September 28, 2004

Earthquakes, and Now Asteroids

I keep telling myself that the earth's cross-section is only 1/62,500 the size of a circle a million miles in radius. And Parkview, CA, is a good 200 miles away. And the hurricanes are a good 3000 miles away. But the universe does not feel like a friendly place this morning.

Obsidian Wings: Missed.: We're going to get a visitor tomorrow: Huge Asteroid to Fly Past Earth Wednesday.

The largest asteroid ever known to pass near Earth is making a close celestial brush with the planet this week in an event that professional and backyard astronomers are watching closely.

The space rock, named Toutatis, will not hit Earth, despite rumors of possible doom that have circulated the Internet for months. Humanity is very fortunate there won't be an impact, as the asteroid is large enough to cause global devastation. Toutatis is about 2.9 miles long and 1.5 miles wide (4.6 by 2.4 kilometers).

On Wednesday, Sept. 29 it will be within a million miles of Earth, or about four times the distance to the Moon.

By the way, if something this size ever actually hits us, the odds that you or I will survive the experience will be so close to zero as to make no difference. Also, there isn't a damned thing that we could have done if Toutatis had decided to hit us, and there won't be even a chance of us ever doing a damned thing about it until such time as we have a viable space program that can meet these things somewhere rather farther away than Near Earth Orbit. In other words, our survival as a species depends pretty much on an extreme long shot never coming in. How long a shot? Beats me, but the Man in the Moon could tell you, you betcha.

Posted by DeLong at September 28, 2004 01:09 PM | TrackBack
Comments

Parkfield, not Parkview

Posted by: seismologist at September 28, 2004 01:16 PM

A better end goal than the Mars Mission for our space program?

Posted by: Waffle at September 28, 2004 01:26 PM

Odds are one "dinosaur killer" per 100 million years, give or take.

The truly, horribly, tragic thing is that even if a killer asteroid was spotted 100 years before impact, humanity would probably be consumed in bickering ("it's not a real threat!", "God will save us!", "spend the money on earth!") until it's too late to do anything about it.

And that that point, and omnicient diety would probably say "those fuckers deserved it".

Posted by: Satan luvvs Repugs at September 28, 2004 01:31 PM

Our survival as a species, feh.

I am so tired of people whose concern for the "species" only manifests itself when there's a chance to shoot big rockets at something.

Any estimates on your odds of dying from:

(a) asteroid impact;
(b) nuclear weapons;
(c) other weapons;
(d) AIDS; or
(e) gastrointestinal disease.

Now put on that Rawslian veil of ignorance and try again.

Posted by: lemuel pitkin at September 28, 2004 01:32 PM

"Any estimates on your odds of dying from:" and a list of several things.

As I understand the statistics, given my age and life style, I need to worry much more about being smashed by a giant SUV than smashed by a giant asteroid. Every time I have to merge into traffic on I-25 here in Denver, I firmly believe those statistics to be true.

Posted by: Michael Cain at September 28, 2004 01:45 PM

The point is exactly that, isn't it. YOUR survival versus that of all species.

Posted by: Aunt Deb at September 28, 2004 01:49 PM

I'd bet that the threat of asteroid impacts will get a lot more attention from conservatives than the threat of global warming.

Why? Because the threat of asteroid impacts can be used to justify funding aerospace boondoggles, whereas the threat of global warming just seems to make them think of regulatory costs.

Posted by: Jon H at September 28, 2004 02:19 PM

"Smashed by SUV" would have been a nice item to follow "smashed by asteroid." Too bad you can't edit blog comments....

Re Aunt Deb, the point is that when you see someone going on and on about the survival of the species without any evident concern for the survival of the individual human beings who make up the specieis, you gotta wonder if they aren't really interested in something else. Like, say, shooting big rockets at things.

Posted by: lemuel pitkin at September 28, 2004 02:21 PM

Um...
If you're standing in the middle of the desert, your odds of dying of heatstroke are much greater than being hit by an SUV. But if you are standing out in the desert and an SUV is barreling down on you, you don't sit around considering the odds. You get out of the goddam way of the SUV.
Likewise, if you get a positive result on your HIV test, you don't say, "That's okay, the odds of a person dying of AIDS in this country is one in 2 million".
You're more likely to be struck by lightning than to win the lottery. But if you hold a winnning ticket in your hand, suddenly the odds that you are going to end up with a lot of cash become very good - 100%, in fact (unless you walk out the door and get hit be an SUV...)
But if the Earth finds itself holding the losing ticket - that is, having an asteroid barreling down on us - then suddenly your 'odds' are out the window, because there is now a ONE HUNDRED PERCENT chance of everyone dying.
- Jeremy

Posted by: Jeremy at September 28, 2004 02:57 PM

Ignorant Question:

Why does a 2.5 mile asteroid collision end all life as we know it? What is the calculation? Weren't the dinosaurs wiped out by something an order of magnitude larger (60 Km or so in diameter)?

Not a hostile question: just an ignorant one.

Posted by: Philipw2 at September 28, 2004 02:59 PM

According to the Impact Effects Calculator, it looks like the author of Obsidian Wings is dramatically overstating the effect of Toutatis hitting Earth. Sure, if it hit in the neigborhood (in asteroidal impact terms - think "halfway across the country") things would be ugly, and if you live in a third-world country the climate change is likely to be fatal, but all-life-on-earth-dying it ain't.

Posted by: Jake McGuire at September 28, 2004 03:16 PM

Jeremy:

"But if the Earth finds itself holding the losing ticket - that is, having an asteroid barreling down on us - then suddenly your 'odds' are out the window, because there is now a ONE HUNDRED PERCENT chance of everyone dying.


Right. Except we're not actually holding that losing ticket. No barreling asteroids around. There might be one someday -- but "might," in this case, means an annual probability somewhere south of 10^-8. So to follow your lottery analogy, right now we're probably better off checking those boring old help-wanted ads than planning what we'll do with our lotto winnings.

Posted by: lemuel pitkin at September 28, 2004 03:29 PM

Jeremy:

"But if the Earth finds itself holding the losing ticket - that is, having an asteroid barreling down on us - then suddenly your 'odds' are out the window, because there is now a ONE HUNDRED PERCENT chance of everyone dying.


Right. Except we're not actually holding that losing ticket. No barreling asteroids around. There might be one someday -- but "might," in this case, means an annual probability somewhere south of 10^-8. So to follow your lottery analogy, right now we're probably better off checking those boring old help-wanted ads than planning what we'll do with our lotto winnings.

Posted by: lemuel pitkin at September 28, 2004 03:35 PM

"According to the Impact Effects Calculator, it looks like the author of Obsidian Wings is dramatically overstating the effect of Toutatis hitting Earth."

http://www.lpl.arizona.edu/impacteffects/

Fun stuff. So we see that if Toutatis hit 1,000 miles from you, which is pretty close in the scale of things, you would experience:

* no direct thermal radiation
* Standing motor cars rocked noticeably
* Glass windows will shatter.

And impacts this big occur only every 10 million years.

Sucks about the windows though.

Posted by: lemuel pitkin at September 28, 2004 03:41 PM

Just sent to a friend from the SE in response to his inquiry:

"If there were such a thing as a looney left-wing televangelist he would attribute [the Florida hurricanes] to God's wrath over the election fraud. Suffer Floridans for your corrupt politicians, suffer. (This incidentally also explains the Cali earthquake.)"

Posted by: ogmb at September 28, 2004 03:48 PM

If it shatters windows, what about ear drums?

Posted by: luci phyrr at September 28, 2004 03:48 PM

If you come looking to learn the rules or the basics of texas
holdem, check out our Poker
Basics, or our Texas
Holdem strategy section. You be able to find out both how the game is played,
and how to play the game better.

Posted by: texas holdem at September 28, 2004 03:51 PM

Lemuel P, I understand this psychological phenomenon. The inability to experience humility while suffering the effects of hubris, I think best describes it...

Posted by: Aunt Deb at September 28, 2004 04:05 PM

i think the real question is what is president morgan freeman going to do about it? and what about intrepid reporter tea leoni? answers people, we need answers.

Posted by: cali_ at September 28, 2004 04:17 PM

There's actually an intriguing twist to the question of how likely you are to die from an asteroid strike.

Consider: If an asteroid kills several billion people every 100m years, the expected number of deaths per year is in the dozens. (In fact it's a little worse than this. The peak of fatalities * probability comes from slightly smaller events.) It turns out your chances of being killed by an asteroid are comparable to your chances of dying in an aircraft accident.

Does this seem counter intuitive? Should we be spending as much on asteroid defense as avoiding air crashes?

The lesson is to beware summarising an entire probability distribution by a single number - its mean. Expectation values are convenient but it can be easy to forget that they throw away most of the information.

Posted by: JK at September 28, 2004 04:32 PM

A couple of quick points:

(1) "I keep telling myself that the earth's cross-section is only 1/62,500..."

Sadly, the effective cross-section is larger than this, though not much, since the Earth attracts the asteroid.

(2) The initial effects of such an impact would not be that huge, but the secondary effects would likely be very large.

My own run of the above tool shows total impact energy of 9.66 million MegaTons, putting it in the 'continent killer' category, and assuming the projectile is light rock.

Weather and secondary seismic events would be considerable.

If this impactor hit in the ocean, most coastal dwellers would be killed though the calculator does not give figures for the resulting tsunami size.

In either case, I estimate that a sizeable fraction of the world's people would likely die (say 10-20%), mostly from secondary effects. It would likely take several decades for humanity to recover.

NASA tends to understate the importance of these events and the only appropriations I have seen discussed seriously have been quite modest.

We could probably head off such an event quite cheaply if we had enough advance warning. It only gets really expensive if there is little warning.

Cheers,

Andy

Posted by: Andrew Price at September 28, 2004 05:02 PM

A couple of quick points:

(1) "I keep telling myself that the earth's cross-section is only 1/62,500..."

Sadly, the effective cross-section is larger than this, though not much, since the Earth attracts the asteroid.

(2) The initial effects of such an impact would not be that huge, but the secondary effects would likely be very large.

My own run of the above tool shows total impact energy of 9.66 million MegaTons, putting it in the 'continent killer' category, and assuming the projectile is light rock.

Weather and secondary seismic events would be considerable.

If this impactor hit in the ocean, most coastal dwellers would be killed though the calculator does not give figures for the resulting tsunami size.

In either case, I estimate that a sizeable fraction of the world's people would likely die (say 10-20%), mostly from secondary effects. It would likely take several decades for humanity to recover.

NASA tends to understate the importance of these events and the only appropriations I have seen discussed seriously have been quite modest.

We could probably head off such an event quite cheaply if we had enough advance warning. It only gets really expensive if there is little warning.

Cheers,

Andy

Posted by: Andrew Price at September 28, 2004 05:06 PM

Todo el mundo sabe que el gran cometa no va a venir hasta el ano dos mil y doce, lo cual es el fin del calendario de los maya. La tierra se pase por una cinta de asteroides, cada treinta mil anos ( mas o menos), durante cual tiempo hay asteroides que pegan el superficie de la tierra. Aveces son grandes, aveces no. Pero con el grande, puede rearreglar los polos y cambiar el mundo como cuando vino el diluvio...o la gran congelada...o algo asi...

Entonces, la moral es no importa quien gana esta eleccion in gringolandia, solo hay 7 anos mas para festejarnos antes el grandisisimo!

(broma)

Posted by: jose at September 28, 2004 05:23 PM

A friend saw a tv show about this that said that to blow one up safely, it has to be a long ways away from us. Several years ago, NASA threw away the plans for the Saturn 5, the only US rocket capable of doing that, leaving the Russians with the only rockets that might be able to do it.

Posted by: masaccio at September 28, 2004 05:35 PM

"If an asteroid kills several billion people every 100m years, the expected number of deaths per year is in the dozens. (In fact it's a little worse than this. The peak of fatalities * probability comes from slightly smaller events.) It turns out your chances of being killed by an asteroid are comparable to your chances of dying in an aircraft accident. ... Should we be spending as much on asteroid defense as avoiding air crashes?"

Interesting calculation. the implict logic is (1) You are as likely to be kille dby an asteroid, as in a plane crash, so (2), we should be equally worried about both possibilities, so (3) we should invest comparable resoruces in preventing both.

Couple problems here. One is that to get from (2) to (3), you need to have some idea of the marginal cost per fatality averted. i don't have the slightest idea about this either, tho, so set it to one side. But the otehr problem is that you are comparing the danger of dying from an asteroid WITHOUT any money being spent to avoid it, to the danger of dying in an aircraft accident WITH lots of money being spent to avoid it.

Logically, you should compare the asteroid danger to the counterfactual case where no money was spent to prevent plane crashes. in which case I'm pretty sure that the risk of dying in an aircraft accdient would be a lot higher.

Or, you should rephrase th question to read, "Should we be spending as much on asteroid defense as avoiding air crashes, above and beyond what we're already spending avoiding air crashes?" In which case, that amount being zero, I'd say yes.

Posted by: lemuel pitkin at September 28, 2004 05:36 PM

Jeebus, drug company spam on Brad's blog now? Almost--note I wrote almost--makes me pine for an Adrian Spittle non-sequitur.

Philipw2, a quick bit of Googling came up with this:

http://www.pibburns.com/catastro/extinct.htm

"In 1980 Walter and Luis Alvarez and their colleagues Frank Asaro and Helen Michel published an historic paper suggesting that an asteroid about 10 kilometers (6 miles) in diameter struck the earth sixty-five million years ago at the end of the Cretaceous."

Not everyone buys this theory. Here's a good page about the 1908 explosion over Siberia that is widely thought to have been caused by an meteorite that exploded in the sky.

http://www.psi.edu/projects/siberia/siberia.html

Posted by: Jim at September 28, 2004 07:55 PM

How confident are we about these asteroid collision-frequency calculations?

Seems like counting up the few craters on our tiny planets and moons might not actually reveal the rubble structure of our vast solar system, galaxy and beyond.

Are asteroids really distributed evenly enough for this kind of calculation to make any sense?

Posted by: Yesh at September 28, 2004 10:05 PM

pitkin:

I don’t think expected loss is a good metric for analyzing low probability high cost events—the “zero times infinity” problem. So I don’t think you can compare airplane crashes with asteroid collisions. You come across something similar with betting schemes where maximization of expectation fails.

Posted by: A. Zarkov at September 28, 2004 11:56 PM

online poker

Posted by: online poker at September 29, 2004 01:09 AM

Zarkov:

Well, how are we supposed to judge the risk of barreling asteroids, then? intuition?

Posted by: lemuel pitkin at September 29, 2004 07:05 AM

I'm surprised that no one has yet pointed out that
this particular asteroid is named after Asterix' patron
deity...

Posted by: Dave MB at September 29, 2004 12:28 PM

The mines, people, head for the mines. Then we can eat canned food and play cards (in the dark) for a few years.

Posted by: Anna at September 29, 2004 12:34 PM

Almost all homeowners are insured against fire. And the insurance companies make money on the transaction. Does that mean we should forego fire insurance, from an expected cost point of view?

No - we pay the insurance company to absorb the risk: the low probability, high cost possibility of fire destroying our house.
It's why we buy Lottery tickets - the small fixed cost (==risk) to chance a high payout.


Since Asteroid insurance is nonsensical ("Lloyds of London will be loaded when we go!" - Tom Leher) the other option is to spend the "insurance" money on preventative measures. How best to spend it is a different post tho.

Exercise for the reader - is it better to spend your money on earthquake insurance with a giant deductible or on retrofitting your house to survive an earthquake?

Posted by: MobiusKlein at September 29, 2004 02:33 PM