Page 1 of 1

1723: "Meteorite Identification"

Posted: Mon Aug 22, 2016 12:10 pm UTC
by HES
Image
Alt text: Click for an actual flowchart for identifying a meteorite. My favorite part is how 'Did someone see it fall? -> Yes' points to 'NOT A METEORITE.' This is not a mistake.

Did Randall just accidentally DDOS wustl.edu? It took a couple of attempts to load it.

Re: 1723: "Meteorite Identification"

Posted: Mon Aug 22, 2016 12:14 pm UTC
by Reecer6
However, if no one on Earth saw it fall, then its true nature of being or not being a meteorite will forever mystify us, because you clearly can't send a sample at that point.

Re: 1723: "Meteorite Identification"

Posted: Mon Aug 22, 2016 12:26 pm UTC
by speising
'Did someone see it fall?' also doesn't have a 'No' exit. very randallian.

Re: 1723: "Meteorite Identification"

Posted: Mon Aug 22, 2016 12:34 pm UTC
by Flumble
*"Did someone see it fall? ----Yes---> It's not a meteorite, but it will be once you find it on the ground."

I expect no less than a lenghty descriptivist-prescriptivist discussion in this thread.

[edit]
For those who meet an unavailable site:
http://meteorites.wustl.edu/check-list.htm wrote:
METEORITE OR METEORWRONG?
Self-Test Check-List
meteorite_flow_chart.gif

Here's an adaptation of wonderful flowchart devised by Deborah Guedes and colleagues in Brazil. Although deliberately simple, we suspect that most experienced meteorite hunters would have little criticism with the logic.


Re: 1723: "Meteorite Identification"

Posted: Mon Aug 22, 2016 12:41 pm UTC
by rhomboidal
"If struck, wait until regaining consciousness to identify meteorite."

Re: 1723: "Meteorite Identification"

Posted: Mon Aug 22, 2016 2:21 pm UTC
by KarenRei
Despite knowing that the vast majority of things people suspect are meteorrites aren't, I once got tricked into thinking that an unusual shaped ("fingerprint"-looking), highly naturally polished piece of basalt might be when it easily deflected my studfinder. It was then that I learned that yes, basalt *can* contain enough magnetite to easily deflect a studfinder... wouldn't have expected that. ;)

Re: 1723: "Meteorite Identification"

Posted: Mon Aug 22, 2016 2:22 pm UTC
by Scott Auld
Start -> It's not a meteorite

Discourages scientific curiosity.

Re: 1723: "Meteorite Identification"

Posted: Mon Aug 22, 2016 2:41 pm UTC
by Caesar
The "Did someone see it fall?" box doesn't have a "No" exit. Took me a while to figure that one out. It means that, if no one had seen it fall, they would have never thought it were a meteorite to begin with. Right?

Re: 1723: "Meteorite Identification"

Posted: Mon Aug 22, 2016 3:03 pm UTC
by water_moon
If you go to the other pages linked on the self-test checklist ( http://meteorites.wustl.edu/realities.htm for those who can't get to the link thanks to xkcd's linkage) they link to several stories of non meteorites falling from the sky, from the sound of it, there's more man made space junk landing back on Earth than space rocks coming down :(

Re: 1723: "Meteorite Identification"

Posted: Mon Aug 22, 2016 3:05 pm UTC
by Shadowman615
Here's an explanation for "Did someone see it fall" :

http://www.meteoritemarket.com/metid2.htm
meteoritemarket.com wrote:You say that you saw the rock fall from the sky. Are you sure?

Meteors enter the Earth's atmosphere at speeds ranging from 14 kilometers/second (31,000 miles per hour) to 45 kilometers per second (100,000 miles per hour). At first they burn on the surface and perhaps explode from the shock. But as they go farther into the atmosphere they slow down. All but the largest meteors (like the one that formed Meteor Crater) quit burning and fall dark from an altitude of from 5 to 20 km (3.2 to 12.4 miles). That's a long fall. No human can trace the fall of a rock that far. In fact, no human can even see a small rock at that distance. Where meteorites have been observed to fall, there has simply been a whoosh and a thunk.

By the time meteorites hit the earth they are traveling at terminal velocity--that is a velocity at which the resistance of the air will not let them go any faster. They are falling no faster than a rock dropped from an airplane--or the Coke bottle in the first scene "Gods Must Be Crazy." Terminal velocity for a small object is not very high--150 to 300 km/hr (100 to 200 miles per hour more or less) or less. These impacts don't make big craters. You are more likely to see a small indentation in the ground, a small hole, or nothing.
With this in mind, here is the bad news:

If you saw a rock burn all the way to the ground and recovered it, you probably did not recover a meteorite. The rock you saw burning probably landed over the horizon.
If your rock was accompanied by a fire, something besides the rock caused the fire.
If your rock was too hot to touch or took days to cool down (as many have reported to me) then well . . . I am not sure what to say. You know what your senses registered.
Many people report that a friend or relative (often an ancestor) saw the rock fall. Are you sure? Stories get told and retold . . . and changed. Old hoaxes or tall tales designed to entertain the grandchildren metamorphose into . . . truth. Well, not exactly.


Perhaps there's no "No" exit on that one because by the time you get to that part of the flow chart, it doesn't matter -- it's already been ruled out by the previous question whether you saw it fall or not. I'm guessing they just wanted to find a place to stick that question because they get so many people bringing them samples they "saw fall out of the sky." As someone said, it's quite a Randall-esque flowchart, and maybe somewhat tongue-in-cheek also!

Re: 1723: "Meteorite Identification"

Posted: Mon Aug 22, 2016 3:07 pm UTC
by Rombobjörn
I guess Randall's point is that a layman with a random stone is more likely to arrive at the right answer with Randall's flowchart than with the one he linked to, simply because meteorites are so vastly outnumbered by terrestrial stones. He's probably right.

Caesar wrote:The "Did someone see it fall?" box doesn't have a "No" exit. Took me a while to figure that one out. It means that, if no one had seen it fall, they would have never thought it were a meteorite to begin with. Right?

It means that if you arrive to that box, then it's already clear that your stone is not a meteorite, and the box is there to deflect the argument that "I know it's a meteorite because I saw it fall".

Re: 1723: "Meteorite Identification"

Posted: Mon Aug 22, 2016 3:18 pm UTC
by orthogon
Rombobjörn wrote:I guess Randall's point is that a layman with a random stone is more likely to arrive at the right answer with Randall's flowchart than with the one he linked to, simply because meteorites are so vastly outnumbered by terrestrial stones. He's probably right.

Typical Bayesian.

Re: 1723: "Meteorite Identification"

Posted: Mon Aug 22, 2016 3:19 pm UTC
by water_moon
Scott Auld wrote:Start -> It's not a meteorite

Discourages scientific curiosity.

I'd disagree, most folks with scientific curiosity would challenge a bald statement. And there are nice links to a real flow chart on a website hosted by a real professor, who to put it in his own words "[In 2015] was contacted 2719 times by 1337 different people from at least 68 countries who thought that they had found, bought, or inherited a meteorite, had questions about funny-looking rocks, wanted to sell me rocks, wanted to chat about meteorites (I don’t chat), or who chastised me because they found my admittedly rude admonishments below to be too rude. I have spent a lot of time answering e-mail messages, looking at photos, examining rocks, and bothering my colleagues. Other scientists who study meteorites have had the same experience. As the public’s interest in meteorites increases and the price of rare meteorites remains high, I expect such inquiries to increase. In order to strike a balance between the use of our limited resources and providing a service to the community, [I'm not going to mess with your junk until you at least try to do your own leg work]"

He goes on to say that if you follow the chart to a positive outcome, contact him, do a recommend $263 lab test for chemical composition, and send him the results, you still only have about a 2 in 470 chance of having found a meteorite.

Re: 1723: "Meteorite Identification"

Posted: Mon Aug 22, 2016 3:45 pm UTC
by cellocgw
HES wrote: My favorite part is how 'Did someone see it fall? -> Yes' points to 'NOT A METEORITE.' This is not a mistake.


So....

If a meteorite doesn't fall in the forest where nobody can see it, is it still not a meteorite?

Re: 1723: "Meteorite Identification"

Posted: Mon Aug 22, 2016 5:14 pm UTC
by da Doctah
Well, no, I didn't see it fall, but the guy who gave it to me? Yeah. I saw him fall. Several times, in fact.

Re: 1723: "Meteorite Identification"

Posted: Mon Aug 22, 2016 6:55 pm UTC
by keithl
I've seen many meteors fall (I go camping during meteorite showers), but I've never found a meteorite.

About a decade ago, I saw an extremely bright meteor east of Portland, Oregon, early in the morning on the way to the airport. Someone else at the airport had seen the same meteor a few miles up the road, so it was likely to be over distant eastern Oregon, and not a firework. Perhaps someday, that object will be found and identified as a meteorite, but we'll never know that it was the fall that I saw.

Between 1964 and 1974. the Smithsonian deployed the Prairie Network, 16 cameras in a 500 km radius of SE Nebraska. They tracked hundreds of meteors, and located and recovered the "Lost City" meteorite in Oklahoma. They were mostly interested in triangulating the orbits of the objects around the Sun before they intersected the Earth.

It would be wonderful if someone would write a meteor observation app for rooted, obsolete Android phones with cameras. Those could be deployed by the hundreds of thousands, watching the skies around the world. If the old phones were connected and powered by USB through a "real computer" with an accurately calibrated clock and an internet connection, then we could triangulate thousands of meteors per year, and characterize the near-Earth solar system. With a dense enough sensor network, we could average out errors and locate more fallen meteorites. A metal-working friend in Canada would like to manufacture the rigid, adjustable "phone mounts".

EDIT: I just realized why we probably can't do this with smart phones. For power efficiency, many smart phone operations are done with specialized hardware, including automatic image compression to JPEG. It may not be possible to access the actual uncompressed bitmap image from the image chip, which would be necessary for proper statistical processing. The image chip is probably too small and insensitive to make proper night-sky photographs, and what survived the imager would be smeared out by wavelet compression. Ah, well. That is why it is a good idea to build Really Big Survey Telescopes with CCD cameras as big as an automobile.

Re: 1723: "Meteorite Identification"

Posted: Mon Aug 22, 2016 8:53 pm UTC
by Copper Bezel
Also applies to arrow heads and dinosaur fossils.

Re: 1723: "Meteorite Identification"

Posted: Mon Aug 22, 2016 9:48 pm UTC
by somitomi
cellocgw wrote:
HES wrote: My favorite part is how 'Did someone see it fall? -> Yes' points to 'NOT A METEORITE.' This is not a mistake.


So....

If a meteorite doesn't fall in the forest where nobody can see it, is it still not a meteorite?

Ask people living in Eastern Siberia

Re: 1723: "Meteorite Identification"

Posted: Tue Aug 23, 2016 3:25 am UTC
by RogueCynic
Reecer6 wrote:However, if no one on Earth saw it fall, then its true nature of being or not being a meteorite will forever mystify us, because you clearly can't send a sample at that point.


Isn't that Schroedinger's Cat experiment?

Re: 1723: "Meteorite Identification"

Posted: Tue Aug 23, 2016 3:28 am UTC
by Soupspoon
RogueCynic wrote:
Reecer6 wrote:However, if no one on Earth saw it fall, then its true nature of being or not being a meteorite will forever mystify us, because you clearly can't send a sample at that point.


Isn't that Schroedinger's Cat experiment?


Mu

Re: 1723: "Meteorite Identification"

Posted: Tue Aug 23, 2016 8:48 am UTC
by taemyr

Re: 1723: "Meteorite Identification"

Posted: Tue Aug 23, 2016 10:10 am UTC
by Gauteamus

Re: 1723: "Meteorite Identification"

Posted: Tue Aug 23, 2016 1:59 pm UTC
by squall_line
keithl wrote:It would be wonderful if someone would write a meteor observation app for rooted, obsolete Android phones with cameras. Those could be deployed by the hundreds of thousands, watching the skies around the world. If the old phones were connected and powered by USB through a "real computer" with an accurately calibrated clock and an internet connection, then we could triangulate thousands of meteors per year, and characterize the near-Earth solar system. With a dense enough sensor network, we could average out errors and locate more fallen meteorites. A metal-working friend in Canada would like to manufacture the rigid, adjustable "phone mounts".

EDIT: I just realized why we probably can't do this with smart phones. For power efficiency, many smart phone operations are done with specialized hardware, including automatic image compression to JPEG. It may not be possible to access the actual uncompressed bitmap image from the image chip, which would be necessary for proper statistical processing. The image chip is probably too small and insensitive to make proper night-sky photographs, and what survived the imager would be smeared out by wavelet compression. Ah, well. That is why it is a good idea to build Really Big Survey Telescopes with CCD cameras as big as an automobile.


So, basically a WeatherUnderground-esque (crowd-sourced) version of NASA's All Sky Fireball Network? I like it!

Re: 1723: "Meteorite Identification"

Posted: Tue Aug 23, 2016 2:46 pm UTC
by DanD
taemyr wrote:http://www.iflscience.com/space/how-hunt-micrometeorites/


I'm pretty sure this technique will also pick up weld spatter. And pretty much any other case where you've got mists of molten iron flying around.

Re: 1723: "Meteorite Identification"

Posted: Wed Aug 24, 2016 1:09 am UTC
by Caesar
Rombobjörn wrote:
Caesar wrote:The "Did someone see it fall?" box doesn't have a "No" exit. Took me a while to figure that one out. It means that, if no one had seen it fall, they would have never thought it were a meteorite to begin with. Right?

It means that if you arrive to that box, then it's already clear that your stone is not a meteorite, and the box is there to deflect the argument that "I know it's a meteorite because I saw it fall".


That logic doesn't follow. In that case, you would expect two arrows, "yes" and "no", pointing to the same answer. There has to be something trickier going on.

Re: 1723: "Meteorite Identification"

Posted: Wed Aug 24, 2016 1:43 am UTC
by Copper Bezel
No, that's right. It's snidely making the assumption that if you got that far, the only reason left you might have been looking at this flowchart in the first place is that "someone saw it fall", and this is a silly reason.

Re: 1723: "Meteorite Identification"

Posted: Wed Aug 24, 2016 4:26 pm UTC
by Caesar
Copper Bezel wrote:No, that's right. It's snidely making the assumption that if you got that far, the only reason left you might have been looking at this flowchart in the first place is that "someone saw it fall", and this is a silly reason.


The important point is that the "No" arrow is not necessary. His comment only explains that the "Yes" arrow IS necessary. Your answer is correct, and it's the same answer I gave.

Re: 1723: "Meteorite Identification"

Posted: Wed Aug 24, 2016 6:40 pm UTC
by Copper Bezel
I'm not sure what to say beyond acknowledging that yes, that was a fairly repetitive exchange.

You: Thing?

Rombobjörn: Thing.

Caesar: Not thing.

Me: No, thing.

Re: 1723: "Meteorite Identification"

Posted: Wed Aug 24, 2016 9:59 pm UTC
by Caesar
My interpretation:

Caesar: Thing?

Rombobjörn: Thign.

Caesar: Not Thign.

Copper Bezel: No, Thing.

Caesar: Thing != Thign.


But I should probably stop. This is fruitless and we probably all mean the same thing.

Re: 1723: "Meteorite Identification"

Posted: Wed Aug 24, 2016 10:39 pm UTC
by Copper Bezel
Oh dear, both of you are you. There any claim of mine to credible reading comprehension in this exchange.

In any case, yes, I think both of you put it more accurately than Rombobjörn did, in that they left out the best bit of the trick.

Re: 1723: "Meteorite Identification"

Posted: Thu Aug 25, 2016 5:51 pm UTC
by Shadowman615
Caesar wrote:My interpretation:

Caesar: Thing?

Rombobjörn: Thign.

Caesar: Not Thign.

Copper Bezel: No, Thing.

Caesar: Thing != Thign.


But I should probably stop. This is fruitless and we probably all mean the same thing.


Or do you mean the same thign?