The article says some 20 km/s, but that's its speed in space. The video is crap (just repeats the same footage four times, no slow motion) but downloading that m3u8 with yt-dlp and running it frame by frame in mpv: it shows up on one frame (well, two because it does interlacing) and then there's already a dust plume. The camera looks pretty high up, maybe 3 meters? With it first appearing roughly in front of the camera (vertically), that would mean a lower bound of 3 meters in 1/30th of a second, or 1/15th if this interlacing makes it 15fps in reality which I suspect, meaning somewhere north of 324 or 162 km/h in either scenario
It looks like a decent-sized stone to me on that one frame, maybe the size of one's hand? Call it a 15cm sphere for simplicity, WolframAlpha says that has a 177 cm² cross-section and https://www.omnicalculator.com/construction/stone comes up with a few kilograms (4-8) depending on the type of rock, with which we can fill in a terminal velocity calculator https://www.omnicalculator.com/physics/terminal-velocity to get an estimate of 320 to 450 km/h depending on its weight (probably towards the higher end given that it's metallic) -- assuming we got the size anywhere near correct, of course
Edit: someone posted (https://news.ycombinator.com/item?id=42822256) a YouTube link where it's visible for 2 frames! (I'll call them frame 1, 2, and plume.) Now the prize question is how far up it was in the first frame and what angle it comes in at. The biggest chunk looks to be about the same size in both frames, at least to within video compression artifacts, so that doesn't help figure out the angle. It also travels nearly twice the distance in pixels between frame 2 and plume as compared to frames 1 and 2, which would be an angle thing or a video compression thing. I don't see an interlacing effect in frame 2 anymore, and frame 1 is duplicated (the youtube video runs at 24 fps). I think this has passed through too many compression / re-encoding layers to say anything useful sadly. For the heck of it: say that it was 5 meters up in frame 1 and it takes 2/24ths of a second to traverse that, that would be 216 km/h
Loading the audio tracks into Audacity, the YouTube version shows nothing until ~180ms before impact if I'm being generous, whereas the CNN version contains inaudibly faint noise starting around ~350ms before (don't know that that's attributable to the meteorite, but at least there's data instead of a flat line), which I guess tells us that CNN's version has undergone fewer recompressions. Sadly, I don't think there's enough samples to measure the doppler effect, also given that it's not going directly towards/away from us (the shift is more like 60° probably?). Three meters are traversed in 24ms at 450km/h, and while there's audible noise on both YT and CNN for ~85ms before impact, the spectra are just too noisy for me to make it out. Someone more experienced with measuring dopplers could have a stab at it, though
More editing: it is only now that I properly read the article (I know, shame on me, I guess DIY is more fun) and it says a "professor of Earth and atmospheric sciences at the University of Alberta" also tried this approach but hasn't figured it out yet: '“We are working on analyzing the video to see if we can say more about the meteorite’s fall — including speed,” Herd said.'
I'm curious if we can figure this out by dropping rocks at different speeds until we get a plume that expands as fast as this one (since we have enough footage to reliably see t...
When I put all the population estimates into a spreadsheet the last time it came up, the number of person-years lived can be roughly divided into thirds, one third from the dawn of anatomically modern humans to the beginning of agriculture, one third from the beginning of agriculture to 1600, and the remaining third from 1600 to today. By current estimates, well acquire an equal slice of lived experience between now and 2100.
Well, the important thing is the number of person-hours not the number of people since meteorite death has to be pretty random.
If 5 people took turns existing in sequence that's not more likely for a human observation than one human over that time. It only matters how many hours where humans were available to die and that's roughly matched between prehistory and the last 100 years.
If you want to get a really good estimate it’s worth considering the odds that either 0 or thousands of people dying in the same event when people are living in dense cities etc.
Tunguska over London would have had a serious death toll where an airburst in a remote area didn’t. But you’re going to see deaths in an area with a population density of 1 per square km or 100,000/km2.
Written records of events didn’t start everywhere as soon as the first group invented writing.
More relevant when considering the criteria of well documented evidence, a hypothetical ship on an ocean voyage in 1500 is likely to just disappear. Let alone a random West African goat herder in 2200BCE.
The Wikipedia page links to this [0] article as a source, where the table states an estimated 9 billion “ever lived” at the 8000 BCE mark, and 55 billion at the 1 CE mark. The end of prehistory isn’t too well defined, but it’s usually taken to be around 5000 BCE (if not later). So while it’s not very precise of course, from those estimates it certainly looks like more humans lived in prehistory than now, by a few billion at least.
Given that the world population is predicted to peak at about 10.5 billion roughly 60 years from now, it also looks like this won’t change anytime soon.
For comparison there’s still 300,000 Wild West African Gorilla today and they’re extremely restricted in range, share that range with modern humans, etc.
Human populations in prehistory were believed to average somewhere in the 1-15 million range which puts 10B at the very low end estimate with 100+ billion being considered plausible. There was at least one major bottleneck but humans predate the last ice age.
PS: Calling anything history before we have any surviving text to work with is also a stretch.
There's probably a Tunguska-sized event at some point that claims 99.99% of all meteorite victims and wasn't registered because everybody that knew about it died in it.
I wonder if any people died as an indirect result - someone startled by the flash of light or the bang, who stepped into the path of a carriage, or just slipped and fell and cracked their skull.
There is one theory that that the biblical story about the annihilation of the city of Sodom was brought about by a large air-burst meteor, think Chelyabinsk and add a couple of order of magnitude.
However, note that when I was trying to find the original paper, that is, something more substantial than the popular news blurbs that are everywhere, I found wikipedia was less than impressed with the evidence.
Anyhow, I suspect a lot of our long term generational history gets turned into religion and legend, that is to say, the legendary events probably happened but like a good fish story the facts get shifted over the years.
According to the estimates in https://www.prb.org/articles/how-many-people-have-ever-lived..., (substantially) less than 55 billion lived in prehistory, and more than that lived since (the current estimated number of humans ever having lived being above 117 billion).
Pro Tip: While the YouTube player is paused, the comma and period keys move back/forward by one frame. (Memory aid, < and > but without using Shift -- assuming US PC keyboard layout.)
1920x1080@20fps, and there's 2 frames with the meteor in it before it impacts.
I'm messing around a bit, but no matter how I turn it over, I'm getting something like 36 m/s for the meteor's (terminal) velocity, while I'm expecting it to be going at least 60 m/s
Possibly I'm making a mess of figuring out the actual scale of the image/video.
Posting here in the hope someone can do this a lot better than me!
Oh! The google drive link also provides the divot with a ruler next to it. Using that instead of going by a rough guesstimate based on just a couple of pixels really helps. In that case the theoretical terminal velocity should have been about 39 m/s , which is pretty close to the speed I estimated. I'll take it.
61 comments
[ 2.5 ms ] story [ 122 ms ] threadThe article says some 20 km/s, but that's its speed in space. The video is crap (just repeats the same footage four times, no slow motion) but downloading that m3u8 with yt-dlp and running it frame by frame in mpv: it shows up on one frame (well, two because it does interlacing) and then there's already a dust plume. The camera looks pretty high up, maybe 3 meters? With it first appearing roughly in front of the camera (vertically), that would mean a lower bound of 3 meters in 1/30th of a second, or 1/15th if this interlacing makes it 15fps in reality which I suspect, meaning somewhere north of 324 or 162 km/h in either scenario
It looks like a decent-sized stone to me on that one frame, maybe the size of one's hand? Call it a 15cm sphere for simplicity, WolframAlpha says that has a 177 cm² cross-section and https://www.omnicalculator.com/construction/stone comes up with a few kilograms (4-8) depending on the type of rock, with which we can fill in a terminal velocity calculator https://www.omnicalculator.com/physics/terminal-velocity to get an estimate of 320 to 450 km/h depending on its weight (probably towards the higher end given that it's metallic) -- assuming we got the size anywhere near correct, of course
Edit: someone posted (https://news.ycombinator.com/item?id=42822256) a YouTube link where it's visible for 2 frames! (I'll call them frame 1, 2, and plume.) Now the prize question is how far up it was in the first frame and what angle it comes in at. The biggest chunk looks to be about the same size in both frames, at least to within video compression artifacts, so that doesn't help figure out the angle. It also travels nearly twice the distance in pixels between frame 2 and plume as compared to frames 1 and 2, which would be an angle thing or a video compression thing. I don't see an interlacing effect in frame 2 anymore, and frame 1 is duplicated (the youtube video runs at 24 fps). I think this has passed through too many compression / re-encoding layers to say anything useful sadly. For the heck of it: say that it was 5 meters up in frame 1 and it takes 2/24ths of a second to traverse that, that would be 216 km/h
Loading the audio tracks into Audacity, the YouTube version shows nothing until ~180ms before impact if I'm being generous, whereas the CNN version contains inaudibly faint noise starting around ~350ms before (don't know that that's attributable to the meteorite, but at least there's data instead of a flat line), which I guess tells us that CNN's version has undergone fewer recompressions. Sadly, I don't think there's enough samples to measure the doppler effect, also given that it's not going directly towards/away from us (the shift is more like 60° probably?). Three meters are traversed in 24ms at 450km/h, and while there's audible noise on both YT and CNN for ~85ms before impact, the spectra are just too noisy for me to make it out. Someone more experienced with measuring dopplers could have a stab at it, though
More editing: it is only now that I properly read the article (I know, shame on me, I guess DIY is more fun) and it says a "professor of Earth and atmospheric sciences at the University of Alberta" also tried this approach but hasn't figured it out yet: '“We are working on analyzing the video to see if we can say more about the meteorite’s fall — including speed,” Herd said.'
I'm curious if we can figure this out by dropping rocks at different speeds until we get a plume that expands as fast as this one (since we have enough footage to reliably see t...
Airspeed, drag, and terminal velocity and all that I would assume it's going subsonic and appears to be quite brittle.
The angle also appears to be consistent with what I've seen other impacts produce in the 40 degrees.
The lack of sonic boom and camera catching it I am curious if this happens on the daily with nominal observation.
https://space.stackexchange.com/questions/12774/what-impact-...
Significantly more people were alive in prehistory than in recorded history. So it’s likely one or more of them were significantly injured.
https://en.wikipedia.org/wiki/Prehistoric_demography#Hominid...
This is one of those things that's probably incredibly difficult to estimate with any level of accuracy.
If 5 people took turns existing in sequence that's not more likely for a human observation than one human over that time. It only matters how many hours where humans were available to die and that's roughly matched between prehistory and the last 100 years.
Tunguska over London would have had a serious death toll where an airburst in a remote area didn’t. But you’re going to see deaths in an area with a population density of 1 per square km or 100,000/km2.
https://en.wikipedia.org/wiki/World_population
More relevant when considering the criteria of well documented evidence, a hypothetical ship on an ocean voyage in 1500 is likely to just disappear. Let alone a random West African goat herder in 2200BCE.
Given that the world population is predicted to peak at about 10.5 billion roughly 60 years from now, it also looks like this won’t change anytime soon.
[0] https://www.prb.org/articles/how-many-people-have-ever-lived...
For comparison there’s still 300,000 Wild West African Gorilla today and they’re extremely restricted in range, share that range with modern humans, etc.
Human populations in prehistory were believed to average somewhere in the 1-15 million range which puts 10B at the very low end estimate with 100+ billion being considered plausible. There was at least one major bottleneck but humans predate the last ice age.
PS: Calling anything history before we have any surviving text to work with is also a stretch.
I wonder if any people died as an indirect result - someone startled by the flash of light or the bang, who stepped into the path of a carriage, or just slipped and fell and cracked their skull.
https://pmc.ncbi.nlm.nih.gov/articles/PMC8452666/
However, note that when I was trying to find the original paper, that is, something more substantial than the popular news blurbs that are everywhere, I found wikipedia was less than impressed with the evidence.
https://en.wikipedia.org/wiki/Tell_el-Hammam
Anyhow, I suspect a lot of our long term generational history gets turned into religion and legend, that is to say, the legendary events probably happened but like a good fish story the facts get shifted over the years.
Having said that, the meteor was probably from the planet Urectum.
Link without tracking for the lazy btw: https://youtu.be/c0gnYzhTL5A
I'm messing around a bit, but no matter how I turn it over, I'm getting something like 36 m/s for the meteor's (terminal) velocity, while I'm expecting it to be going at least 60 m/s
Possibly I'm making a mess of figuring out the actual scale of the image/video.
Posting here in the hope someone can do this a lot better than me!
That said, there's entire international communities dedicated to fireball trajectory analysis.
Starting with the OG: https://fireballsinthesky.com.au/
They do set up well in advance and like to have overlapping coverage from multiple sites, of course.
It looks pretty though, which we wouldn't have known a century ago - a small, pretty, damp rock. It's not much, but it's all we'll ever have.
Original CBC article is here (2).
(1) https://islandscholar.ca/people/jvelaidum
(2) https://www.cbc.ca/news/canada/prince-edward-island/pei-char...
https://news.ycombinator.com/item?id=42694277
https://news.ycombinator.com/item?id=42719167
https://news.ycombinator.com/item?id=42699537
https://news.ycombinator.com/item?id=42733580
https://news.ycombinator.com/item?id=42742065
https://news.ycombinator.com/item?id=42721622
https://news.ycombinator.com/item?id=42807205
https://news.ycombinator.com/item?id=42750821
https://news.ycombinator.com/item?id=42780394
https://news.ycombinator.com/item?id=42745712
https://news.ycombinator.com/item?id=42736932
https://news.ycombinator.com/item?id=42767349
Oh wait...