The Atlantic Monthly Death Spiral Watch (Gregg Easterbrook Asteroid Devastation Edition)

Why oh why can't we have a better press corps? The Atlantic Monthly features Gregg Easterbook, who writes:

The Sky Is Falling: The odds that a potentially devastating space rock will hit Earth this century may be as high as one in 10. So why isn’t NASA trying harder to prevent catastrophe?

If the odds that a devastating space rock will hit the earth in a century are one in ten, then the chances that we have gone...

• one millennium without a DSR hitting the earth are 0.35...
• two millennia without a DSR hitting the earth are 0.12...
• four millennia without a DSR hitting the earth are 0.014...

It's possible a devastating space rock hit the earth between eight and four millennia ago and we know nothing about it--but it's not terribly likely. It's very hard for me to believe that a devastating space rock has hit the earth since 3000 BC. We have Tunguska--and that's pretty much it[1].

That means that if you started out with a 50-50 prior probability that Gregg Easterbrook knows what he is talking about, your posterior probability that the lead of his Atlantic article is better than birdcage liner given no rock since 2000 BC is 0.0138. But we start with a lower probability than that, don't we? Gregg Easterbrook has a history, doesn't he? I would start with a prior probability that Easterbrook knows what he is talking about of one in a ten, in which case our posterior judgment, given no rock since 3000 BC, is 0.0014. If the Atlantic published an article by Gregg Easterbrook every month, we would have to wait 41 years before there was a 50-50 chance that even one of the Easterbrook articles was right.

"Odds that a potentially devastating space rock will hit Earth this century may be as high as one in 10." Feh!!

---

[1] Yes, I know that Easterbrook claims that the abnormally cold weather of 536-537 was caused by a dust cloud raised by a "space object about 300 meters in diameter hit[ting] the Gulf of Carpentaria, north of Australia, in 536 A.D." But I had thought that sulphur left in ice cores in 536-7 was strong evidence that the cause was a volcanic eruption: see http://www.realclimate.org/index.php/archives/2008/03/536-ad-and-all-that/. Easterbrook doesn't mention SO₄ concentrations in ice cores.

Comments

[bigTom]

I think it might depend upon ones definition of a DSR. Something that could cause mass casualties in a populated area, could be much smaller than Tunguska, perhaps by ten to a hundred times smaller. Such an impact would be too minor to leave much of a record (except perhaps locally). The spectrum (number of versus size) of spacerocks is heavily dominated by little ones. Of course spending large sums of money to defend against rare calamities no more deadly than (say) a major earthquake would be money poorly spent.

[Kieran]

I look forward to the emergence of a Not a Penny for Climate Change But Invest in Anti-Space-Rock Tech Now coalition. It would likely be strongly supported by the Star Wars people.

[Fan]

Apparently Easterbrook did some good reporting on the Space Shuttle about 30(?) years ago and has been living on the reputation he got ever since. In baseball, starting out well and then sputtering is called "Sophomore Slump." In the rest of the world, regression to the mean. Of course, in baseball they track your average on an ongoing basis, and if you stay slumped you get sent down. What a pity the pundit business doesn't work the same way.

[jerry]

On the other hand, he was on TOTN yesterday discussing his WSJ Op-Ed regarding how he really enjoys all the free ice cream he's been getting from the economy and wondering why Americans think we're doing poorly.

http://online.wsj.com/article_print/SB121331500809069989.html

He sounded somewhat familiar, I thought he should be given the chance to listen to save the rust belt and some others around here....

How did Easterbrook end up at Brookings? Jebus, any respect I may have had for that place suddenly went town the toilet.

[Paul G. Brown]

*ducks*

[Bloix]

Of course the odds may be as high as one in ten. They may be as high as one in one. We won't know until 2100. But we really don't care what the odds may be. We only care what the odds are. And Easterbrook doesn't have a clue about that.

[Bloix]

Of course the odds may be as high as one in ten. They may be as high as one in one. We won't know until 2100. But we really don't care what the odds may be. We only care what the odds are. And Easterbrook doesn't have a clue about that.

[Bloix]

Of course the odds may be as high as one in ten. They may be as high as one in one. We won't know until 2100. But we really don't care what the odds may be. We only care what the odds are. And Easterbrook doesn't have a clue about that.

[Bloix]

Jeebus H Christ, how did that triple-post? I typed in the secret code only once.

[shah8]

You know...The Volcano Menace offers much higher odds, if you wish to speculate on world ending events.

[David Lloyd-Jones]

Yee-Hooo!

I'm just bayesing with laughter...

-dlj.

[jerry]

"Jeebus H Christ, how did that triple-post? I typed in the secret code only once."

Yep, that's what I've been saying. Any system that makes you type in a unique identifier and then STILL lets you triple post, well that's beyond craptastic.

Most blog applications are just huge steaming piles proving once and for all that mankind really has no sense of aesthetics or quality or pride. We really deserve to be nuked from orbit.

[George Smiley]

Be very, very careful, Brad. Easterbrook's prose destroys the reader's IQ at a rate of -1 point per 10k words read.

I should know. I used to be a pretty bright guy. Then I read hisA Moment on the Earth: The Coming Age of Environmental Optimism; it turned me into a two-time G.W. Bush voter.

[christofay]

There have been so many of these death watch posts that generally I feel the sentiment is "The United States Death Spiral Watch" in 24 point helvetica bold. It's a sign of myopia at least or really this is "The United States Death Spiral Watch" that we haven't seen many posts labeled "Federal Reserve Death Spiral Watch" here as really Fed practice the last two decades has been Greenspam/Bernankespam thought. And look what it has got us, the greatest financial crisis since the Great Depression.

Why study foreign SWF funds? Well, it's a hot topic, and the foreigners have the money, so you study it now and maybe some dollars will come later from that quarter.

I would ask why not study domestic SWF funds like the New Jersey SWF that just invested in the Lehman Bros secondary offering; the one in the news right before the latest stock price swoon.

Anyways, just like Mr Magoo Greenspan himself calling it a conundrum about foreign central planners investing in U. S. central gov paper, Setser has this post title up "Central banks — not sovereign funds — are doing the heavy lifting these days; they financed much of the US deficit in the first quarter."

[christofay]

SWF is buzz. The foreign SWF which are tasked at preserving value or increasing it have already been burnt by American Innovative Finance Sphere, you know the aspect of the U. S. economy that Greenspan gets a boner about, so I don't think the foreign SWF will reappear at that corner. Domestic SWF like New Jersey's as we're Americans and slow learners didn't read the memo; they bought in six months after the Citi and Merril examples.

The story which we don't hear enough about is the Communists, Chicom especially, and the dumb Japanese continuing supporting the United States during the Bush/Cheney/Cheney/Bush Administration years. You would think the Japanese would be more conservative after their decade of buying peaked-priced American golf courses, well, we will see. Unfortunately.

[christofay]

Cheney is Miley Cyrus and Bush is Hannah Montanna. Or is it the other way around?

Invest in the Cheney SWF now via your tax dollars, all money received will be forwarded to Halliburton.

Hold yr Nose till Victory.

[christofay]

"We really deserve to be nuked from orbit."

There's no way the Iranians are going to be able to do it for a decade or two, American hater.

[low-tech cyclist]

Time to get back to the Washington Post Death Spiral. On today's front page, upper left, is an analysis piece by someone named Neil Irwin, who can't make sense of why consumer confidence is at its lowest level in 30 years.

http://www.washingtonpost.com/wp-dyn/content/article/2008/06/17/AR2008061702463.html

He doesn't mention that job creation stays down a lot longer around each recession than it used to, that incomes haven't made it back to where they were before the last recession, that the unemployment rate seems to have lost its connection to the number of Americans employed, and that there's no sense that anyone knows how high the price of oil will go, how low the value of their houses will go, or how far the dollar will fall. People know they had no leverage to get a raise during the boom, and they know it's only going to be worse during the bust.

In other words, people know their personal situations are shaky, and they're afraid the economy's fundamentals may be screwed too. That oughta do it. But Irwin doesn't see any of this.

[allan_in_upstate]

I totally disagree with Brad's analysis.
The same reasoning would rule out the possibility
that both The New Republic and The Atlantic could be
destroyed by their owners in the same decade.

[ogmb]

Easterbunny is talking about a "potentially devastating" space rock. There are people, Gene Shoemaker among them, that believe Tungusta-scale events are fairly common but rarely observed. Hence, "potentially damaging" -- the chances of such an event happening are relatively high but the potential of it creating catastrophic damage is still small. Simple scaremongering. But nevermind calculating the odds, how about this particular piece of crap turd:

"If Abbott is right, then you may be here today, reading this magazine, only because by sheer chance those objects struck the ocean rather than land."

Yeah, if a DSR hits the ocean it'll disappear with a few ripples.

http://www.sandia.gov/news/resources/releases/2007/asteroid.html

[Monte Davis]

There are persuasive arguments that above a certain impact energy, ocean strikes "couple" better to the atmosphere, transferring more energy faster, ergo more destructively.
The chemistry (e.g. sulfuric acid and sulfates) depends sensitively on local geology, whether land or sub-sea.

[Monte Davis]

There are persuasive arguments that above a certain impact energy, ocean strikes "couple" better to the atmosphere, transferring more energy faster, ergo more destructively.
The chemistry (e.g. sulfuric acid and sulfates) depends sensitively on local geology, whether land or sub-sea.

[AndrewBW]

I pretty much like the Atlantic and would not be so happy if it were to die. Easterbrook excepted -- why does anyone publish him? -- they have a lot of good writers. James Fallows does good work. I like reading Christopher Hitchens' literary articles (as opposed to his political articles in Slate). They have Ambinder and Yglesias now. William Langewiesche is good. I rarely agree with Sullivan about anything, but I enjoy him (so long as he's not writing about Hillary Clinton). And their archives are full of interesting stuff. Oh, they should get rid of the awful Sandra Tsing Loh too. And Nicholas Carr's current cover story, "I Google Making us Stoopid?" is pretty clear evidence that at least it's making him stoopid. But overall I can usually get a couple of good hours of reading out of it and come away satisfied.

[Bloix]

Low-tech cyclist is right. Shorter Irwin: the hedge fund operators are doing fine, why are the construction workers so bent out of shape?

[Christer Watson]

I typically agree with your posts about the uncritical media, but I honestly believe you've gone overboard with this criticism.

[Well, what do you think the chance of a city-killer is in the next century? What do you think the chance of a dinosaur-killer is in the next century? It's not one in ten.]

As an astronomer, I personally know that there are many people in the field who consider the danger of asteroid impact as a serious long-term threat to human civilization. For example,

http://www.latimes.com/news/nationworld/nation/la-sci-schmidt31-2008may31,0,3765620.story

You may disagree with them for good scientific reasons (some are given in posts above), but when leading scientists speak about an important issue, I believe it is the journalists job to explain their ideas as clearly as possible. Science reporting is so difficult and typically terrible, I think it's a shame that you pick out this article as an example.

[Christer Watson]

I typically agree with your posts about the uncritical media, but I honestly believe you've gone overboard with this criticism. As an astronomer, I personally know that there are many people in the field who consider the danger of asteroid impact as a serious long-term threat to human civilization. For example,

http://www.latimes.com/news/nationworld/nation/la-sci-schmidt31-2008may31,0,3765620.story

You may disagree with them for good scientific reasons (some are given in posts above), but when leading scientists speak about an important issue, I believe it is the journalists job to explain their ideas as clearly as possible. Science reporting is so difficult and typically terrible, I think it's a shame that you pick out this article as an example.

[Neal]

So this is how the son of star wars will be born?

Find an ignorant bozo that suffers from uncontrollable writers diarrhea and get them worried about meterites as opposed to football or hockey?

Find more ignorant bozos to build a groundswell of worry?

Build a defense agianst meteorites that just happens to work against Chinese satellites?

Lets have a movie about it, perhaps with heroic astronauts who struggle to blow up the approaching meterorite. Build some public support that way.

Nothing is new under the sun.

[Joe S.]

Very mild DeLong smackdown:
SO4 is not a good way to say "sulfate." If you prefer the non-English symbol, SO4-- (with the "4" a subscript and the "--" a superscript) is a bit less confusing, at least to this chemist. It took me a second or so to figure out that Brad did not mean the unstable neutral sulfur tetroxide molecule, and instead was referring to the boring old sulfate ion

[Touche...]

[harold]

The race is not to the swift nor the battle to the strong (or smart).

[kharris]

"I pretty much like the Atlantic and would not be so happy if it were to die."

My very first response was "Crap, that would have to be a really big rock..."

[NeilS]

As a geologist I think Brad's comments are exactly correct.

The odds of future hits are calculated from historic hits. Even Tugusnksa was not very devastating because it hit far away from where people live. Most metoerites do.

Either meteorite strikes are very rare or not often devastating. Take your pick.

[bcamarda]

One thing I'd thought we'd learned from all the "long tail, black swan" stuff that been going around is: we're not quite as good at predicting the actual likelihood or significance of longshots as we like to think we are. Why can't I be deeply concerned about global warming and spare a little worry for killer meteors too? Why can't we get action on both?

I'm not defending Easterbrook's middle school math skills: I'd just like the policy judgments on killer-meteor-prevention to be based on as scientific a foundation as possible, just like global warming. Rather than simply assuming: no way it could happen, because it hasn't happened recently.

Not a perfect analogy, but I'm reminded a bit of the argument that as long as people are going hungry, global warming ought to be a low priority.

[eugene]

> Either meteorite strikes are very rare or not often devastating. Take your pick.

The problem with this argument is that this is based on the assumption that the world is still the same as it was in 1908. It is not. One point is that the population is larger, so the probability of hitting a population center is larger. But the larger problem is that we now have the ability to overreact in ways unimaginable in 1908. Even a modest-sized rock (100m) would have the appearance of a nuclear strike, and possibly give someone the excuse to launch a much more devastating counter-strike. Not to mention that a 100m rock could kill very large numbers of people if it did hit a population center. The current best-guess for a 1km sized rock to hit is something like 1/5000 in the next century. But the size distribution is steep, so 1/10 for a 100m rock would not be surprising.

On Dec 25, 2004, the a-priori probability of a devastating ocean-scale tsunami was small: we had only had a couple of known local events in Indonesia. On Dec 27, 2004, the world gnashed its teeth in anger at the failure of Geologists to track and warn against the Indian Ocean Tsunami.

I'm not going to address the creditability of the Gregg Easterbrooke or his reflection on The Atlantic. I'm not sure why Brad is discounting the opinion of experts working in this field without apparently talking to either local experts or looking into the matter more himself.

[So are you saying that the chance of a devastating impact in the 21st century *is* one in ten?]

[eugene]

Full Disclosure: I am an astronomer, and I am in charge of the data analysis system for the PS1 (Pan-STARRS 1) telescope mentioned in Easterbrooke's article. The NEO threat is one, but only one, of the many science goals of the Pan-STARRS project. It is the one that gets the most public reaction.

[mic]

The fact that someone is writing in his blog about DSR and using Bayesian analysis at that seems to be somewhat dependent on us not witnessing no DST in the last year, right? Because after a DSR impact there will probably be no electricity or people caring about Bayesian analysis for a while.

[Now that's just stupid. On that definition of DSR we ought to see many, many near-DSR elements killing half the human race every millennium.]

[harold]

I guess "stupid" like David Brooks and his "regular guy" act, if Easterbrook is fronting (using religion as a mask) for big weapons contractors.

[Nathan]

"On Dec 25, 2004, the a-priori probability of a devastating ocean-scale tsunami was small"

There was a devastating tsunami in practically the same spot 100 years ago. When was the last life-ending meteor strike?

[Maynard Handley]

Jesus. 30+ comments and not a single definition of "devastating"!

If we define devastating as killing off 50% of genera, yeah, it's pretty unlikely.
If we define devastating as killing at least 2 human beings, I guess it's pretty likely.
There's a heck of a lot of range between those two.

How about, as a practical matter, given that this is what Easterbrook is famous for, we define devastating as "likely to cause more economic damage than global warming"? This gives us deaths (converted into dollars using whatever standard mechanism you want) plus auxiliary damage.
By this metric, are the odds 1 in 10? No fscking way.

[Kaleberg]

It sounds like the posterior probability of Easterbrook being a posterior is 100%.

[eugene magnier]

> "On Dec 25, 2004, the a-priori probability of a devastating ocean-scale tsunami was small"

> There was a devastating tsunami in practically the same spot 100 years ago. When was the last life-ending meteor strike?

1908, Tunguska. an unknown number of nomadic herders were killed, some names are known.

Before that, it seems to be 1490 in Shanxi, 10,000 people were killed (http://en.wikipedia.org/wiki/Impact_event).

[Very interesting...]

Also, this shows the importance of defining terms. The Krakatoa tsunami was definitely catastrophic, with ~36,000 dead, but almost an order of magnitude smaller than the 2004 Boxing Day Tsunami (230,000 dead), and much more limited in scale (affected area was in the vicinity of Krakatoa / Java). Before 2004, people did not expect a massive devastating tsunami in India and Sri Lanka caused by events in Indonesia. That event was very much the Black Swan of tsunamis. Any yet, people are now puzzled how there might not have been an Indian Ocean wide tsunami alert system (as there is in the Pacific). We seemed doomed to always put in the warning system after we have had the big catastrophe...