should be presented. And we are not earthquakeologists here. We just took a stab at presenting the data the way that we thought would be relevant to homebuyers.
Speaker: Yeah, ultimately, I think we just presented all the data that is available, like from a, like a pure, this is how much this area shook versus like looking at fault lines. I feel like you could go down a rabbit hole. I think maybe we consider adding the seismic rib risk map later on if if enough people have feedback. And if any listeners are big Earthquake people and really opinionated would love to hear your opinion. Shoot us an email at podcast at akia-mart.com. But Keita, any other thoughts on building this? Was this challenging?
Speaker: Is there any any other filters like this? It was really interesting, you know, going back, getting all the data. um And so there was a couple discussions that we had, like how should we ah how should we calculate this data? And so we've been talking about the maximum intensity the whole time. But at the beginning, like we were talking, maybe maybe we would take average or like some kind of a weighted average, like focusing on recency or something like that, I don't know, but we landed on going just for the maximum over 100 years. And obviously, like you you kind of have to do that because you can't put everything into one thing and gets too complicated. So I think or what we were trying to make, this maximum was the the the best way to do it. But like you said, there are definitely different approaches that we can take. And you know maybe we will if there is feedback on this in the future. yeah
Speaker: Okay. So is it inaccurate to say like, if there was a, yeah, so like I'm, I, the way I would interpret this without seeing anything, max recording intensity five, that would not be magnitude. This is Matt Shindo five, which is different. Right. Yeah.
Speaker: Correct. Uh, is it less than a magnitude again? How do I, how do I correlate this in my mind? Like, can I correlate them as like, Hey, was there a magnitude five earthquake here around this area? Or how would you explain that?
Speaker: – magnitude it it depends on like how far away the earthquake was like how deep the earthquake was if you have a magnitude. Mike Wilberg – course earthquake but it's very shallow it's like just one kilometer under exactly where you are, then it would be very intense. Magnitude is like a ah ah metric about the the specifics of the earthquake itself, and Shindo is more about the shaking on the ground at that point. That's another reason why we used the Shindo for the map itself, because we wanted to get ah an an image of like how much actual shaking ah occurred at this place on the Earth. We should definitely make a legend here. So just like if you're looking at the no-till earthquake, max recorded intensity at this edge is 6.1.
Speaker: According Wikipedia, this is, oh my gosh, standing or moving is only possible by crawling. People may be thrown through the air. Low earthquake resistant houses are likely to lean or collapse. Larger cracks are possible. We were looking at a samurai compound in no-till area that definitely looked messed up and had been hit by a Shindo 6 earthquake. So... Looking at Japan as a whole, it seems like over the past hundred years, it it looks like it's averaging around four, like four-ish everywhere, right? Just for people listening on the Shindo. Four is startles most felt by nearly everyone walking wakes up most. Okay, so Shindo four will wake you up. I was looking living in Tokyo for about 10 years. And I think just from my experience living there, we would get maybe on average a Shindo for earthquake once a year or something like that. so Can you go to the map, put in ICU? This is the university we all went to, International Christian University in Western Tokyo.
Speaker: I was there during the... um Tohoku earthquake, and that's probably the biggest earthquake. I'm just going guess for this area. I want to know what what the shaking was. Okay. Okay. Yeah. What was the shaking? 4.5. when you were the legend, I grabbed onto something stable. i remember that. have that memory. So I was like, okay, what were we at?
Speaker: You're on 4.6, I would say. So that's, you're in between a wake up and frightens many. I feel like I have this weird curse where I have all these earthquakes that follow me. I've had like five natural earthquakes and like,
Speaker: The Japan earthquake, i was at the airport. Iceland, had an earthquake. And I was in Nepal with the the Everest earthquake. And that was like, that must have been a six, I think. Rocks were moving up. It like it was it was weird. There was like a liquefaction almost. I was like, uh-oh, we're screwed. But anyways, good good that we have this earthquake map.
Speaker: Hopefully, we all never experienced a seven seven is unrecorded that is not uh something we've experienced yet there's a lot of work that has gone behind this intensity calculation we have just scratched the tip of the iceberg you're a seismologist out there and you see some some large blunders we've made here let us know but it sounds like the tldr here if you are buying in japan you're pretty much looking at experiencing an earthquake in the next hundred years that will at least wake you up
Speaker: go back to noto for a second on the map like it's not to say that lightning will always strike here again or like lightning never strikes twice in the same places like this is the max over the past 100 years had we made done this analysis two years ago before this earthquake it probably wouldn't be as red um as it is uh so you know it's with earthquakes maybe lightning does strike twice in the same spot and maybe it doesn't it's like no one really knows Keita, great work on this. I can't believe we had found the perfect person on our Yakushima trip. It was, ah it seems like it was destiny.
Speaker: Keita, I know you're spinning up your own company. Where can people find more about you or what are you up to nowadays? i yeah I guess you can find me on on Twitter on on X. I'm a sleepy underscore Keta on X. I work on doing maps, working on this kind of data. um We have a bunch of raw data. um I'll process it into some kind of a map that you can use to create value elsewhere. Kotoba Media is the name of his business. So you can search for that or hit him up on X. And we got ah some cool stuff cooking up right now. I think at least for the next year, hey kate I'm excited for that some of the projects we brainstormed over the whiteboard. We are excited also to see you. We will be coming back to Yakushima hopefully soon.
Speaker: will have another Hotel Iwasaki brainstorm session and hopefully we can buy an avocado farm together. Cool, Keita. Thanks for being on the podcast and we look forward to meeting you with soon.
Speaker: Thanks for listening to this week's episode of Buying a House in Japan. If you find the show helpful or at least entertaining, give us a five-star rating. Takes five seconds and really goes a long way. If you or someone you know is looking for a Japanese home, check out Akiyamart.com, our English-friendly site that lets you browse over half a million listings. That's A-K-I-Y-A-M-A-R-T.com to find your dream home in Japan. If you've got questions or feedback, drop us a line, email contact at akiyamart.com. You can also find us on Instagram as well as TikTok, where we post properties, tips on the home buying process, and updates on our own purchases and renovations. Thanks again for listening, and we'll see you next week.
