Clark R. Chapman Southwest Research Inst. Boulder, Colorado, USA

31
Clark R. Chapman Southwest Research Inst. Boulder, Colorado, USA Interface 2003”: 35th Symposium on the Interface 2003”: 35th Symposium on the Interface: Computing Science & Statistics Interface: Computing Science & Statistics Salt Lake City, Utah 15 March 2003 Salt Lake City, Utah 15 March 2003 Predicting and Comprehending Asteroid Impacts Invited Session: Prediction of Catastrophic Events Fertile Ground for Applications Fertile Ground for Applications of Statistics of Statistics

description

Predicting and Comprehending Asteroid Impacts. Fertile Ground for Applications of Statistics. Clark R. Chapman Southwest Research Inst. Boulder, Colorado, USA. Invited Session: Prediction of Catastrophic Events. - PowerPoint PPT Presentation

Transcript of Clark R. Chapman Southwest Research Inst. Boulder, Colorado, USA

Page 1: Clark R. Chapman Southwest Research Inst. Boulder, Colorado, USA

Clark R. ChapmanSouthwest Research Inst.Boulder, Colorado, USA

Clark R. ChapmanSouthwest Research Inst.Boulder, Colorado, USA

““Interface 2003”: 35th Symposium on the Interface 2003”: 35th Symposium on the Interface: Computing Science & StatisticsInterface: Computing Science & Statistics

Salt Lake City, Utah 15 March 2003Salt Lake City, Utah 15 March 2003

““Interface 2003”: 35th Symposium on the Interface 2003”: 35th Symposium on the Interface: Computing Science & StatisticsInterface: Computing Science & Statistics

Salt Lake City, Utah 15 March 2003Salt Lake City, Utah 15 March 2003

Predicting and Comprehending Asteroid ImpactsPredicting and Comprehending Asteroid Impacts

Invited Session: Prediction of Catastrophic Events

Invited Session: Prediction of Catastrophic Events

Fertile Ground for Fertile Ground for Applications of StatisticsApplications of Statistics

Page 2: Clark R. Chapman Southwest Research Inst. Boulder, Colorado, USA

The processes that formed the planets 4.6 billion years ago left many small remnant objects: comets (beyond the outer planets) and asteroids (in a “belt” between the orbits of Mars and Jupiter). Some of them occasionally cross the Earth’s orbit and can strike our planet...if it happens to be there at the same time.

The processes that formed the planets 4.6 billion years ago left many small remnant objects: comets (beyond the outer planets) and asteroids (in a “belt” between the orbits of Mars and Jupiter). Some of them occasionally cross the Earth’s orbit and can strike our planet...if it happens to be there at the same time.

Comets and Asteroids

Jupiter’s orbit

Sun

Asteroid Belt

Two asteroids colliding

Two asteroids colliding

CometsComets

Comets come

from fa

r beyond

Jupiter

We are Here!

We are Here!

NEOs

Page 3: Clark R. Chapman Southwest Research Inst. Boulder, Colorado, USA

The Hazard from Asteroids and Comets: Overview

The Earth encounters interplanetary projectiles, ranging: (a) tiny, harmless ones; (b) gigantic, destructive ones… (and everything in between).

The most dangerous ones are very rare but very destructive. Smaller impacts, with greater chances of happening soon, also merit practical concern by relevant public officials.

This extreme example of a natural disaster (tiny chances of happening, but with huge consequences) challenges a rational response by citizens and policy-makers.

The Little Prince

Meteorite punctured roof in Canon City, CO

Global catastrophe

Page 4: Clark R. Chapman Southwest Research Inst. Boulder, Colorado, USA

Sizes, Impact Frequencies of NEOs

DustDust

Boulder

Boulder

Build

ing

Build

ing

MountainMountain

Second

Second Week

Week

Mill

enniu

nm

Mill

enniu

nm

500,000 yr500,000 yr

100 Myr

100 Myr

Leonid meteor showerLeonid meteor shower

Peekskill meteoritePeekskill meteorite

Tunguska, 1908

SL9 hits

Jupiter 1994

SL9 hits

Jupiter 1994

K-T mass extinctor, 65 Myr ago

Smallest, most frequent

Smallest, most frequent

Huge, extremely rareHuge, extremely rare 15 km

15 km

Page 5: Clark R. Chapman Southwest Research Inst. Boulder, Colorado, USA

What Do We Know About the Impact Hazard?

How many asteroids and comets there are of various sizes in Earth-approaching orbits (hence, impact frequencies are known).

How much energy is delivered by an impact (e.g. the TNT equivalence, size of resulting crater).

How much dust is raised into the stratosphere and other environmental consequences.

Biosphere response (agriculture, forests, human beings, ocean life) to environmental shock.

Response of human psychology, sociology, political systems, and economies to such a catastrophe.

WE

KN

OW

TH

IS…

Ver

y P

oo

rly

S

om

ewh

at

Ver

y W

ell

Ver

y W

ell

I’ll be coming back to this!

Page 6: Clark R. Chapman Southwest Research Inst. Boulder, Colorado, USA

Potential Impacts of Practical Concern

Page 7: Clark R. Chapman Southwest Research Inst. Boulder, Colorado, USA

Visualize the Widely Different Impact Scenarios

Global, civilization-threatening horror (>2 km diam., 1 chance in 10,000 21st C.)

Regional catastrophe (e.g. tsunami destroys everything within few km of Pacific Rim) (300m-1.5km, 0.2% chance 21st C.)

Devastating local disaster (30m - 300m, 40% chance 21st C.)

Blinding Hiroshima-scale flash in sky (happens every few years)

Media hype, false alarm

(happens every few months) OVER KASHMIR? OVER ISRAEL? HOW WOULD THE GENERALS RESPOND?

Page 8: Clark R. Chapman Southwest Research Inst. Boulder, Colorado, USA

Tsun

ami

Str

atos

pher

ic D

ust

Risk vs. Scale of Impact

Annual fatalities peak for events near the global “threshold size”, about 2 km

Orange/yellow zone illustrates our range of uncertainties for agricultural disaster due to stratospheric dust

Tsunamic risk very uncertain

Page 9: Clark R. Chapman Southwest Research Inst. Boulder, Colorado, USA

What Can We Do about This?What Are We Doing about It?

We can use telescopes to search for asteroids and comets that might be on a collision course with Earth during this century (Spaceguard Survey)

If one is found (among all those that we can certify as not a threat), then we could mitigate (evacuate, amass food supplies, move the asteroid so it won’t hit, etc.)

Page 10: Clark R. Chapman Southwest Research Inst. Boulder, Colorado, USA

How Does a Scientifically Illiterate Public React to Asteroids?

Here’s a case of very low statistical Here’s a case of very low statistical odds, but the potential catastrophe odds, but the potential catastrophe could destroy civilization...could destroy civilization...

Page 11: Clark R. Chapman Southwest Research Inst. Boulder, Colorado, USA

Asteroid Impacts are “Chancy”

We’re all familiar with people who think they We’re all familiar with people who think they will win the lottery, who build homes on the will win the lottery, who build homes on the 100-year floodplain because last year’s flood 100-year floodplain because last year’s flood waswas the 100-year flood according to the Army the 100-year flood according to the Army Corps of Engineers…and so on. This is Corps of Engineers…and so on. This is notnot encourtaging about the possibility for rational encourtaging about the possibility for rational thinking about asteroids!thinking about asteroids!

Public Education about ProbabilitiesPublic Education about Probabilities

Page 12: Clark R. Chapman Southwest Research Inst. Boulder, Colorado, USA

(2003 UPDATE: This statistic has changed in the last few years as we have discovered most of the mile-wide asteroids and learned that those won’t strike Earth this century: now there’s a slightly better chance of getting a Royal Flush than death-by-asteroid next year!)

(2003 UPDATE: This statistic has changed in the last few years as we have discovered most of the mile-wide asteroids and learned that those won’t strike Earth this century: now there’s a slightly better chance of getting a Royal Flush than death-by-asteroid next year!)

A Royal Flush

It is more likely that a mile-wide asteroid will It is more likely that a mile-wide asteroid will strike Earth next year than that the next poker strike Earth next year than that the next poker hand you are dealt will be a royal flush.hand you are dealt will be a royal flush.

Odds:1 to 649,739

Page 13: Clark R. Chapman Southwest Research Inst. Boulder, Colorado, USA

Chances from Dying from Selected Causes (for U.S.A.)

By terrorism (mostly due to

Sept. 11th attacks)

Page 14: Clark R. Chapman Southwest Research Inst. Boulder, Colorado, USA

Source: John Pike

20th Century Catastrophes: We have much more to worry about than impacts!

Averaged over long durations, the death rate expected from impacts is similar to that from volcanoes.

(Asteroids similar)This is what we have to

worry about this month

Page 15: Clark R. Chapman Southwest Research Inst. Boulder, Colorado, USA

Fatality Rates Compared with Accidents and Natural Hazards

Page 16: Clark R. Chapman Southwest Research Inst. Boulder, Colorado, USA

1010

55

00

-5-5

-10-10

-15-15

(100

000

km)

(100

000

km)

-15-15 -10-10 -5-5 00 55 1010 1515

(100000 km)(100000 km)

The “Scary” Case of 1997 XF11

In March 1998, head-lines warned of pos-sible impact in 2028.

The next day, old data ruled it out…but the prediction was badly mistaken.

This Week is the Fifth Anniversary of the Scare

Page 17: Clark R. Chapman Southwest Research Inst. Boulder, Colorado, USA

1997 XF11 Error “Ellipse”

Original “back-of-the-Original “back-of-the-envelope” calculation reported envelope” calculation reported

in a Press Statement by the in a Press Statement by the Minor Planet Center at HarvardMinor Planet Center at Harvard

Actual Monte Carlo calculations Actual Monte Carlo calculations of very elongated error ellipse, of very elongated error ellipse, done by Muinonen from same done by Muinonen from same data available to Minor Planet data available to Minor Planet Center (it took several days of Center (it took several days of computer time)computer time)

Earth

Part of very elongated error ellipse

Page 18: Clark R. Chapman Southwest Research Inst. Boulder, Colorado, USA

1997 XF11 Error Ellipse: Details

Page 19: Clark R. Chapman Southwest Research Inst. Boulder, Colorado, USA

Returns in 2040 from 2027 Encounter with 1999 AN10

Many asteroid orbits are in physical Many asteroid orbits are in physical resonances (they do resonances (they do not not intersect intersect Earth’s orbit randomly like particles-Earth’s orbit randomly like particles-in-a-box). And they are moving in-a-box). And they are moving chaotically.chaotically.

Within an error ellipse for a particular Within an error ellipse for a particular near-miss, there are many, tiny “keyholes” near-miss, there are many, tiny “keyholes” through which the asteroid could pass that through which the asteroid could pass that would bring it back to Earth impact some would bring it back to Earth impact some years or decades hence -- so it is difficult years or decades hence -- so it is difficult to rule out future impacts.to rule out future impacts.

Page 20: Clark R. Chapman Southwest Research Inst. Boulder, Colorado, USA

1998 OX4 Virtual Asteroids for January 2046 (Milani et al., 2003)

Earth

The Problem: In 1999, asteroid 1998 OX4 passed close to Earth, but telescopic The Problem: In 1999, asteroid 1998 OX4 passed close to Earth, but telescopic positional data were sparse. positional data were sparse. Could it have passed through a keyhole that would Could it have passed through a keyhole that would aim it AT Earth in 2046?aim it AT Earth in 2046? Monte Carlo calculations show a circumstance that gets Monte Carlo calculations show a circumstance that gets uncomfortably close to Earth. We can accurately trace back the orbit of this uncomfortably close to Earth. We can accurately trace back the orbit of this “virtual asteroid”, look for it at “its” next close approach, and see if the real “virtual asteroid”, look for it at “its” next close approach, and see if the real asteroid is actually there. If not, we’re safe from 1998 OX4.asteroid is actually there. If not, we’re safe from 1998 OX4.

Dangerous V.A.

Page 21: Clark R. Chapman Southwest Research Inst. Boulder, Colorado, USA

Asteroid Size Distribution: How Often Impacts of Different Energies Happen

Page 22: Clark R. Chapman Southwest Research Inst. Boulder, Colorado, USA

Two Weeks of Near-Earth Asteroid Searching by Spaceguard Survey

What are the observational biases of these various search programs? How What are the observational biases of these various search programs? How do they affect our knowledge of the size- and orbital- distributions?do they affect our knowledge of the size- and orbital- distributions?

Page 23: Clark R. Chapman Southwest Research Inst. Boulder, Colorado, USA

In the Post 9/11 World...

Does the remote threat of an asteroid catastrophe have anything to teach us?

Page 24: Clark R. Chapman Southwest Research Inst. Boulder, Colorado, USA

NEO Impacts in the Context of Natural Hazards and Civil Defense

Impact hazard has similarities and dissimilarities compared with more familiar disasters Similarities include: nature of damage partly caused by

familiar forces (fire, high wind, quake, falling debris, flood) Dissimilarities: impacts happen anywhere; no analogs to

“aftershocks”; no radioactivity or enemy soldiers

Though a major impact could happen, it is much less likely than a familiar natural disaster For every future impact that will kill thousands to

hundreds of thousands of people, there will likely be hundreds of floods, typhoons, and earthquakes that will each kill just as many

Page 25: Clark R. Chapman Southwest Research Inst. Boulder, Colorado, USA

Public Perception

While “known” to many from movies and the news, a serious impact disaster has never been experienced in recorded history.

The tiny chances and huge consequences are extremely difficult for people to relate to.

The impact hazard is “dreadful” (fatal, uncontrollable, involuntary, catastrophic, increasing…) and apocalyptic (with religious or superstitious implications for many). Public response to a real impending impact is expected to be exaggerated (e.g. Skylab falling).

Experience with news media hype and misinformation suggests we need more science literacy among journalists and citizens in general.

Page 26: Clark R. Chapman Southwest Research Inst. Boulder, Colorado, USA

Badly Misleading News Stories in 2002 Alone

2002 EM7 came from a “blind spot,” it was a near-miss, so the Spaceguard Survey is inadequate Many NEOs are found departing. Goal never was to catch one just before impact.

Reflects basic misunderstanding of survey approach.

2002 NT7 “is on an impact course with Earth” (BBC, July 2002) and other hyperbole It was a scientifically interesting case of a very small chance of impact many years

from now; as usual, a few days of further observations reduced the chances to zero.

“Impact dangers less than we thought” (attributed to Brown et al., Nature, Nov. 2002) Brown et al. studied harmless objects 1 to 10 m in size; no implications for

“Tunguskas” let alone for the several km NEOs, which actually dominate the hazard,

statistically.

Page 27: Clark R. Chapman Southwest Research Inst. Boulder, Colorado, USA

Hazard Scales: The Challenge of Simply Communicating Risk to Citizens

The well-known Richter Scale has been refined over decades.

Americans wrestled last month with how to deal with “orange”.

The Torino Scale has had modest effectiveness, challenges.

Richter Scale (Earthquakes) Terrorism Scale Predicted Asteroid Impact Scale

Page 28: Clark R. Chapman Southwest Research Inst. Boulder, Colorado, USA

Comparisons of Terrorist Threat and Impact Hazard

Similarities Threats are new, “dreadful”, poorly understood, raise fears Few have been (or, in all likelihood, will be) killed...but many could

be killed Both strike randomly, in place and time

Dissimilarities Terrorism is consciously done (so retribution is demanded) while

impacts are an “act of God” We can do something concrete about the impact hazard (search for

the body that may impact, deflect it, or evacuate ground-zero); battling terrorism is like the “war on drugs”

We spend vast sums to battle terrorists, a few million $ annually to search for asteroids (and how much on flu vaccination programs that might save 20,000 annually?)

Page 29: Clark R. Chapman Southwest Research Inst. Boulder, Colorado, USA

Post-September 11th Insights

We’ve “learned” to fear the unexpected in what seems like an ever more risky world -- whether or not it truly is. (Asteroids appear to be an increasing danger, even though they aren’t, due to increasing “near misses”.)

“Objective” measures of death and damage (e.g. ~3000 deaths and property damage in lower Manhattan) do not begin to predict the nature of public responses and the resulting potential losses (e.g. economic recession).

“Who was to blame for not foreseeing this kind of disaster?” (we seek villains in government, for 9/11, for the Columbia disaster…who will be blamed if an asteroid strikes?)

Page 30: Clark R. Chapman Southwest Research Inst. Boulder, Colorado, USA

Two-tiered Approach to Dealing with Irrational Risk Responses

Public officials must be prepared to deal with disproportionate responses The public politically demands that they do There are real psychological, economic, and other

consequences

Politicians, educators, and science journalists must endeavor to teach citizens how to evaluate more rationally the risks that affect them Generally, fear would be reduced; rational concern

would lead to constructive response Our national and personal resources would be

employed more cost-effectively

Page 31: Clark R. Chapman Southwest Research Inst. Boulder, Colorado, USA

Current Dilemma. (FEMA has shown little interest, for example.)

What should we do to develop the institutional capability to respond to a predicted or actual impact?

What should we do to develop the institutional capability to respond to a predicted or actual impact?

Astronomers have tried to take the first step... Astronomers have tried to take the first step...