4 ms·
I posted a visualization of the shock wave as a submission, but I"ll repost it here. This is a satellite measurement of integrated atmospheric water vapor (IIUC
by lnwlebjel 5y ago
I posted a visualization of the shock wave as a submission, but I"ll repost it here. This is a satellite measurement of integrated atmospheric water vapor (IIUC), so changes in the density of air change the integrated measurement, such that you can see the shock wave travel around half the globe.
https://cimss.ssec.wisc.edu/satellite-blog/images/2022/01/220115_goes17_waterVaporTimeDifference_anim.gif https://cimss.ssec.wisc.edu/satellite-blog/images/2022/01/22...
(it take a few seconds to load, but I believe this is one of the most incredible things I've ever seen).
- roieki 5y ago
- _qua 5y agoThis is a 140 MB gif ◑﹏◐ I hope you don't mind, I reuploaded as an MP4: https://imgur.com/BbaLE5U https://imgur.com/BbaLE5U
- LeoPanthera 5y ago139MB is utter insanity for an animated GIF. Crushing it with gifsicle can get it down to 125MB but clearly GIF is the wrong choice of format here. Here's a HEVC compressed video version which is only 25M and suffers only extremely mild softness: http://david.gloveraoki.net/f/32M.mp4 http://david.gloveraoki.net/f/32M.mp4 It would probably be even better if I could work with the source video and not the gif version.
- wingworks 5y agoidk, I had no issues loading the gif and playing it. For some reason your video doesn't want to play in Chrome. I had to download it first.
- jcun4128 5y agoIs interesting Gif size. With ShareX I get massive Gifs, but Kap on Mac I can get much smaller files (this is 14fps). Same for screentogif massive files... not sure what makes them so big.
- mlindner 5y agoYour video doesn't work either but the GIF worked fine. I get MIME type not found for yours. Also the video is full of very fine scale noise. Something that won't compress by HEVC very well at all. You want a lossless compression method (or a very high perceptive index on HEVC).
- graderjs 5y agoI love how the helpful engineers of hacker News come to the aid of the poor soul who's inadvertently uploaded their video of the biggest eruption in the last 30 years in the wrong format or without the proper compression. I love that...it's like, 'happy to help to make everything more efficient!' In other news IT still takes days to get back to me about getting my BYO device onboarded... I'm not saying it's a paradox just noticing the juxtaposition
- 5y ago
- tailspin2019 5y agoIncredible! Thanks for posting this.
- fghorow 5y agoVery cool! Gives great intuition for geometric spreading on the surface of a sphere! Modulo phase velocity and attenuation, it makes clear that the amplitude/energy density could actually increase past the great circle!
- nomel 5y agoI wonder if the eruption could be heard on the exact opposite side of the globe, where the waves converged? This would have been somewhere around 20.570000, 4.620000, in Guezzam, Algeria, according to: https://www.antipodesmap.com https://www.antipodesmap.com
- lnwlebjel 5y agoI was wondering the same thing. One factor might be the variation in propagation speed, which is visible for example due to winds near low pressure systems.
- graderjs 5y agoAnother fact is the Earth is not a uniform sphere so the waves wouldn't exactly converge in a uniform circular condensation on that exact spot. Unless you already factored that in...
- _moof 5y agoNot sure if anyone heard it but I know it registered very clearly on barometers in London, so it's not out of the question.
- cjblomqvist 5y agoIt was apparently registered 3 times on Iceland (almost opposite side of earth) - apparently it was big enough to travel 3 times around the globe!
- austincheney 5y agoWhat’s happening immediately east of the eruption? There is a point water vapor variability in the center of the shock wave start and another single point of water vapor variability of equal size without a shockwave.
- dmissp 5y agoHonest question - is this animation in real time, or sped up?