Pragmatic 31: It's Supposed To Be Raining

30 July, 2014


Since information could travel faster than the weather we’ve strived to predict the weather more accurately into the future. Joel Housman joins John to dig deep into just how hard weather prediction can be and why.

Transcript available
Welcome to Pragmatic. [Music] Pragmatic is a weekly discussion show contemplating the practical application of technology. Exploring the real world trade-offs, we look at how great ideas are transformed into products and services that can change our lives. Nothing is as simple as it seems. This episode is sponsored by ManyTricks, makers of helpful apps for the Mac. Visit for more information about their apps, Butler, Kimo, Leech, Moom, Usher, Desktop Curtain, TimeSync, NameAngler and Witch. If you visit the URL, you can use the code PRAGMATIC25 in the shopping cart to save 25% on any ManyTricks product. This episode is also sponsored by LIFX. Visit for more information and to take advantage of a special discount off their amazing LED smart bulbs exclusively for Pragmatic listeners. We'll talk more about them during the show. I'm your host John Chidjie and I'm joined today by my guest host Joel Hausman. How are you Joel? Hi John, doing well. How are you? I'm doing very, very well. Thank you very much. Thanks for coming on the show. Yeah, sure. Thanks for having me. a real treat. Yeah, it's when I got in touch and sent you through the topic list, you picked a topic that I was a bit concerned if I'd be able to find someone who was keen to cover it, but I'm really glad that you did. But before we dive into that, I just a little quickly want to say a few thank yous. After our last episode, I've just had an overwhelming response to episode 30 about coffee. I was concerned about covering coffee but I've had many dozens of tweets back, particularly just great feedback. So clearly a lot of people have enjoyed it. So thank you everyone for that feedback. Really appreciate it. But I also want to specifically want to thank Dean Johnson for the lovely review on iTunes. It's been a while since I've mentioned iTunes reviews, but just a reminder, if you are enjoying the show, please leave a rating and iTunes really would appreciate it. It does help. So, thank you for that. And without further ado, I guess we should probably start talking about the weather or well, maybe not specifically what the weather is like right now, but more about predicting the weather. Right. So, could you just start by telling me why you chose that topic because I think it's an interesting story. Well, I've always been fascinated by weather and I think it dates back to when I was a child and where I grew up in southwestern Virginia, we were in the base of the Appalachian Mountains. I really, really, really enjoyed snow in the winter because the Virginia Highway Department is so poorly equipped to deal with winter storms it meant that we got out of school. So whenever I would hear, you know, the TV would be on the background and I would hear the weather forecaster mention that there's maybe snow coming, I would be glued to the news to hear the forecast or when we finally got cable TV around the early 90s, mid 90s, I would watch the weather channel back when that station was still a good channel to watch. And would be glued to that channel for days on end prior to any storm coming to determine "Are we going to get any snow?" and "Could we get out of school?" So the fascination was initially with just snowstorms in the winter. And then I started to pay attention to things like hurricanes in the summer and the fall during hurricane season and then tornadoes, although with tornadoes you usually don't get live coverage of them on TV because usually they're done and over with before anyone has a chance to get on the scene of where it's happening and they're a bit more dangerous. But yeah, so it dates back to a childhood fascination and then I almost majored in meteorology at university until I realized how much math was required for a meteorology degree. And I had a brief fascination and dream of being a storm chaser in the Midwest chasing tornadoes until I realized that it's actually really really hard to get a job being a storm chaser because it's pretty much only funded by grants at universities and it's pretty much it's very low-paying and you pretty much have to be like a grad student and you only do it for a few years until you've got your degree and moved on to an actual real job. - It's also kind of dangerous, don't you think? - Yeah, well, that too. And these days, there's actually become a bit of a problem in the Midwest because there are so many storm chasers. And at least for a system that has a lot of warning that they know many days in advance it's gonna come through an area, The storm chasers all descend on that area and clog up the roads and impede emergency responders from getting to the scene. All sorts of things that give storm chasers a really bad name. I'm actually very grateful I didn't choose that path now because I think I would have regretted it. So, do you think that Jean de Bont has something to answer for with the movie Twister? I think maybe, Twister, I remember really enjoying that movie when I was younger and then I went back and watched it about a year ago and having seen it as an adult, I was like, "Wow, this movie really isn't as good as I remember it being." And then you get the kind of really cheesy 90s special effects. Yeah. Yeah. I still have a soft spot for that movie, but there's just so many gaping holes in it. It's like, "Oh, God, really?" Yeah, it's kind of like when you're a technology geek and you watch any sort of movie and they mention technology in it and they get details wrong and you just kind of wince. I imagine Twister is like that for meteorologists. I suspect you're right, yes. Well, cool. Thank you for that background. That's good to hear. I've always been fascinated by weather prediction because, and for me, I guess it's sort of, it started with a barometer for me. I'll get to that in a minute. I'm jumping ahead here. I really think we should probably dive in. So I guess the problem with weather prediction of any kind of accuracy, and I mean, long-term, beyond just what you can see on the horizon, that's gonna require some kind of communication method that travels faster than the weather travels. And the problem with that historically that we really haven't had that until recently. It's around about the 19, sorry the 1830s I should say, when the telegraph became widespread enough it was actually possible to do that. But in any case, early on they relied heavily on the synoptic chart and surface level barometric pressures and everything. And actually the word synoptics derive from a Greek word which is synoptikos which means seen together which is when I looked up what that meant I'm like okay sure I get how that has something to do with weather but anyway never mind. The point is that the surface level barometric pressure was much easier to measure yeah along with temperature and wind speed and direction everything but it really only tells a very very small part of the story because once you go up a few you know well really even 30-40 meters or whatever you know you know everything changes. It's not just this little thin skin at the ground level that's the problem. So, anyway. So, the idea of a barometer when I was a kid fascinated me and that's when I first sort of got interested in it was when I successfully managed to kill the barometer. Because I mean, I wanted to know what was inside because, you know, that's just what you do when you're a kid. So, I pulled the barometer. Oh yeah, you know, I got a screwdriver, an inflated screwdriver and and opened it up and, oh, cool, it's all empty inside. What was that funny hissing noise when I opened up the metal bit? So yeah, these things, I don't know if you've ever seen one, for listeners, if you've ever seen one, it's like, think of it like a, it's like a metal can, but it's relatively thin. And the, well, this particular one I was looking at was circular and it had a series of ridges, circular ridges that were concentric and they got smaller as you got towards the outside. The whole thing was really only about an inch, if that thick, so, you know, at 25 millimeters thick, maybe a bit more than that, maybe 30 millimeters, something like that, a bit over an inch. And it's fully sealed to a set known pressure. And as the barometric pressure changes, the air pressure on the outside changes, then it pushes that actual pressure vessel in and out. And that movement in and out is picked up by a spring, and that spring drives the needle on the dial show you the barometric pressure. So of course if you puncture that pressure vessel whoops you just busted your barometer. So yay this was my first foray into... yeah anyway never mind that. I also grew up... yeah I know I feel terrible about it even now but never mind. Well you have to take things apart to figure out how they work. Exactly it's just my problem was I couldn't put them back together again. Well I think that's a common problem. Yeah it is it is. A clock suffered a similar fate, what can I say? And later computers, although I actually got good at putting computers back together so that wasn't such a bad sad story that one. I used to do that with old hard drives and a hammer. Seriously? Yes. I would have back in the mid to late 90s when the largest hard drives were say 20 or 40 gigabytes, and hard drives had gotten cheap, I would upgrade them frequently and old ones that I had no more use for or let's say you know it died or something like that if it was out of warranty and I couldn't RMA it I would get a you know flathead screwdriver and hammer and prise its case apart. Cool nice nice silver spinning disk inside. Right yep. Oh dear that's yeah that was fun anyway okay there you go that's that's our geek cred for the episode there we go we'll move on. So numerical modeling is pretty much around about I think it was about the 1920s or thereabouts. It was all well and good but it was very well numerically computationally mathematically intensive which is something you alluded to just earlier with your career decisions and it really wasn't all that useful until the 1950s or so where computers started to happen. Initially analog computers, of course, but you know, and digital computers and as their power increased, numerical models could actually start to be more useful. And of course, also in the 50s, they were launching satellites. So, you know, they started to launch satellites to monitor the weather. So all of that collected more data always need more data, you know, more data, more data, because your models are going to run better if you have more data. And, and obviously, then you collect all the data, then it becomes an exponentially bigger problem. So you need a faster and faster computer to compute all this extra data that you're dragging in. I knew that they could look ahead a week or two, but when I actually just brushed up on this before we did the episode, I was surprised that they're saying that some of the best models will give you six to seven days of relatively accurate prediction. And that's the best they can do. That's the best they can do now. And some of these, that's seriously crazy. So some and even then some of the models at least in the east coast of the US where I actually feel qualified to speak about I don't know about how it is and maybe even on the west coast or on other countries but if we have a say winter storm coming in in February and we know it's going to hit sometime between you know 8 p.m. on Wednesday night and you know 9 a.m. on Thursday morning, you'll get model runs that will say as far out as say 14 days that it's going to happen. But you rarely see at least good forecasters actually produce forecasts on their website or on their TV show or whatever it is, their mechanism for forecasting, to say, "Oh, well, there's going to be a snowstorm two weeks from now," because they know that the models Models will deviate enough that they'll look like an idiot if they come out and say there's going to be a blizzard and nothing happens. Now that's not to say that still doesn't happen, but certain models at least have a reputation even for being accurate 48 hours out, but as you get closer to the event they become less accurate. And then certain other models have a reputation for being inaccurate 48 hours out, but when they get in the last 12 hours they become more accurate. So at least for the forecasters in our area where we live, they will take bits and pieces of information from different models depending upon their accuracy and their reputation for being accurate during certain time periods prior to the event happening. And when a model predicts something and it turns out to be accurate, that's the terminology they use to tell that it's called verification. So the model verifies the actual prediction. And the two primary models that we use here are the GFS, which is known as the North American model, which is funded by NOAA. It's paid for by the federal government. And the European model, which I'm a little hazy on how that gets funded, but I think it's money from the European Union and different European countries. And the European Union, the European model actually has a reputation for being a bit more accurate on average than the GFS model but it also is less accurate at predicting certain things like a certain type of severe storm in the winter may be more accurate under the GFS than like a certain type of winter storm from the European model you know. Okay interesting cool well those models I was hoping you talk about those because I don't know as much about them so But I just want to quickly talk a little bit more about inputs to those models just real quickly. So that's really cool and hold those thoughts. One of the things that I always found interesting was getting the data for these models. So you've got ground stations that monitor these things, but you also have, which I think are far more interesting, are weather balloons. And I say far more interesting, I mean as interesting as a balloon could possibly be. know. So these the weather balloons that they launch, they launch them usually twice a day. And what they'll do is they have something on, they have they basically carry a payload and that and that payload has all sensors in there typically, for pressure, temperature and relative humidity, but they have a transponder that also allows their position to be tracked. So you can use them to estimate wind speed direction at different altitudes as well. So and they'll they'll transmit every one or two seconds, the data and everything. And I think the more modern ones have also got data loggers on board and everything. So the thing is that they launched with either hydrogen or helium, and they started about six feet in diameter, which is about 1.8 meters, which is quite reasonable. So average, well, decent sort of height and size. But as they go up in the upper atmosphere and there's less air pressure, of course, the balloon swells because PV equals NRT, which can be rewritten as P1 V1 over T1 equals P2 T2 over V2 T2. Since if you, if you decrease the external amount of pressure, then the volume will increase proportionally. So by the time you get up to the upper atmosphere, way up high, it actually swells up to about 20 feet in diameter, which is huge. I said that's six, about six meters in diameter, at which point, then the actual balloon material will pop and then the payload will drop down like a, like a rock pretty much. And it has a, a balloon, not a balloon, it has a parachute mechanism, which is triggered by pressure. So when it reaches a certain altitude, it pops parachute and then it glides back down and then they recover it and then they refit it ready for the next release. So that was something that I always found kind of interesting the way that they did that. When I was younger, I used to think that you just set these balloons up. I used to think, gee, that's really wasteful. Yeah, 'cause they're just throwing away all these balloons. Right, where they're just gonna float away. Anyway, so there you go. And they'll feed that data into those models that you were talking about. And that's key information because you need that information from higher up because you need to know what the higher level air masses are doing. And two other sources of data, especially when you're dealing with, say, cyclones or typhoons or hurricanes, depending on what you call them and where you live, are buoy data off the coast of a lot of countries, and even some of them way out in the middle, the Atlantic or the Pacific, They'll, you know, of course anchor a buoy off the ocean floor and they'll put a little weather station on the buoy. And then at least for, I'm not sure how it's done with cyclones and typhoons, but at least for hurricanes in the Atlantic, NOAA also funds something called the Hurricane Hunters that are, I think they are usually flown by military personnel, but like operating under the Elvis of NOAA and they have these specially equipped planes that are filled with radar and all sorts of weather sensors and that sort of thing. And they will take off and make regularly scheduled flights into the hurricane at various altitudes. I mean usually they're pretty high up but they basically will fly like a crisscross zigzag formation or zigzag flight route diagonally across the hurricane from one side of the eye to the other through the middle of the eye and then they'll turn and do another pass. So basically they want to hit all four sides of the eye wall and get different readings at different altitudes. And as they're flying through they drop little sensors from the plane that fall down through the hurricane and measure the pressure and the wind speed and all that sort of thing at various heights within the storm. And those are really important, especially for hurricanes, because otherwise all of our sensors are just at ground level and you can't exactly float a weather balloon up through a hurricane and have it go straight up in the air. No, that's not going to work out, is it? No. But yeah, so all I've been especially with with with storms coming in off the coast Though and I think I think even some some ships like cargo ships Tankers and all will have weather stations on board that also feed into the models Yeah, absolutely. I mean you got all these ships out in the ocean. It makes sense to be collecting data from them I wouldn't be surprised if that yeah, I'm pretty sure I don't think all of them by any means but some of them if they're equipped with modern enough weather equipment that's able to radio its data back. And I'm not sure how that works, whether or not the owners of the particular models pay them for the data, or it's just like a volunteered or whatnot. Although I think it's in the best interest of the ships to volunteer the information, it's just going to help them too. Yeah, I'd suspect that would be the case. but yeah, it's more data, better models. And that's what it comes back to. - Right. - 'Cause the model is only as good. - And the more data, the bigger your computer has to be. - Yeah, that's it. That's it. Okay, cool. So before we go any further, I just wanna talk about our first sponsor just really quickly, and that's ManyTricks. So ManyTricks, a great software development company whose apps do, well, you guessed it, ManyTricks. Their apps include Butler, Kimo, Leech, Moom, Usher, Desktop, Curtain, TimeSync, NameMangler and Witch. There's so much to talk about for each app. We're just going to focus on a different one each week. And this week we're going to talk about Witch, spelt W-I-T-C-H. So think about Witch as a supercharger for your command tab app switcher on your Mac. Without Witch, you just get whichever window is on top for a given app. And if you're like me and you've got three or four documents open at once in pages or in Word, let's say, you know, then it takes another step after that to grab the specific document that you want, with which you get a beautifully simple pop-up that lists every window that's open. Across spaces, displays, you get to pick exactly the one that you're looking for. There's even a neat quick look preview of the windows content, so you can be sure it's the window you're looking for before you select it. And that alone makes it worth grabbing, but you can tweak the time delay before the switcher actually opens. You can show or hide specific apps or windows that you do or you don't want included, as well as you can customize the zoom, close, minimize behaviors for when you do select specific windows. So I've had a play with Witch this week, and I can see why people rave about it. If you haven't tried it, download the trial, have a play with Witch yourself. You can download a free trial of it from W-I-T-C-H, and you can try it out for yourself. If you do fall in love with it, then you can buy it from that page or you can buy it through the Mac App Store for $14 US. However, if you visit that URL before the 17th of August, and you can take advantage of a special discount off their very helpful apps exclusively for Pragmatic listeners. So simply use Pragmatic 25 in the discount code box in the shopping cart to receive 25% off. This offer is only available to Pragmatic listeners for a limited time. So take advantage of it while you can. Thank you to ManyTricks for sponsoring Pragmatic. So one of the other things, Joel, that I find just blows my mind about this whole weather prediction thing is when you consider all of the variables. So I got to thinking, okay, the sun is shining, well, presumably. - Right. - Let's say you've got cloud cover that just comes over for whatever reason, multitude of reasons that there's cloud cover. So that's going to reduce the amount of sunlight that heats the ground up in a localized area. That's going to reduce the surface level temperature. That's going to change your evaporation rate. It's going to affect, therefore, the local air pressure in a localized area. And that's going to affect the pressure gradients between low and high pressure zones around that area. That will then change the wind speed around that area as the high pressure air flows in the lower pressure zones. And suddenly this whole thing just got so damn complicated. It's like, now I understand just how hard this is. All from some heavy cloud cover, right? So what you're alluding to is something that I think many people have heard of this phrase, but it's referred to as, the slang term is the "butterfly effect", which deals with chaos theory. I think the movie Jurassic Park exposed more people to the concept of the chaos theory in the general public than any other pop culture reference. I think you're right about that, yes. The whole little... the drop of water on the back. Right, exactly. Like when he's chatting up the next future Mrs. Malcolm. Right, exactly. Yeah. Yeah. That was a wonderful scene. What's the name of that actor? He's so good. Jeff Goldblum. Yeah, Gold... and he's always... he plays the same character in every movie he's in. Yeah, that was good. But yes, exactly. Yeah, Chaos Theory, Butterfly Effect. That's the one. So this this was pioneered by an American mathematician named Edward Lorenz Who was a mathematician and a meteorologist? This was back in the 60s, I think 1961 Lorenz was using a computer model to run a weather prediction basically and the the story goes that and I want to be accurate when I say this, that as a shortcut on a number in the sequence he entered the decimal .506 by mistake instead of entering .506127. You know, he left off three decimal places. And the result of this prediction was a completely different weather scenario. And he noticed that the prediction didn't really do what he thought it was going to do and then he went back and checked his data and realized that he had left off three decimals and when he ran it again with the three decimals a completely different scenario came out and and based on this observation he wrote a he published a theoretical study that that was well known in at the time that was called deterministic non-periodic flow which basically is the scientific way of saying the butterfly effect or the butterfly theory. And he would then present this paper at some sort of mathematical or meteorological conference and in summing up his talk on the brochure or however it was published, another of his colleagues suggested saying that if the theory were correct, one flap of a seagull's wings could change the course of weather forever. And then Lorenz kind of borrowed this idea of this phrase and changed it around to be more elegant and basically said, "Does the flap of a butterfly's wings in Brazil set off a tornado in Texas?" And thus, the concept of the butterfly effect was born. Yeah, exactly. fascinating how that came about but you're absolutely right and there's so many interactions and so many things to model that I suppose theoretically, and this is the, you know, if you could monitor every single air molecule in the world and mathematically simulate all of them and have the big enough supercomputer to compute all of that data, that's right, then it should be possible to predict with complete accuracy. However, that is incredibly unrealistic when you think about it. So it's the sort of thing that may never happen or if it does happen it's gonna be a long, long, long way off. So one of the things that they do is they run what essentially it's Monte Carlo simulations which I'm familiar with through engineering but something that I believe is referred to as ensemble forecasting. Right yeah so this is this is pretty much the sort of the peak of modern weather forecasting ensemble yes yeah yeah that's it so you put multiple model predictions and you get a distribution of possible results and you pick the essentially the most statistically likely possibility yeah yeah you might throw out if you get if you if you let's say let's say there's gonna be you know a storm coming across the Midwest and you run it against European model and the GSF model and the Canadian model and the the North American model and the the British model or the UK model the UK met and and let's say oh and then there's there's a couple I'm leaving out there's the mid-range model and the SRF the short-range forecast model let's say two of those models have these way crazy outliers that don't say that aren't even close to what all the other models are doing so so most modern forecasters will just kind of throw out those two outliers and then average the others together. And then whatever the average is, is what you read in the paper that it's going to do tomorrow or not on the paper. I'm dating myself here. You know, on your favorite weather website or, or, or whether, yeah, right. Absolutely right. Yep. Couldn't put it better. Exactly. So one of the things that you mentioned just there briefly was the short-term forecast. And one of the things that I found interesting just about the from the app, in the app point of app world, because often you want to, I mean, okay, this, I guess there's two real scenarios. There's I'm sitting at home, I'm about to go out, or I'm in the office, and I'm about to go out. And I want to know what the weather's doing. Apart from the obvious, look out the window. You know, you could use an app on your desktop, however, or on your PC. However, the other use case is that you're out and about, you're not next to a computer, you've got your phone with you, a smartphone, and it's running on Android, Windows Phone, iOS, whatever it's running, and you want to check what the weather is. No one looks at the paper anymore. That's not true, they do, but so many more people, I think, are turning to apps. And that's one of the things that in the last few years, there's been a massive explosion of different apps that are available for looking at the the weather and some of them are just a window they simply report on data supplied by different meteorological you know divisions departments companies right and and those those forecasts are put out at set intervals during the day and and some of the apps may only pull their data you know based on whatever the last you know forecast was which may be six hours old ten hours old you know Yeah, exactly. And yet there are other apps that actually have a dedicated service where they actually do their own interpretation They do their own ensemble forecasting. Some of them just look at the short term models And that was one of the ones I first sort of came across was one called Dark Sky And Dark Sky, I think, I'm not sure if it's still popular or not, but it's not available in Australia But I've heard people rave about it I think it is Dark Sky? I it dark sky is is for I still love it. I use it daily. I've seen I see it recommended all the time on social media. I think it's still very popular. Absolutely. I'm right. Yeah, you're right. And I think it is very popular, but it's still not available in Australia, which and I sort of looked at this, and it's only currently available in the US, Puerto Rico and the British Isles. And there's all sorts of reasons for that. And it comes back to the data and the quality of the short term forecast models. And it's interesting because if you have a look into it, a choice quote from one of the, I think it was one of the founders or whatever of the company that makes Dark Sky. And the quote, it goes like this, "Dark Sky doesn't make any effort to identify "and compensate for chaotic storm behavior." So its primary point is to tell you what the rain's gonna do in the next very short period of time, five, 10 minutes, half an hour, hours, something like that. It's not interested in the long-term forecast. It's not interested in what it's, you know, like it says, it's not trying to compensate. It's doesn't have a complex model. It just looks at a very short term forecast. And the beauty of that is that actually short term weather prediction is kind of, is relatively straightforward, provided you have enough localized weather information. - The reason, yeah, and I think the reason that is an emphasized short range prediction, you know, like in most of your traditional avenues for getting weather predictions is that Most people don't care enough about the weather to constantly check their phone every 30 minutes all day long. And if you did that, you would know what was going to happen. Most people only care enough to look at it in the morning before they had to work to know whether they need to bring a coat with them or bring an umbrella. Yeah, absolutely. Now, when we were talking just before the show, you also mentioned another app called RadarScope. Can you tell me a bit about Yeah, so RadarScope, if you look for at least a radar app, an application on any device, Android, Windows computers, Macs, iOS devices, if you want an app where you can get nice views of a radar loop for your area and you want to know what the professionals use, you should get RadarScope. And don't let me scare you by saying that. It's very, I forget how much they charge for it, but it's very inexpensive on the App Store. I'm checking now. Like on the, it's a universal app for iOS, so it's $10 US and you would have both the iPhone and iPad version. The Mac version is $30 US. But when you have the app, you get access to all of the different weather forecasters refer to as the products. So when you're like, I'm going to launch RadarScope right now, in RadarScope you get access to, for instance, the Super Res Reflectivity Tilt 1, which is usually what you, it's your average view of the radar, Doppler radar in your area, you get things like the vertically integrated liquid. So basically you're looking at liquid in the atmosphere 20, 30,000 feet up. I don't know if you've ever been driving along a road and you've looked in the distance and see rain falling from a cloud and it looks like it's falling to the sky but it's not quite to the ground yet. That product would tell you, "Oh, there's rain above you. It's coming down it just hasn't gotten to you yet, you know, because it's 20,000 feet up and falling from a big thundercloud. RadarScope is wonderful in that you can adjust its settings a little bit and say, "Well, show me 6 frames of data or 20 frames from data," and hit loop and it kind of loops like an animated GIF. So you can, and especially with the iPhone and iPad version, you can use the location services on your phone to have it pinpoint exactly on the map where you're at and then very easily tell based on the loop, "Oh, well that storm's going to come, that storm that if you turn on the TV and listen to the weather guy, he's like, "Oh, there's a storm coming in the area." You can look on the apps, like, "It's going to completely miss me," based on the speed it's going and where it's going or, "Wait a minute, this one's going to go right over top of us." And Dark Sky basically gives you a little snapshot of, and I love Dark Sky because I own dogs and I have to walk them outside quite a bit. So Dark Sky was great for when I lived closer into the city and couldn't really get a good view of the horizon to know what's coming. That I could say, "Oh, can I take the dogs out right now and not get wet?" And it would tell you, say, "The rain is supposed to start in 20 minutes." I'd say, "Great, I've got 10 minutes to get them out, walk them, and get them back inside before it's gonna begin raining. Radar scope will give you a little bit bigger view because of course you can zoom out and see what storms are doing five six hundred miles away from you but based on how fast they're moving and you get little time stamps on each frame you can kind of roughly say oh well this big front that's 500 miles west of us you know right now it's it's 830 a.m. here oh this one should be to us by four or five o'clock this evening so if I'm going to mow the yard today I need to do it early afternoon, you know, before the rain gets here. Radar, I love RadarScope. It kind of allows you to become your own, you know, amateur weather forecaster and then the more you look at the data and the more you watch storms pass overhead you'll just get better at it. Cool. Well from my perspective in Australia one of the things that we sort of have rolled our eyes at for years is the is Yahoo weather because Right. Yeah, because the iPhone comes with Yahoo weather and it's generally accepted that the Yahoo weather is not really up to snuff compared to The main local source in Australia is the Australian Bureau of Meteorology or the right everyone calls them BOM for short So yeah, they're not they're not the BOM, but you know, anyway, never mind that. Jeez. That is such a bad teenage joke Anyway, point is point point is oh god. I I see I'm just I'm trying to be cool like they are right you know anyway no because if you use that sort of teenage slang that just instantly makes you cool anyway sorry okay so over the years though that's improved now the Yahoo data in Australia comes from a company called weather zone and what they do is I actually draw their data from a multitude of organizations in the Asia Pacific region one of them is the Bureau of meteorology but they have their own meteorologists that interpret and model the data themselves. So that's all well and good but even so I don't use the native weather app on my phone because I find it to be the Yahoo data even though it's coming from WeatherZone and everything is still not as good or is not as accurate as the Bureau of Meteorology. So I actually use Pocket Weather which is by ShiftyJelly and they what they do is they poll you know this is just from their blog they poll various FTP and HTTP sites from the Bureau of Meteorology at various intervals during the day. They pass all of that data and that's what is the data, that's the data that they use in Pocket Weather. So everything they do is purely just a window into the Bureau of Meteorology because the BOM website is really not that great. Yeah, the same here in the US, the National Weather Service which is run by NOAA, the National Oceanic and Atmosphere Administration. Their website looks straight out of 1994. - Yeah, yeah. - But to their credit, all of the data that they have available, they give to you on the website. You can get at everything because it's taxpayer funded. But some of the products they put out look really, really horrible as far as the quality of the map and what the colors are and everything. But they give you everything. They present you all of the data that they themselves are using to predict. Yeah. Yeah, absolutely. And the problem that I've got, though, in Australia is that it's great that that data is up there. It's all there. Just like you're saying for National Weather Service, same here in Australia with the Bureau of Meteorology, all the data is there. But, you know, Pog, the software here can display that information, but we don't have dark sky in Australia. We don't have some of the short term modeling stuff. and you know still the models aren't that great and it's mainly because there's a massive area Australia is a big country in terms of surface area it's really not that much smaller than the United States and the population is a fraction of that it's like a tenth and most of your population is on the coast that's it exactly so all of the weather stations or the vast majority of the weather stations Basically, we can't afford, well, the government is not going to put money into putting weather stations absolutely everywhere all over the countryside because they are inaccessible, because they're nowhere near populated centers, which makes them expensive to maintain, difficult to install. And they would be hard to get out to, to prepare. Absolutely. So, it's all those reasons, all the same sorts of reasons why everything like telecommunications is more expensive here. All those reasons apply to weather forecasting. So I do feel it's unfortunate, but I don't think we'll ever get apps like Dark Sky in Australia. I don't think we'll ever get that. I'm curious, which way does your jet stream flow? Does your weather pretty much come west to east or is it east to west? West to east. So most of the weather is coming off of the Indian Ocean on to the continent, crossing the continent before it gets to the population? Generally, yes. Generally, yes. The fronts tend to move up also from typically from south to north. Okay. That would blow my mind because most of our fronts come in from the west. Sometimes they'll come from the southwest or northwest, but it's usually a westerly direction. We have rare instances where there'll be weather coming from the east or something, usually that's related to a storm coming off the Atlantic, but whenever it happens and you look at the radar and it's going the other direction, you're like, "Whoa, that's weird." Yeah, there is an effect at the equator where there is a swirling effect and we're not close enough to the equator to get too much of that. But in any case, it is interesting. Before we get into, because I want to start to admit it's some extreme weather, which we sort of touched on briefly before, but I want to go a little depth about that. Before we do talk about that, I just want to talk about our second sponsor for the episode, which is LIFX, spelled L-I-F-X. It's a smart light bulb that gives you previously unheard of control of your lighting. Each light bulb is Wi-Fi enabled and can give you light in whatever color of the rainbow you like. It's an energy efficient LED light bulb that you can control with an app on your smartphone. It's got over a thousand lumens available and it's incredibly bright, only consumes 18 Watts of power. That's at maximum, though most rooms will only use about half that. So controlling the brightness, color, and there's a range of cool effects. It's very easy on a smartphone, on your smartphone with the LIFX app. And it's made to, and the light bulb itself is made to last. They're rated for about 27 years of operation and that's at four hours a day, which is equivalent to 40,000 hours. So you're gonna be moving house before you're gonna be changing the light bulb basically. The LIFX bulbs ports both standard Edison screw and bayonet connectors and it'll work at all standard voltages around the world between 100 and 240 volts AC. It has a developer friendly SDK currently available for iOS, Android, and Ruby, which means that if you can think of a great way to control them, you can go out and build it on whatever platform you like right now. In fact, there's currently a competition open to developers until the 25th of July this year. So if you submit an app by that deadline, you'll be in the running to win enough LIFX bulbs to fill every light socket in your house, as well as get free advertising for your app through LIFX. Check out for more information and be quick 'cause that's closing really soon. Now I've been testing some demo bulbs and my kids took control and had a little bit of disco in the lounge area. They had a great time actually. It was a lot of fun. And my wife's 40th birthday party is coming up a few days and we're going to be doing something rather similar then. So they are a lot of fun and I really do enjoy them. So LIFX bulbs are shipping today for only $29 US with free shipping worldwide. Simply head over to LIFX that's spelled to learn more and enter the coupon code Pragmatic for 15% off the total price of your order. Thank you to LIFX for sponsoring the show. So, before we move on to extreme weather, I just wanted to touch a little bit on the computing power of models and just what goes into, you know, we talk about you're getting all of these different data points. One of the problems, of course, is the more data you need, the more computing power you need. And I had came across a, doing a little research prior to the show, I came across an article that is about a year old and I couldn't find any more recent information on this. So take this with a grain of salt. Things might have changed since this was put out. This was as of August 2013. But it basically said that the GFS model, which stands for Global Forecast System, which is the official model that the US government funds as a part of NOAA, is currently run on a supercomputer out of Western Virginia which is nicknamed Tide and it has 213 teraflops. And looking at the, there's a great little website called which keeps track of the top 500 supercomputers in the world. Based on that teraflop ranking, that would rank it somewhere around number 180 to 200 on the supercomputer list. And this article also stated that they were planning to upgrade this computer in 2015 from 213 teraflops up to 1950 teraflops. Which based on that supercomputer list would bump it somewhere into, you know, from the upper hundreds and 200 range down to, hold on a second I want to get this an accurate number. Oh, it would be above 100. That would put it in the top 30 of supercomputers in the world. Well, actually, based on Teraflops, that would put it somewhere, and this is as of June 2014, the list I'm looking at, this would put it at ranking number 16 to 17 for the top supercomputer. Now, granted, by the time 2015 rolls around, that list is going to change a lot. And they said that when they did the previous upgrade, which took the computer from 90 teraflops to 213, they saw an average of a 15% accuracy bump in the GFS model. And I know at least for winter weather forecasting, a lot of the winter amateur meteorologists that I, there's this great weather forum that I use in the winter to track these winter storms, they prefer the European model over the GFS because it's a bit more accurate. Although everyone is hopeful for this 2015 upgrade that maybe our accuracy will bump enough that we're on par with the European model for a change. Now when I say it's more accurate than the European model, it's like a 5-10% accuracy difference. It's not too far behind, but it's enough behind that if the GFS is sometimes saying that we're we're going to maybe get 15 inches of snow and the Europeans saying we're going to get 9 inches of snow, chances are it's going to be more like 9 to 10 inches rather than 14 to 15, you know. Yeah, I understand. Well, I mean, it's just that's the wonderful thing is that I know that some people wonder, what on earth do you use a supercomputer for? Well, guess what? This is what you use them for. This is actually practically useful, you know, which is one of the reasons I want to talk about on the show. So, well, thank you for listening to stuff because I was hoping you would. I have no idea about any of the other supercomputers for other systems I'm sure that the information is out on the web somewhere if you did a few Google searches but I just found that interesting because I knew it was a big computer I just didn't know exactly the numbers on it. Sure now I'm glad we talked about that because yeah there is a lot of grunt mathematical computing power just raw power that goes into predicting the weather these days and people still complain about it but anyway. Okay so extreme weather and I guess specifically when I say extreme weather I'm not talking about small localized storms I'm talking about the two big ones which are tornadoes quite quite obviously and of course they're they're bigger brothers the cyclone / hurricane / / / typhoon which are basically just big bad tornadoes. They don't spin as, they don't blow as fast generally in the middle, but they have such a huge area that that sort of brings its own set of problems. So the thing is, I just want to just quickly touch on the whole naming conventions, because you'll have people say, "Oh yeah, Cyclone's different from a hurricane and that's different from a typhoon." Well, no, they're not. They're just different names for the same things. Different parts of the world call them different things. And the The funny thing is also that, and there's a link in the show notes too, there's a good Wikipedia article that goes through all the differences. There's a table that shows all the different wind speeds between which, and the wind speeds are usually rated at one minute sustained, 10 minutes sustained, and peak gusts. You look at this table and it'll show you, well, between this range and this range of speeds, then that is called, in this part of the world, it's called a hurricane category one. Cyclone category too, it's called a Typhoon, it's called a Severe Typhoon, you know, it's called a Tropical Depression. Depending upon where you are, what lingo they use, is all based on how fast the speed is at the eye of the storm, pretty much, and the wind gusts and such. So, that's the first one. So, because I had someone try to convince me years ago that they were different. So, like, somehow a Cyclone was different from a Hurricane. I said, "No, it's... no." It's just a different name for the same thing. Fortunately, tornadoes are generally called tornadoes wherever you are, I think. It's certainly that I'm aware of anyhow. Yeah, you'll occasionally see some of them referred to as water spouts. Oh, sure. Which is basically a tornado, but it exists over, you know, you'll see them off the coast, you know, off the beach, sometimes Florida, Florida here in the US gets a lot of them off, like say, you know, down near Miami or Daytona or, you know, off the coast. off the coast. We just had a hurricane come through very recently that was the first storm of the season that started off the coast of Florida and came up the coast and grazed the outer banks of North Carolina as it passed by. And there were quite a few water spouts that were created off the coast as it ventured its way up the east coast along the Atlantic seaboard. I think the most dangerous part about water spouts is if there's lots of sharks in them and you get a sharknado right that is really terrifying we cannot and if there is a time if there's ever a time when you find yourself in that situation apparently a chainsaw is all you need to survive I swallowed by a shark I'm happy to say or I guess unfortunate enough to say that that I actually did not see that movie but I think it basically but by viewing my Twitter stream while it was going on I felt like I saw Joel you have not lived, mate. You have not lived until you've seen that movie and you have just been aghast at the fact that they made that movie. And even more shocked that they've made a second one. Yeah. I am so looking forward to seeing so that I can laugh at it and it's going to be brilliant. Anyhow. Okay. So, sharknadoes to one side. We're going to stick with tornadoes and hurricanes. So, we'll stick with those. Okay. So, sorry. I was going to say, I didn't grow up with, we don't have as many tornadoes here on the, at least in the east coast of Virginia as they do out in the Midwest. And most of my experience with tornadoes were related to hurricanes. Usually the tornadoes here will happen when we have a hurricane coming off the Atlantic and that hurricane's rotation will cause smaller rotations, smaller vortices to form and cause tornadoes in the area. I remember about the mid-2000s, there was a particularly strong hurricane called Hurricane Ivan that hit the Gulf Coast and then came up through the southern US and up the coast over the land. It didn't go back out over the ocean. But the storm was strong enough that it caused tornadoes all up and down the coast as it made its way up the coast. And we had quite a bit of them here in Virginia. and quite a bit for us is half a dozen. We don't get a lot of them. And I have a lot less experience with, I've never seen a tornado in person. I've been near one when one has been a few miles away, but I've never actually seen the effects of it directly. Now my wife, she grew up in Indiana. And all through her growing up, in school, they were made to practice tornado drills, and there are tornado sirens, and the children are instructed to get safe places within a building you can get in order to protect yourself in the event that one happens. In contrast, here in Virginia, we don't even have tornado sirens. And this is something that once we were married and there was a tornado nearby, she's like, "Why aren't the sirens going off?" I'm like, "What are you talking about? We do not have tornado sirens here." And she was horror struck like, "Really? This is awful." You know? How can you not have them? Right, exactly. She was like, "This is irresponsible." You know? It's like, - That's right. - We really don't have tornadoes here. - Yeah, well, I mean, I'll echo that sentiment because we don't generally get tornadoes in Australia. And we had an incident only a year and a half ago where an ex-tropical cyclone triggered a bunch of water spouts just off the coast. And one or two of them actually touched down on land. They were gone inside 60 seconds. They were really, really brief. But it was the same kind of thing. But the really big bad tornadoes require that right set of circumstances geographically, which, of course, is why they call the center of the United States Tornado Alley, because that's one of the parts of the earth where all of the conditions are right to create tornadoes. So bottom line is just real quickly to compare and contrast, hurricane cyclones, they're huge. They cover hundreds of kilometers or miles in area. the wind speeds in the center do not get anywhere near as fast as in a tornado. So you're looking at a category three cyclone, for example, hurricane, similar, you know, kind of wind speeds greater than 130 kilometers an hour, which is greater than 80 miles per hour, one minute sustained wind speeds. And you're going to cause damage to your house. Whereas even a small tornado like an F1 or the modified F1 scale is going to cause damage to your house. But the difference is tornadoes are highly localized. So it's possible for a small tornado to rip through a suburb and take out five houses. But if you've got a cyclone that's blowing at those sorts of wind speeds, it's going to take out all of the houses or most of them. Right. So, and so, so the, the two scales that these, these types of storms are measured on is the tornado scale is measured on something called the Fujita tornado damage scale, which was developed in 1971 by Theodore Fujita at the University of Chicago. And basically the way tornadoes are rated are on a scale from F0 to F5. And an F0, and you'll have to help me out here with metric, but F0 is basically anything less than 73 miles per hour. An F1 is 73 to 112. An F2 is 113 to 157. An F3 is 158 to 206. An F4 is 207 to 260 and an F5 is 261 to 318. And of course there have actually been tornadoes measured at F5 above 318, but they're super, super, super rare. Now to contrast that, the hurricane scale is known as the Saffir-Simpson hurricane scale, which was developed I think by two people whose names were Saffir and Simpson, so thus they share the name. And you'll hear hurricanes referred to as category 1 to 2 to 3 to 4 to 5. And just to contrast that, a category 1 hurricane is 74 to 95 miles per hour. A category 2 is 96 to 110. A 3, which a 3 is, a category 3 hurricane is known as the first major size hurricane. It's a 111 to 129. A 4 is 130 to 156. And a 5 is 157 or higher. So a category 5 hurricane, which is known as the most intense size hurricane, which on the scale is listed as catastrophic damage will occur. A high percentage of framed homes will be destroyed with total roof failure and wall collapse. Fallen trees and power poles will isolate residential areas. Power outages will last for weeks and possibly months. Most areas will be uninhabitable for weeks or months. that's at 157 mile per hour wind. Comparing that to tornadoes, a 157 mile per hour wind tornado is only an F2. So it goes up to F3, F4, F5. So even the largest most destructive hurricanes is just a mid-sized tornado as far as wind speed. Yeah that's it. But the thing that's interesting is that everyone looks at the damage that the different storms will cause and they'll say "oh wow these tornadoes are terrible because they're ripping everything apart but what they don't see it or is less visible perhaps is the fallout afterwards and the fallout from a cyclone hurricane typhoon is always greater in terms of things like localized flooding, widespread damage to central services so I mean a tornado will take out you know worst case maybe a suburb or two I don't know but yeah a small town if it's a small town or a small section of a large town you know yeah don't get me wrong that's terrible and I seriously you know in all of this I don't want to make light of the fact you know people die in these extreme weather events okay that's it's a horrible terrible thing don't don't misunderstand what I'm trying to point out here is all I'm saying is that it's a different set of tragedies and the tragedy of uh of a cyclone in in my experience has been the constant ongoing winds for hours or days at a time. It just, overhead power lines just get destroyed, absolutely destroyed. And so, you're then without power for days, weeks at a time. So I read to you the description of the damage of a top hurricane, the description of the damage of a top tornado, an F5, reads as saying "strong frame houses are leveled off foundations, swept away. Automobile-sized missiles fly through the air in excess of a hundred yards. Trees are debarked." The bark is stripped off the tree. And then it just sums up by saying "incredible phenomena will occur." And so you don't get that level of damage in a hurricane, but like for instance, and again this is my personal experience on the East Coast, down in North Carolina there's a little spit of land that sticks out from the coast, referred to as the Outer Banks. It's basically a string of sandbars that are strung together by a road and the little islands are connected and there's a lot of, it's a great vacation spot. There's tons of beach houses, rental properties, that sort of thing. But that little spits of sandbars are so far out from the main shore that there's quite a large sound in between those sandbars and the actual mainland. And one of the problems that the Outer Banks has is when a hurricane comes through, it's not the wind damage. I mean, granted, you get wind damage and that sort of thing. But the problem is all of that wind from this giant rotating storm forcing the water off of the ocean through the inlets into that sound and the sound depth, which is normally only like four or five, six, ten feet at most, will rise to say 15 or 16 feet and then when the storm passes over and the wind shifts, because remember it's a swirling cyclone and when it passes overhead the wind will shift from being one direction to the other and it will start to suck that water back out of the sound and there's so much water and there's not anywhere for it to go that rather than it going out the inlets it will go over the land and push everything back out to sea or in the case of the banks which the problem they've been having over the last couple years is it's been making new inlets it split the island in half and they've had to build a temporary bridge over this spot which luckily with this last hurricane it didn't do very much damage to they were able to repair it fairly quickly. But, but the, the, and the banks, there are houses there that, that 15 years ago there was a house there, now that's, that's the ocean. You know, part of that island is now gone, just from erosion. The, the, yeah, hurricane, the most of the damage is just what they call, they refer to a storm surge. Yes. Which is, which is the, the, the water coming on shore from these strong winds over a 400-mile area just pushing the ocean up onto the land, and then when the storm recedes, of it all flowing back out. Almost like a very slow, methodical, deliberate tsunami, except in coming in one big wave, it's steady over a 12-hour period. Exactly right. has caused a lot of damage particularly you know because wherever they in order to actually gain and maintain their strength cyclones hurricanes typhoons they need a body of water that's got you know that's warm in order to actually maintain their intensity so what typically happens is once they cross onto land they very very quickly lose speed they lose energy and they start to die so inevitably the storm surge and the coastal areas are the ones that hit the worst. So yeah, and that's typically the way it goes. But when they do devolve, they turn into what we would call a tropical low or a tropical rain depression. They can still cause flooding from that point. It's just the wind is less of an issue. Yeah. And it's just the fact that, you know, they can dump 10 inches of rain on a small localized spot in a short amount of time, you know. Yes. Exactly. Exactly. So one other issue you were saying about, you know, as they come on shore, you know, they will dissipate. Also I actually was unfortunate enough to miss out on this just this past hurricane. I keep mentioning the Outer Banks. Some very good friends of mine were actually vacationing there when the storm hit. And my wife and I were supposed to be there with them. And due to personal circumstances, we weren't able to go with them this year. But they elected to ride out the storm. They did not evacuate inland and they just stayed in their beach house through the storm. And the reason they did so was that when, of a hurricane, you'll see a hurricane like the photo on radar of this massive storm that stretches like it's the equivalent of two or three states in size. You know, states or provinces. I'm not sure how big your provinces or your local areas of government are in Australia. But in the US at least, you know, a hurricane sometimes when you overlay it over top of the US, it'll take up like two or three states worth of landmass. So this is a massive storm, but generally the really strong winds, the hurricane force winds are at least on a category one or two or three size storm, are generally right around the center of the storm, right around the hurricane's eye. That's the strongest part. And at least on the Gulf, when the landmass is to the north of the storm, or on the Atlantic coast when it's to the west of the storm, there are parts of the land that where the eye hits that say 30 miles down the beach to the left will only experience say 50 to 60 to 70 mile per hour winds, where 30 miles to the beach to the right will get 110 mile per hour winds because of the difference on the side of the eye that you're on when it comes ashore. And they lucked out and and their beach house was on the quote good side of the eye, the west side of the eye, because at that point the eye had came on shore, blown over the shore, rotated, and was coming back off the shore and had lost enough strength, where they only experienced between 50 and say 75, which is like a weak category one strength wind. And just up the coast from them, 30 or 40 miles away up at Cape Lookout, they received category two, some category three force winds. So, you know, between 100 and 120 miles per hour. And where he said that they were at, their house, you know, these are elevated beach houses, you know, kind of on stilts. He said the house shook a little bit, the windows shook a little bit, it rained a lot, it wasn't that big of a deal. Whereas up the coast, there was like, you know, major storm surge and major property damage. And so you'll have interesting phenomena like that with hurricanes based on where you're at relative to the eye and where it comes on shore. - Cool, so just 'cause you brought up a couple of cyclone or hurricane experiences there, just tell you about a couple of mine, just the two. Well, we'll see, maybe three. See, I grew up in a town called Rockhampton and it's right on the Tropic of Capricorn. So we are on the southern edge of where cyclones will typically come ashore. So Rockhampton though is a 45, 40 kilometers, something like that inland from the actual coast, which is about 30 miles inland, something like that. And as a result, it's usually taken the edge off by the time it gets into the town. Anyway, so in 1991, X-Tropical Cyclone Joy basically crossed the coast north of Rockhampton, well, relatively north, but it dumped all of its rain on the catchment area for the Fitzroy River. The Fitzroy River is the river that goes through Rockhampton. And it was the third largest flood event in recorded settled history. And I say settled history is since, you know, since White Settlement, you know, since it was settled there in like 18, early 18s, hundreds, wherever it was on there. Anyway, so that was quite an interesting experience, because the city was cut off for three weeks because of flooding. So the airport is built on low lying land, so they couldn't take off land, any planes. So that was, you know, interesting. There was another one in I think it was 93. And it basically was another cycle by the time it got to us it was an ex-tropical cyclone but it was still gusting pretty impressively and it rained for essentially what felt like about three weeks, it was less than that but it was about two weeks and two days I think but it literally just rained and it was always that it was a driving rain that you get in a cyclone where it's just like wave after wave and it's just I don't know how to describe it but you know what I mean? We experienced the same sort of thing with when Hurricane Sandy, I don't know if you remember when that? Yes. So Sandy was a weird circumstance where it's formed as a hurricane in the Caribbean. It came up the coast well out to sea and then right as it passed by Washington D.C. and Baltimore, it just took a hard left hand turn and crashed into the coast right around New York City. And it was large enough that as it got up the coast kind of due east to us. For about a 16, 18 hour period, we were buffeted by its bands, the different bands as it came on shore. And we were at the time living in a condo unit that was on the first floor of the building that was a corner unit. So we had windows facing, I believe, north and windows facing to the east. And for the first, say, 8 to 12 hours of these bands coming in, the rain was pretty much blowing steadily up against the east face of the building, almost like horizontal rain coming right into the windows. And then as the storm got further north and made its turn, and due to the way it rotated, all of the wind was coming in from the north-northwest. And we had leaking in our condo under the windows because this building was about 15 to 20 years old and the caulking around the window had had kind of you know cracked and dry rotted and that horizontal rain forcing into those cracks where the caulking forced rain in under the the window and actually you know got our carpet wet along that the walls right on it was first on the one side of the room as the wind was coming in that way and then as it shifted it shifted to the other side of the room yeah and and that that's like I said with hurricanes that that's cyclones That's what you get. It's just relentless pounding for hours and hours and hours. Well, that's right. And it's a very different, it's a very eerie experience. The fact that the sound of that, the wind and the rain just unrelenting. And yeah, it was also more recently, we went through that again recently with X Tropical Cyclone Oswald that was in 2013 and it was so widespread in Southeast Queensland so that was actually hit down here in Brisbane and it was so, the damage was so widespread that we personally were out of, we had no electricity for about four days and everything in the fridge and freezer spoiled, we had to throw it all out and we basically, you know if we were fortunate that it was only I say only if you consider Greater Southeast Queensland something like three million three and a half million people out of the Greater Southeast area one quarter of houses had their power lost for a period of four or more hours during that that four or five day period we were just unlucky to be in one of the areas that the repair crews got to last or towards the end of the list lucky us anyway so we were without power for four days but fortunately my in-laws they had power so we I remember we went to their place to charge our iPods iPhones and so on and yeah we ate at their place because we couldn't do anything we had we had a barbecue but the barbecue was waterlogged and yeah we couldn't get it to fire and to stay fiery because it was still even even after the rain stopped it was still very windy for a few days afterwards so keeping it something burning was kind of difficult. But anyway, that's nothing of course compared to you know, I've had friends where I grew up who were from Townsville and you know, and Cairns and they'd been, they'd actually live through cyclones where they'd lost the roof on their house. So I mean, this is all nothing compared to you know, sort of damage it can do. But just to just to wrap up quickly on the extreme weather though, is that I find that I find cyclones, I think do more damage in terms of terms of overall net cost. But because you tend to see them coming, you have an opportunity to prepare, you have an opportunity to get out and evacuate. There you have many, hopefully, unless it's a weird circumstance. And sometimes there are weird circumstances. But most of the time you have minimum three to five days warning sometimes longer than that. Exactly. And if the again, if the storm doesn't change a lot, and the models do a good job at predicting, you know, they pretty much know where it's going to hit and what populations are going to be affected and they have an accurate prediction of what the storm surge will be. So, based on your elevation, they'll issue evacuation orders, at least here in the US, they'll issue evacuation orders along the coast to tell residents to get inland a bit at least for the next 48 hours to board up the windows of your house, all that sort of thing. Absolutely. Absolutely. It's exactly the same here. And I find that though, because I actually I did some work in Richardson many years ago and I was stunned walking along the corridors and you'd see tornado shelter and a little sticker on the next to the doors, some of the doors, and it was like tornado shelters. And I was down there at a particular time when a tornado had actually ripped through Fort Worth, which was not too far away from Dallas. And it was all very fresh. It was only three or four weeks previously. And it was all very fresh in everyone's minds. I'm thinking to myself, I'm walking around this building in Richardson and thinking, "Okay, where's the nearest tornado shelter?" I've never had to think about this before, but if there was a tornado, that's where I would have to go and hide. And I guess that's the thing that makes tornadoes so terrifying is not so much the fact that... I mean, they have highly localized damage, however, the fact that they are far less predictable, you don't tend to see them coming specifically, that's what makes them, I think, more terrifying. Yeah, and the warning for them can be as little sometimes as 5 to 15 minutes. That's right. And just think about all the time during the day in which you are not near a computer or don't have the TV on or aren't looking at a weather app and, you know, let's say you're taking a shower or you're doing the dishes or you're in your car listening to a podcast and not, you know. So there's so many instances in which a storm could be coming in and they are sure a tornado is going to hit and they've got five minutes to warn you, but people just don't see the warnings because it's hard to get a warning out when you only have five minutes notice and get everyone's attention. Hence the tornado alarms that you were mentioning. And that's how the Midwest copes with it is they just put on these big poles in every town around the whole area this giant siren that goes off and everyone can hear it. It's a they do it at this it's almost like an air raid siren back from World War II. In fact I think after World War II happened they repurposed many of the air raid sirens as tornado sirens. Yeah it's just so everyone in the entire town no matter where you are what you're doing you can hear the sound of it and so you know you know there's a tornado on the ground somewhere get into shelter. Yep that's it. So one of the other things just quickly want to touch on before we wrap up is spaghetti forecasting with regard to extreme weather. So that's something that you mentioned specifically you just want to have a quick chat about. So spaghetti this is the the term is spaghetti model and and the reason it's referred to as a spaghetti model is that When the different models will try to predict a cyclone or a hurricane, it will, you know, with say a winter storm or a summer storm, you're mainly trying to figure out where the front's going to move through and what the rainfall amounts are going to be or what the snowfall amounts are going to be, that sort of thing. With a hurricane, you want to figure out where the eye is and where it's going to be. So a spaghetti model will actually plot a colored line on a map to have that model's prediction of exactly where they think the eye is going to go and thus they could then predict where the most damage is going to be based on what I was saying earlier with on one side of the eye it'll be less than the other side so it's still really hard for them to accurately predict more than just a few hours away exactly where that bad side of the eye is going to hit but the way the spaghetti models work is say the North American model will plot a line you know, line A and the European model will plot line B and the Canadian model will plot line C, well then, when you superimpose all of these together over a map, it'll look like strings of like, say, spaghetti. They're different colored little lines and it's all of the different possible routes the models are saying it will take. And just like with this ensemble forecast over land, they generally will take the ensemble of all the lines and average them together to figure out where the hurricane's gonna go. - Yeah, exactly. Yeah, cool. - I'm looking at a great map here. NOAA maintains a website, a section of their website, which is known as historical hurricane tracks. And after a hurricane actually happens, they're able to say, okay, let's throw out all the models. this is the actual route that the eye took. And here's the actual route that every eye of every hurricane has taken over the past ten years. And so this map is wonderful in that it's just, it has tons and tons and tons of these little colored lines on it with the name beside each one of showing the route of every hurricane. So you can see generally where they all hit on the coast. And you'll occasionally see this little line that's separated from all the others where this one storm took this little weird turn that most of them never take and that's due to you know the butterfly effect something in the atmosphere that particular week calls that storm to do something that usually none other storms do which just again reinforces how hard it is this you know for forecasters how exactly is exactly and and that's where I would like to wrap it up right there is is the fact that this is not easy predicting the weather is very very difficult it's a very very complicated problem in fact yeah what you're essentially doing is you're trying to predict the future and you know there's so many inputs there's so many variables and we don't have the technology to model absolutely every molecule of air in the entire world and we don't have you know we don't have the computer technology to compute that we don't have the models it's just you know it's hard is very hard and I feel like you know whether prediction is better today than it ever has been in all of human history and yet people are still, they say it's going to rain this weekend but you know, whatever. It's like well, wind the clock back 20 years, 10 years even. I mean things have come so far and they're getting better all the time and they'll continue to get better but ultimately, stop blaming the weather person and I'm careful not to say weather man, weather person because you know honestly weather prediction is hard very hard and honestly if if you don't take that away from what we've talked about this episode then yeah maybe maybe haven't been paying attention and something else I would just want to mention like like if you pull up a I have this I have the one app that I use it's kind of give you a general forecast is this one one app called today and I like it because it has a very nice little interface and sometimes its data isn't the greatest, but I like its interface. Like I'm looking at today's weather, it's saying there's a 61% chance it's going to rain today. What many people don't realize is that the forecaster will look at the models, and the models will basically say, "Well, this area right here is due to have some rain." So the forecaster will basically draw a little circle around that area, and they'll see, on the model what chance that there's going to be rain somewhere in that area. And when you see there's going to be a 60% chance, that's 60% chance of the small little geographical region that you live in. And that area may be 10 miles wide, 30 miles wide, 50 miles wide, that depends on your local forecast area. But and then if later in the day it rains in a couple places in that area, they consider that a successful forecast. Maybe it didn't rain in the exact spot over your house, but it rained in that area and it was almost impossible for them to know whether your particular front lawn is going to have rain or not versus maybe the storm passed and just missed you two miles away. But that was a successful forecast because it's almost impossible without a lot more data and a lot more computing power before they will be able to do forecasts that are that accurate. If you'd like to talk more about this, you can reach me on Twitter @johnchidgey and check out my writing at If you'd like to send any feedback, please use the feedback form on the website and that's where you'll also find the show notes for this episode under podcasts pragmatic. You can also follow Pragmatic Show on Twitter to see show announcements and other related stuff. This episode was proudly sponsored by ManyTricks and I'd like to thank them for sponsoring the show. If you're looking for Mac software that can do ManyTricks, remember, specifically visit this URL, for more information about their amazingly useful apps and use the discount code pragmatic25 for 25% off the total price of your order. Hurry, it's only for a limited time. I'd also like to thank LIFX for sponsoring this episode. If you're looking for a great LED light bulb that's energy efficient, remotely controllable, colorful, and just plain fun to use. Remember to specifically use this URL,, and use the coupon code pragmatic for 15% off the total price of your order. And of course, finally, I have to thank Joel for coming on the show. Thank you so much for coming on and talking about the weather. If people want to get in touch with you, what's the best way they should do that? - Well, these days I'm pretty Twitter prolific. So at Twitter, I'm just my first name and last name, which is J-O-E-L-H-O-U-S-M-A-N. So it's @joelhausman. And then, or you can reach out via the contact form on my website, which is also - Fantastic, awesome. We can all stop worrying about whether or not it's supposed to be raining and just smile and be happy. - Yeah, thanks for having me on. I love to get geeky about the weather. Our local, in my opinion, best weather sources It was a former weather blog. Now they're part of the actual Washington Post's weather section of their website. There's a few of those guys I follow online. And then I warn anyone, if you follow me on Twitter and there's a hurricane about to happen or there's a blizzard in the winter, be prepared for weather tweets. Cool. Awesome. Well, look, thanks again for coming on and I really appreciate your sharing. And yeah, thanks again for listening, everyone. Thanks a lot, John. No worries, mate. [MUSIC PLAYING] [MUSIC] (upbeat music) (upbeat music) (upbeat music) (music) (upbeat music) (upbeat music) (upbeat music) (upbeat music) [MUSIC] [Music] (explosion) [BLANK_AUDIO]
Duration 1 hour, 25 minutes and 23 seconds Direct Download
Episode Sponsors:

LIFX: LIFX is a smart lightbulb that gives you previously unheard of control of your lighting. Each bulb is Wi-Fi enabled, can give you light in whatever colour of the rainbow you like, and is an energy efficient LED light bulb that you can control with your smartphone. For developers LIFX are running a competition with great prizes please check it out and submit your app by the 25th of July, 2014. Visit and use the Coupon Code PRAGMATIC for 15% off the total price of your order.

Show Notes

Related Links:

Weather Apps Joel and John Use:

Movies Mentioned in this Episode:

Premium supporters have access to high-quality, early released episodes with a full back-catalogues of previous episodes


Joel Housman

Joel Housman

Joel writes on his blog.

John Chidgey

John Chidgey

John is an Electrical, Instrumentation and Control Systems Engineer, software developer, podcaster, vocal actor and runs TechDistortion and the Engineered Network. John is a Chartered Professional Engineer in both Electrical Engineering and Information, Telecommunications and Electronics Engineering (ITEE) and a semi-regular conference speaker.

John has produced and appeared on many podcasts including Pragmatic and Causality and is available for hire for Vocal Acting or advertising. He has experience and interest in HMI Design, Alarm Management, Cyber-security and Root Cause Analysis.

Described as the David Attenborough of disasters, and a Dreamy Narrator with Great Pipes by the Podfather Adam Curry.

You can find him on the Fediverse and on Twitter.