“Don’t blame yourself. The apocalypse wasn’t your fault. Actually, it was just as much your fault as it was anyone else’s. Come to think of it, if you’re an American, it was probably about 80-90 percent more your fault than the average human. But don’t let that get you down. It wasn’t exclusively your fault. Unless you’re the president. Then it might be your fault. But you’ll have plenty of interns to tell you that it wasn’t, so you’ll be fine.” -Meghann Marco

Nothing gets a scientist in the press quite like telling everyone that we’re all gonna die. Remember when there was talk of creating black holes that destroy the world at the LHC?

Remember the hype that the world’s going to end on December 21st, 2012?

Remember the supposed end-of-civilization-as-we-know it from Y2K? (Actually, younger readers may not!)

As you may have guessed, none of these things either has destroyed the world or is likely going to. These stories are always out there, and they’re almost always full of gross misinformation. One of the public services that science (and scientists) often perform is getting the correct information out there, out of the hopes that this will allay the hysteria and the wasted resources that these scares cause.

But what do you do when it’s one of your own who causes scares like this?

I’m looking squarely at you, asteroid- and comet-hunters. Remember, a few years ago, there was speculation that there was a 1 in 37 chance that the large asteroid, Apophis, would hit the Earth with devastating results?

Of course, now that we’ve taken better data, that chance is now down to less than 1 in 100,000. And this seems to be a recurring theme in popular media; I recall watching the worst episode ever of History Channel’s “The Universe” on comets and asteroids destroying the Earth.

To what should be no surprise, a new asteroid, relatively large in size, with poor data mapping its trajectory, is in the news. I’ve gotten to read headlines like:

  • Asteroid Could Threaten the Earth in 2182,
  • Giant Asteroid Headed Towards Earth; Could Level London, and
  • Will an Asteroid Kill Us All in 2182?

The reports are that there’s a 1-in-1,000 chance that we’re going to get hit. And I get it. I get why this is so scary. After all, a giant asteroid impact is what wiped out the dinosaurs 65 million years ago. If that happened to Earth today, just as surely, we’d all be goners.

Here’s my problem with this type of reporting, both by journalists and by scientists. Remember the story of The Boy Who Cried Wolf?

It only works if the boy cries wolf when there’s actually a wolf! Despite what’s being reported, the odds are almost certainly much, much safer than “one-in-a-thousand.” But you’d never know that from reading the news. Take Apophis, for example. The odds were reported at 1-in-37 because the data was poor at that point, the asteroid’s trajectory was very uncertain, and solid conclusions couldn’t be drawn. It wasn’t until better data was taken and its trajectory was better determined that we could come up with a realistic estimate. And we did, and it was more like 1-in-250,000. But this means Apophis never had a 1-in-37 chance of hitting us, we just didn’t have enough data to know better.

Does this latest story feel like déjà vu? It should. When you tell stories like this, giving 1-in-1000 odds of our destruction I don’t believe you. I’ve seen the pattern of bad reporting, and I assume — correctly — that you give “scary” odds because your data isn’t good enough to make a realistic prediction. (Guess what? The data for this new asteroid, 1999 RQ36, isn’t very good yet.)

And you do it, alarming the public, every couple of years. And when more data comes in, we find out — every time so far — that the object is actually going to miss the Earth entirely by many thousands (or tens of thousands, or hundreds of thousands) of kilometers.

So STOP.

We’ve got lots of asteroids, and we need to learn to track them, because when a potentially devastating one actually does wind up on a collision course with Earth, we’d like to be ready. And even though “extinction” events (like the one that wiped out the dinosaurs) only happen once every hundred million years or so, smaller, substantial impacts that cause significant damage happen thousands of times more frequently.

But by causing unfounded hysteria like this, based on inconclusive data, you make all of us look bad. Worst of all, you add to this awful misconception that science doesn’t know what it’s talking about most of the time.

And that’s not true. Science knows what it’s talking about. But loudmouth scientists (and science journalists) who cry “wolf!” (or “Asteroid!”) harm us all, and I’m speaking out against it. We’re not all going to die, not from this asteroid, not anytime soon.

And your great-great-grandkids probably won’t either; give us a few more years to better track this asteroid and we’ll have an accurate, meaningful estimate of the odds for you then. But until then, please don’t lose any sleep over it.

Comments

  1. #1 Nathan Myers
    July 29, 2010

    That’s easy for you to say. If you’re wrong, you’ll be dead and won’t care. Even if you do care about that now, we’d be dead too and couldn’t blame you. Of course, with everybody dead, there’s nobody left to care one way or the other, so I suppose it doesn’t matter. OK, you’re right.

  2. #2 Nathan Myers
    July 29, 2010

    One other thing: Y2K was real. Billions of dollars were spent on real, mostly dull work to keep the billing and accounting systems from going all to hell come 2000. That you got your paycheck deposited right on schedule was because those billions were spent, and the work was done. I suppose the ozone hole was a hoax, too? And acid rain? Some people say so. Some people are already saying that global warming, mass extinctions, worldwide famines, and world war will turn out to have been a hoax. We can only hope they will have the opportunity to insist on that after the fact because we fixed things, but I’m not holding my breath.

  3. #3 crd2
    July 29, 2010

    Great post.

    I was browsing The Silicon Valley Lectures. Dr. David Morrison’s lecture mirrors this topic so I’d like to share it.

    This is the most recent posted lecture and will be the 1st one on the list. Its streaming audio with no video so there is no buffering pauses which is nice. I cut/pasted the summary below.

    Dr. David Morrison (NASA Lunar Science Institute & SETI Institute): “A Scientist Looks at ‘Doomsday 2012′ and the Rise of Cosmophobia”
    April 21, 2010

    Many people have heard the rumors that the world will end in 2012 — and that some astronomical event or alignment is to blame. Dr. Morrison discusses the public fears and how they have been enflamed by the media. He sets our minds at ease, showing why there is no reason to worry more in 2012 than any other year.

    Listen @: http://www.astrosociety.org/education/podcast/index.html

    There are quite a number of other lectures that cover all sorts of topics including dark matter and dark enegery. I enjoyed them all very much and I’m sure some of you will as well. Keep up the good work Dr. Siegel!

  4. #4 Jonathan Lee
    July 29, 2010

    On technicalities, Apophis only got to 1-in-37 because we determined it’s trajectory quite well. For a rock that will do a close approach, as you bring the error down the impact probability goes up, initially.

    It is *plain wrong* to assert that 1-in-n odds should not be reported. After a 1-in-n assessment of a rock, statistically n-1-in-n such rocks will, with better data, be found to miss. 1-in-n of them will be found to hit. The claims are meaningful, assuming you accept the model assumptions (normality in noise).

    On the more general point, you seem to be claiming that existential *risks* should never be reported – only certainties. Why? What is it that a priori makes this one class of statement about the future a protected one?

  5. #5 CaptainBlack
    July 29, 2010

    You write:
    “Despite what’s being reported, the odds are almost certainly much, much safer than “one-in-a-thousand.” But you’d never know that from reading the news. Take Apophis, for example. The odds were reported at 1-in-37 because the data was poor at that point, the asteroid’s trajectory was very uncertain, and solid conclusions couldn’t be drawn. It wasn’t until better data was taken and its trajectory was better determined that we could come up with a realistic estimate. And we did, and it was more like 1-in-250,000. But this means Apophis never had a 1-in-37 chance of hitting us, we just didn’t have enough data to know better.”

    You do realise that this is nonsense don’t you?

    Any asteroid will either hit or miss within a particular prediction horizon. So as you refine the calculations based on better data the probability associated with the prediction horizon that you get should eventually approach 1 or 0.

    Your post seems to imply that the impact probabilty of Apophis always was “really” more like 1-in-250000, and that the 1-in-37 was at the time it was calculated in error.

    If the probabilities are reported correctly (and I doubt that they are, and if you like I could factor that doubt in) then ~2.7% of asteroids with similar predictions will impact. It does not matter that when further data is collected our estimate of the impact probability falls.

    CB

  6. #6 vagueofgodalming
    July 29, 2010

    The Reverend Bayes demands an apology.

  7. #7 kai
    July 29, 2010

    But we are all going to die. Just not necessarily all at the same time…

  8. #8 Birger Johansson
    July 29, 2010

    Hmm…the Chixchulub impact occurred just after the Deccan Traps magma flood event already had pushed the global biosphere to the edge of stability, so not even an asteroid (of that size) can do that much damage unassisted.

    For a fun take of asteroid deflection, read the one-page short story at Nature magazine: “Goliath; It’s all about timing” (Vol. 460 pp 298, 9 July 2009)

    Another astronomical tidbit: “Astronomers Find Planets in Unusually Intimate Dance around Dying Star” http://www.physorg.com/news199594786.html

  9. #9 Birger Johansson
    July 29, 2010

    I forgot the link. Here we go:
    http://www.nature.com/nature/journal/v460/n7252/full/460298a.html
    “Goliath; It’s all about timing” (Vol. 460 pp 298, 9 July 2009) by Bruce W. Ferguson.

  10. #10 Benjamin Franz
    July 29, 2010

    Sorry dude. You usually post good stuff, but this time you just messed up: You are mis-using statistics. There is no such thing as the “real” probability of an impact the way you are using it. There is only the probability given the data we have. The probability is an expression of our ignorance. The probability of 1 in 37 was no less “real” than the probability of 1 in 250000. They are both expressions of our current ignorance.

    Ultimately an asteroid will either strike or not strike – and as the potential events get closer the probability will either rise to 1 or fall to 0 as our ignorance decreases. It is perfectly possible (just not currently likely) that that 1 in 250000 will become 1 in 1 a few years before the close approach. That would in no way means the prior probabilities were in any way wrong.

  11. #11 Fred Magyar
    July 29, 2010

    Standby for Glen Beck and Rush Limbaugh reporting on hackers accessing email between astronomers researching asteroids. Run for your lives, Asteroidgate in 4, 3, 2, 1 …. >;^)

    Hey, we’re already pretty much fucked even if we don’t get hit by an asteroid! Go talk to the guys over at RealClimate. Now that’s some *REAL* scary shit!

  12. #12 Rory Kent
    July 29, 2010

    I was going to type a comment, but #4 (Jonathan Lee) expressed my views better than I ever could.

  13. #13 Florian Freistetter
    July 29, 2010

    (Ok – I try it again without the links)

    “When you tell stories like this, giving 1-in-1000 odds of our destruction I don’t believe you. I’ve seen the pattern of bad reporting, and I assume — correctly — that you give “scary” odds because your data isn’t good enough to make a realistic prediction. (Guess what? The data for this new asteroid, 1999 RQ36, isn’t very good yet.)”

    Sorry – but calculating impact probabilities *is* science and you do not have to believe anything. You can just read the paper or do the calculations for yourself. And the “pattern” you speak of is just the way this stuff works: Someone finds an asteroide and preliminary calculations show that there might be a collision somewhere in the future. One cannot say for sure because there are not enough observations so one assigns a probability. The news is then spread through the community; more observations come in; better calculations can be done and in most of the cases the impact probability goes down. To say that the scientists deliberately publish “scary” predictions is nonsense (and a little bit offense to all people working in this field!). They give exactly the probabilities that can be calculated according to the data available. Should they hide scientific data just because some stupid newspapers will cry doom?

    And the whole story of 1999 RQ36 is not really news. The paper with the probabilities was published in February 2009 (by Andrea Milani et al; its on arXiv). Since then, there were no new developments. It just seems that some spanish news service published an interview with one of the researchers some days ago. And of course now every blog and newspaper is spreading the “killer-asteroid” story…

  14. #14 jj
    July 29, 2010

    wow. For a smart guy you sure misunderstood the statistics of this. I don’t think I have anything to add to earlier comments, but I’ll restate: in one sense there is no probability an asteroid will hit, it either will or it won’t. Probability describes how likely it is given our current knowledge, and assuming some distribution of outcomes for the unknowns. 1-in-1000 is a perfectly reasonable and rational estimate! If you have beef with the methodology then speak up, otherwise you’re just griping.

  15. #15 Chicken Little
    July 29, 2010

    We’re all screwed!

    http://weareallscrewed.org/

  16. #16 Frederik Douglass
    July 29, 2010

    If you live a nice, long, lifetime, you’ll hear of an asteroid that’s got a 1-in-37 or so chance of hitting us, oh, I’d guess, about 37 times. And yet, NONE of those actually destroy the earth. Ethan’s right; it’s hype.

    How do we know to be skeptical of “1-in-37″ odds? Because major earth-destroyers aren’t once-in-a-lifetime events, they’re once-in-an-era events. But crying wolf happens every day.

    When the weather forecast says 30% chance of rain, that means that, under conditions like these, rain follows about 30% of the time. When an asteroid forecaster says one chance in 37 of destroying the earth, that means that they want to be quoted in the news.

    OR, just maybe, they’ve got some INCREDIBLE data. But then, why aren’t they publishing?

  17. #17 Thomas Neil Neubert
    July 29, 2010

    Maybe chicken little isn’t right every time; but if we like to worry, there are a lot of things to worry about all of the time. The statistics only help us to worry better or worse, depending upon your point of view.

    But as #14 says, something “either will or it won’t” happen.

    Personally, I’m more interested in whether some CERN experiment will hit a hypothetical Higgs boson; than whether a super asteroid will hit the Earth.

    Politically, one headline reads, “NASA asteroid-tracking program stalled due to lack of funds.” Now, how does someone scare up more funding for their asteroid research?

  18. #18 Sili
    July 29, 2010

    kai beat me to it.

    And while it’s annoying, I understand why scientists do try to communicate this information to the public. The search is desperately underfunded and there is a real risk in the very long run. But it’s hard to get people to care about what may happen in the next thousand years, when they don’t even care about the next two decades.

  19. #19 Ethan Siegel
    July 29, 2010

    I think many of you have missed my point about probabilities, so here’s a little quantitative food for thought. A typical asteroid is much smaller than the Earth, which has a cross-section of about 127 million square kilometers.

    Many of these asteroids come within a few 100,000 km of the Earth. Let’s say we’ve got one that — according to our best estimates — will miss the Earth by only 100,000 km. But let’s also say that we are uncertain of its trajectory by about 200,000 km. That means you can assume that the asteroid will be somewhere in an area of pi*(200,000 km)^2 when it passes Earth.

    Divide 127 million square kilometers by that area, and do you know what you get? That will give you 1-in-1000 odds that this asteroid will hit us.

    And here’s the problem. 1-in-1000 is a reflection of your uncertainty, not a reflection of actual probability. It’s why you need good data before drawing a believable conclusion. Apophis’ 1-in-37 odds were not a reflection of the asteroid’s trajectory; they were a reflection of our ignorance.

    That is not news. You can do science with that information, to be sure, but you need to be really responsible, or you’re going to wind up creating hysteria because people don’t understand the difference. And that was my point.

  20. #20 Thomas Neil Neubert
    July 29, 2010

    OK, I think I now understand your point about probability, uncertainty and ignorance.

    But I disagree that scientists are “causing unfounded hysteria”. My friends and acquaintances aren’t even remotely worried about being hit by an asteroid. Rather they are indifferent to asteroids except for entertainment value, i.e. a good joke or movie.

    Hence, the gross revenues from movie Armageddon 1998 was $561 Million; meanwhile public funding for asteroid tracking isn’t getting popcorn money.

    “Arecibo was set to make radar measurements in January 2013 that could help rule out an impact by Apophis, but due to funding cuts, it might need up to $3 million extra per year to continue.” March 25, 2010. http://www2.macleans.ca/2010/03/25/asteroid-tracking-hurt-by-funding-cuts/

  21. #21 Michel
    July 29, 2010

    But but I love to post these things on museumofhoaxes forum
    And I will keep posting these end of the world things. It´s such a laugh.

  22. #22 TBRP
    July 29, 2010

    But let’s also say that we are uncertain of its trajectory by about 200,000 km. That means you can assume that the asteroid will be somewhere in an area of pi*(200,000 km)^2 when it passes Earth.

    Divide 127 million square kilometers by that area, and do you know what you get? That will give you 1-in-1000 odds that this asteroid will hit us.

    Wouldn’t that yield an accurate probability that is pretty unambiguous? I mean, at that point we’d expect if 1000 asteroids whipped by with that exact same set of orbital elements and uncertainty level, that one would probably hit.

    And if we were more uncertain (location uncertainty>200000km) then that would actually yield a lower probability of impact, right? So I guess I’m having a hard time seeing what is being misrepresented. Where’s my math/reasoning wrong?

  23. #23 Jonathan Lee
    July 29, 2010

    What do you mean by “actual probability”? Unless you honestly think that celestial mechanics is unknowable, there either will be an impact or there will not be. The true “probability” is 1 or 0.

    So *obviously* your posterior probability of impact is a reflection of your ignorance; if you had perfect knowledge you’d assess to 0 or 1. This doesn’t make your posterior beliefs any less valid. As vagueofgodalming (#6) noted, you owe Bayes an apology. On the original reports of Apophis, there was a 1-in-250000 or so chance; it escalated because we got more data and that data didn’t push the best trajectory estimate off the earth.

    Your statements about posteriors are just plain wrong. Given 37 Apophis-like sets of observations, there will likely be one impact.

    As to hysteria, what magically distinguishes existential risks from any other kind? Why do statistics become unreportable when playing roulette with a large rock? In this case the numbers were actually reported. Drop the numbers, and you will desensitise people; every 1-in-10^6 rock will get played up. See LHC black holes for an object lesson here.

  24. #24 Ethan Siegel
    July 30, 2010

    If you look historically at asteroid impacts on Earth, you find that on average, we get a km-sized rock running into us about every 500,000 years. Larger ones (5 km or above) happen every 10 million years or so, and huge ones (10 km or above, like the one that wiped out the dinosaurs) happen every 100 million, give or take.

    Apophis and this new one, 1999 RQ36, are about half a km in size. According to historical records, we can expect an impact from an asteroid that size every 50,000 to 100,000 years.

    An impact that could level a city, at worst.

    If you take the estimates that these papers have come out with seriously, you’re talking about large impacts happening about 100 to 1000 times more frequently than they actually happen. It isn’t a question of whether their methods are correct, their statistics are computed correctly, or anything else. If you look at what they predict and what actually happens, they don’t match up by a large amount. You’re free to do what you like with that information, but for me, it means I’m going to speak out against these misleading reports that don’t accurately reflect what happens to the Earth.

  25. #25 vic
    July 30, 2010

    I think that any object large enough to create an explosion, either by hitting the atmosphere or the ground, that passes within the moon’s orbit should be slightly alarming: and watched and warned about without statistics. When a hurricane is about to hit land people don’t yell out statistics, they give an approximation of where it could hit and sound and alarm. It is good for the people to know, “hey this astroid’s orbit passes pretty close we need money to watch this thing.” All it takes is the right revolution and we get hit. Also it takes longer to prepare for an asteroid impact than any other natural disaster. It could take until 2100 to prepare our great great grandchildren for that disaster. I’m sure they would appreciate it. I understand what is said about the statistics being misleading though. I don’t think that media should intentionally cause hysteria. Good argument either way.

  26. #26 Birger Johansson
    July 30, 2010

    Off-topic, but quite cool (from New Scientist):
    “Every black hole may hold a hidden universe” http://www.newscientist.com/article/mg20727703.000-every-black-hole-may-hold-a-hidden-universe.html
    “Shields up! Force fields could protect Mars missions” http://www.newscientist.com/article/mg20727701.300-shields-up-force-fields-could-protect-mars-missions.html

  27. #27 Emlyn
    July 30, 2010

    And its funny how American Republican rubes reactively deny pretty much anything that science has to offer when it might in some way contradict their God-country-flag-freedom doctrine, but who still routinely use this very same pathetic fear-mongering come election time.

    Terrorists!? Vote Republican! Skyrocketing unemployment!? Vote Republican! Asteroids!? Vote…er, Republican!

  28. #28 K/T event
    July 30, 2010

    So STOP. Won’t happen: the “asteroid hunters” get funding by promulgating fear.

  29. #29 rob
    July 30, 2010

    you want something else to worry about besides rocks from space? try rocks from below!

    check out http://ashfall.unl.edu/ashfallstory.html

    it is a state park in Nebraska where lots and lots of fossilized animals were discovered covered by several feet of volcanic ash. volcano? in Nebraska? say what? there ain’t no volcano in Nebraska! the actual volcano was hundreds of miles away–in Idaho.

    the volcano erupted 12 million years ago and dropped ash all over. you may be saying to yourself “there ain’t no volcanos in Idaho!” true. however, after 12 million years, due to plate tectonics, the volcano has moved. today you would know it as Yellowstone park.

    that’s right! the whole FREAKIN’ park is a FREAKIN’ volcano! the caldera is something like 40 miles across. you can only see it from aerial photographs.

    geologic evidence shows it has erupted many times in the past, averaging about every 600,000 years. sometimes it is a giant explosion and sometimes it is “just” massive lava flows, kinda like the Deccan Traps.

    so when was the last eruption?

    about 600,000 years ago.

  30. #30 daedalus2u
    July 31, 2010

    The problem is not how risks like this are reported, it is how they are dealth with. The commenters have pointed out the defects in your analysis from a scientific view point, I will point them out from a psychological viewpoint.

    What are the changes that global warming will do bad things to the Earth? A lot higher than 1 in 37. What is being done about it? Little or nothing. Certainly nothing that will be effective.

    If society can’t do something effective to counter the existential risk of global warming, the problem is not with the science or with how it is presented. The problem is with the denialists and those who would profit by allowing disasters to happen.

  31. #31 MadScientist
    August 3, 2010

    Oh, you don’t have it terribly bad yet. You don’t have the equivalent of the Climate Modelers – now THEY make the whole global warming research community look really bad by making absurd claims like: “There will be more frequent and more powerful typhoons. It is difficult if not impossible to prove that the frequency and intensity of typhoons are increasing” – in short, pseudo-scientific claims. In any other field if you tried to pass something off as fact and told your colleagues “it is difficult if not impossible to prove” the kindest thing they’d do is call you a moron and a crackpot.

  32. #32 MadScientist
    August 3, 2010

    @daedalus #30: Where does that “1 in 37″ come from? I can’t imagine how anyone puts odds on such a thing.

    Predicting effects is generally tough because we cannot readily manipulate large ecosystems and thus any information has to be drawn from analogous events in nature. In rare instances you have relatively easy predictions – for example, with regards to failure of certain crops due to temperature related flowering behavior.

    Now if hypotheses are correct that rainfall and even ocean circulation can be affected, that will be disastrous indeed.

    I agree that nothing is being done to curb CO2 rise due to human activities and personally I don’t expect anything to be done in my lifetime. I advocate better and more numerous observation systems in the hope that scientists years down the line improve their understanding well enough and can use the data to come up with reliable predictions.

  33. #33 MadScientist
    August 3, 2010

    @Florian: However, for events like meteorite/asteroid impacts which do not have a predictable interval between events (for example, Gaussian distribution around 100 years with standard deviation of 20 years), claims like “we are hit by a big one every 1M years” is meaningless. Well, not really meaningless, but you certainly cannot draw any conclusion other than “in the past we were hit by a big one every 1M years” – which is absolutely irrelevant to the future.

  34. #34 MadScientist
    August 3, 2010

    @#16: Nope, you have a very common but absolutely wrong idea of statistics. Even if reported odds were 1/37 and you had 100 such reports and no impact you cannot conclude that the guys reporting the odds were wrong (but you can calculate the odds that they were right and nothing happened as well as the odds that they are wrong).

  35. #35 higginsdj
    September 23, 2010

    Put the blame where it is due…… WE do not report what we discover in the popular press, journalists do. Scientists deal in facts and figures, we report the facts and figures AS THEY ARE known to us at the time they are reported.

    If you think the reporting should wait – tell us when you would like to know? The day before an impact?

  36. #36 tiffany willington
    May 21, 2011

    no one wood be alive because it is Millin years in tilt that
    wood happen every one noses that

  37. #37 Ms Trixie
    March 13, 2013

    As a scientist, working in a field where I do a lot of outreach, I agree that people are blowing things out of proportion. However, people need to be informed that there is a possibility that an asteroid can hit Earth and cause catastrophe. It has already happened. Millions of years ago, an asteroid hit the Earth and wiped out the dinosaurs. Panicking about these dangers is not going to get us anywhere but neither is ignoring them. The public just needs to be informed.