Cosmic disasters real and imagined
1 / 43

Cosmic Disasters, Real and Imagined - PowerPoint PPT Presentation

  • Uploaded on

Cosmic Disasters, Real and Imagined. Alan W. Harris A presentation for the Night Sky Network Astronomical Society of the Pacific and Jet Propulsion Laboratory November 18, 2010. Introduction.

I am the owner, or an agent authorized to act on behalf of the owner, of the copyrighted work described.
Download Presentation

PowerPoint Slideshow about 'Cosmic Disasters, Real and Imagined' - rosalba

An Image/Link below is provided (as is) to download presentation

Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author.While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server.

- - - - - - - - - - - - - - - - - - - - - - - - - - E N D - - - - - - - - - - - - - - - - - - - - - - - - - -
Presentation Transcript
Cosmic disasters real and imagined l.jpg

Cosmic Disasters,Real and Imagined

Alan W. Harris

A presentation for the Night Sky Network

Astronomical Society of the Pacific and

Jet Propulsion Laboratory

November 18, 2010

Introduction l.jpg

Cosmic impacts as a natural hazard are obviously real, but so infrequent that we have not experienced a verifiable disaster (the word means “bad star”) in all of recorded history. We have the capability to measure this risk at least approximately, but it is so infrequent that it can give rise to mythic and irrational beliefs. How should we respond? What should we believe?

Outline of presentation l.jpg
Outline of presentation

  • NEA population and impact rate

  • Impact damage vs. impactor size

  • The impact hazard revisited

  • Survey completion, “residual risk”

  • Disasters imagined: Neocatastrophism and Holocene impact showers

Neo population how do we know l.jpg
NEO Population: How do we know?

When a survey keeps re-detecting the same objects without finding any new ones, one can infer that the survey has found them all. Going to smaller sizes, one can estimate the fraction discovered from the ratio of re-detections to total detections. Still smaller, where there are insufficient re-detections, one can estimate the relative detection efficiency versus size, and extrapolate the population estimate to still smaller objects.

Computer simulation of surveys l.jpg
Computer Simulation of Surveys

Differential Completion

Differential Completion

(same plot, log scale)

dm = Vlim - H

dm = Vlim - H

Survey simulations yield completion vs. relative magnitude plots that are remarkably similar over a wide range of input assumptions. Thus, from such models we can determine, for example, if a survey is 40% complete at a given absolute magnitude H, it will be ~20% complete at H+1, and so forth. At extremely low values of dm, discoveries occur only very close to the Earth and the discovery problem can be treated as “particle in a box”, leading to theoretical detection efficiency proportional to 0.8dm.

Cumulative population l.jpg
Cumulative Population

Until recently, including the 2003 and 2006 NASA reports, a straight-line (log-log scale, or constant power law) size frequency population was assumed for estimating the impact frequency vs. size of impactor. This fits fine for NEOs larger than ~1 km diameter. But below that, down to the smallest size that can cause ground damage (~30 m diameter), the straight line over-estimates the population, and hence the frequency of impacts, by as much as a factor of 3 or 4.

New population estimate l.jpg
New Population Estimate

This new population estimate (done just last week and still not fully checked or released to the public or even to NASA) is not substantially different from the 2007 population estimate.

It appears to confirm the “dip” in the size range around 0.1 km, and suggests that “Tunguska” and smaller impacts are even less frequent than the earlier estimates.

Impact damage the kill curve l.jpg
Impact Damage: The “Kill Curve”

  • Below a diameter of ~30 m, the atmosphere shields us and very little ground damage occurs.

  • In the size range ~30-100 m diameter, most of the energy is released in the atmosphere as an airburst, but low enough that ground damage resembling a nuclear explosion occurs.

  • In the range from ~100 m to 1-2 km diameter, a land impact produces a crater and extensive regional damage; an ocean impact can produce a tsunami affecting a continental shoreline.

  • Above 1-2 km diameter, the greatest damage comes from global environmental effects resembling “nuclear winter”

Revised kill curve airbursts small land impacts l.jpg
Revised Kill Curve, Airbursts & Small Land Impacts

The “kill curve” used in the 2003 NASA SDT report was based on Hills & Goda (1993) impact damage estimates, shown here. Their estimates would suggest that the Tunguska impact was ~30 MT, which is higher than practically any estimate currently, and about six times higher than a recent upper limit estimate of ~5 MT, by Mark Boslough.

30 MT

Revised kill curve airbursts small land impacts10 l.jpg
Revised Kill Curve, Airbursts & Small Land Impacts

I have revised the “kill curve” from that used in the 2003 report by “sliding” the energy of impactor for a given area of destruction over by two size bins (a factor of 1/4 in mass or energy of impactor, or a factor of 0.63 in diameter of impactor) in the size range of airbursts (<100 m diameter), and blended to the original curve by a diameter of ~200 m. This shift leads to a minimum size impactor for ground damage of about 26 m diameter, or about 1 MT energy.


7.5 MT

3 MT

Slide11 l.jpg

Revised Kill Curve, Ocean Impact Tsunami

Here we adopt the inundation estimates of Chesley and Ward (2006, Natural Hazards 38, 355-374) that were used in the NASA SDT report of 2003. The 2003 report recognized that tsunamis, even on short or no notice, do not kill everyone in the inundation zone. Historically, earthquake-generated tsunamis kill about 10% or less of the population in the inundation zone, and because of smaller run-up and run-in, impact tsunamis may have an even lower mortality rate. Nevertheless, in that report, the full “numbers affected” were added in to the “fatality” numbers from other types of impact (airbursts, global climatic disasters) where the numbers really do represent estimated fatalities. While it is difficult at best, and politically incorrect at worst, to place a dollar value to lives, we tend to “value a life” at least ten times higher than per capita net worth (e.g., in the U.S., several millions of dollars per life, versus per capita net worth in the $100K range). Thus, in scoring “damage” from tsunamis in units of lives, it is probably closer to reality to count about 1/10 the number “affected”, either as the number of lives actually lost or as a proxy for property damage. Thus, in my revised estimate of the impact hazard, I have reduced the “number affected” per event by a factor of 10 to represent estimates of the number of actual fatalities.

Slide17 l.jpg

Lifetime Risk of Death from Various Causes, USA unless otherwise noted

Asteroid risks before any NEAs were discovered

Slide18 l.jpg

Lifetime Risk of Death from Various Causes, USA unless otherwise noted

Asteroid risk remaining at present level of survey

Slide19 l.jpg

Lifetime Risk of Death from Various Causes, USA unless otherwise noted

Asteroid risk remaining after next generation goal is reached

Survey completion short term warning l.jpg
Survey completion: Short-term warning otherwise noted

Impactors on terminal approach (“death plunge”) trajectories come preferentially from the anti-solar (opposition) direction or the solar direction. The half coming from the anti-solar direction have a high probability of being spotted days or weeks before arrival by current and future surveys. The recent discovery of 2008 TC3 the day before it entered the atmosphere was not a fluke, it was about the expected rate of such discoveries by current surveys.

Survey completion short term warning22 l.jpg
Survey completion: Short-term warning otherwise noted

Current surveys, Vlim 20.5, can detect anything large enough to cause ground damage about 10 days before arrival. Next generation surveys, Vlim 24.5, should be able to provide at least a month warning.

Short term warning from surveys l.jpg
Short term warning from surveys otherwise noted

Current surveys, with perhaps a bit of optimization, can have a 35-40% chance of discovering even the smallest NEAs that can cause ground damage, days to weeks before they impact, thus providing time for civil defense measures.

Richard Kowalski, discoverer of 2008 TC3, in front of the telescope used to discover it, holding a fragment of the Almahatta Sitta meteorite recovered from Sudan after it arrived on Earth.

Conclusions real hazard l.jpg
Conclusions: Real Hazard otherwise noted

  • Within a few years, if not already, we will have found essentially all NEAs large enough to be a risk of global climatic effects. We will be left with some fractional probability that even one such object remains undiscovered.

  • Mid-size impacts, mainly tsunami risk, are less frequent and probably less damaging than previously estimated. Should programmatic importance be re-evaluated?

  • In the smallest size range capable of causing ground damage, the next generation survey may find ~25%, providing long-term warning. That’s hundreds of thousands of objects, some which will present ambiguous risk for extended time – like Apophis only smaller.

  • Ground-based optical surveys could be capable of about 35-50% chance of detecting a “death plunge” object providing days to weeks’ warning, sufficient for some civil defense measures. This is the most likely scenario to actually happen – like 2008 TC3 only bigger.

Risks imagined impacts in historical or holocene times l.jpg
Risks imagined: Impacts in Historical or Holocene Times otherwise noted

Throughout history, people have looked to the skies for explanations of Earthly events, e.g. Velikovsky’s (1950) “Worlds in Collision”. More recently (1982), Clube and Napier have promoted the concept of “coherent catastrophism”, claiming that the Earth is subjected to enhanced periods of impacts, which they blame for various environmental and societal ills over recent millennia. More recently still (2006),

  • NEA population and impact frequency

  • Comet impact frequency

  • Impact showers: onset and clearing

  • Supernova triggers

  • Direct transfer of SN material

Firestone et al. have blamed supernovae with injecting large comet-like projectiles into collisions with the Earth, causing the Younger Dryas climatic disturbance and extinction of North American megafauna, among other ills of the planet. Here I examine some aspects of that claim:

Firestone s claim l.jpg
Firestone’s Claim otherwise noted

  • ~41 Ky ago, radiation pulse from a supernova explosion ~100 LY distant from the sun, possibly Geminga, although that SNR is currently ~500 LY distant.

  • ~34 Ky ago, ion and gas blast wave arrives.

  • ~16 Ky ago, another shock wave, increased comet and asteroid impacts.

  • ~13 Ky ago, multiple comet-like impacts (total equivalent to ~4 km diameter, carbon-rich body), leads to YDB, megafaunal extinction, end of Clovis culture, Carolina Bays as impact scars, etc.

Curiously, this scenario has accumulated a number of adherents, who, like the six blind men feeling an elephant, latch on to and flesh out parts of the beast in terms that agree with their favorite ideas, even if they fail to address the big picture. A very recent example is Napier (MNRAS 2010, in press), who wants to make the 13 Ky BP comet shower part of his beloved

Taurid Complex, in spite of the fact that Firestone has gone to great lengths to declare that the impacting material must come from a recent SN in order to explain carbon isotope anomalies. Other researchers can’t seem to agree on these isotope anomalies and similarly ignore Firestone’s insistence on a SN.

Nea population and impact frequency l.jpg
NEA Population and Impact Frequency otherwise noted

4 km dia.

107 years!

10-3 odds in 13,000 years

Long period comet impact frequency l.jpg
Long-period comet Impact Frequency otherwise noted

In the NASA report on NEO surveys (Stokes, et al. 2003, NASA Office of Space Science), Don Yeomans reviewed the literature on the flux of LP comets crossing the Earth’s orbit, and concluded that the impact frequency of comets of all sizes is ~2x10-8 y-1 for all sizes, and ~1x10-9 y-1 for comet nuclei 4 km diameter. This is about 1% of the asteroid impact flux.

Thus, the a priori odds that a 4 km diameter long-period (Oort Cloud) comet would have struck the Earth in the last 13,000 years is about 10-5.

Some notes on odds l.jpg
Some notes on odds otherwise noted

In the above estimates, it must be conceded that we have “framed” the discussion in terms of size of impactor (4 km) and time since impact (13,000 years) of the specific claim. Consider the analogous “claim” of the Tunguska impactor, presumed to be 3-5 MT in size, occurring 101 years ago. The impact frequency estimated for this size is once in 500 years. If we ask “what are the odds of such an impact in the last 100 years?” the answer is one in five. Two years ago, that seemed like an unusual statistic, because there was an event in the last 100 years. Today, it is not an unusual statistic, because there has not been such an event in the last 100 years. Thus, we have to allow that some level of improbability is expected for a “framed” proposition.

Returning to the case of a 4 km diameter impactor 13,000 years ago, the extremely low probability is not to say that such an event could not have happened, but the a priori probability is a measure of how extraordinary the claim is, against which one must weigh the evidence that the event actually did occur. For Tunguska, the evidence for the event is quite extraordinary, and the odds are not that extreme.

Oort cloud comet showers l.jpg
Oort Cloud Comet Showers otherwise noted

500,000 years half-life of decay from peak flux

200,000 years from impulse to peak flux at Earth

From Hut, et al. (Nature329, 118-126, 1987)

Jupiter family comet showers l.jpg
Jupiter Family Comet Showers otherwise noted

Levison and Duncan (Icarus 108, 18-36, 1994) calculate a dynamical lifetime of Jupiter Family Comets of 450,000 years. The dynamical half-life of Short-Period Comets (P < 200 y) is 17,000 years.

Asteroid Impact Showers

Bottke et al. (Nature 449, 48-53, 2007) hypothesize that the K-T impactor was a member of a “shower” due to the catastrophic disruption that created the Baptistina asteroid family. They calculate that the peak flux of impactors reach the Earth ~50 million years after the breakup event, and the decay is similarly slow.

For either of these classes of bodies, one could imagine a scenario leading to a rapid onset of one or multiple impacts, but the majority of pieces of a disrupted body (asteroid or comet) would not impact promptly, but instead the rest of the stream would take tens of thousands to millions of years to dissipate. Thus, the flux that existed 13,000 years ago is essentially the same today.

Supernova triggers l.jpg
Supernova triggers otherwise noted

The commonly invoked mechanism of triggering a comet shower is gravitational perturbation of Oort Cloud comet orbits by passing stars, galactic tide, etc. Could thermal heating, leading to gas jetting by nuclei, cause a comet shower?

No. A Type II SN at peak brightness would be about as bright as the sun at a distance of 1 light year. Thus, the heat from a SN at 100 LY would be about the equivalent of the sun at 100 AU, way too little to evaporate even the most volatile components of a comet.

How about the momentum impulse of a SN blast wave?

No. A 20 solar mass cloud of gas and debris expanding out from 100 light years away would deliver ~40 kg of mass to a 4 km diameter comet nucleus. Even at a velocity of 10,000 km/sec, this would alter the velocity of the comet nucleus by less than 0.01 cm/sec. This is at least a million times too little to trigger a comet shower.

In any case, as already pointed out, a comet shower from whatever cause takes 200,000 years to peak following a trigger, and half a million years to die down.

Direct transfer of sn material l.jpg
Direct transfer of SN Material otherwise noted

Firestone (AGU December 2009 meeting) states, “Nearby SNe … may eject dust clouds and shrapnel that can impact the Earth”; and further, “There is no mechanism … unless the impacting object came directly from a recent nearby SN…”

If all ~20 solar masses of a SN were ejected isotropically as dust from 100 light years away, the Earth would intercept ~4x105 tons of that mass, or a ball about 100 m in diameter.

If the entire 20 solar masses were condensed somehow into 4 km diameter balls, the odds that even one such “comet” would strike the Earth is ~10-5, about the same as the odds that a regular comet of that size would happen to strike in the last 13,000 years.

To reach the Earth from the distances and in the times suggested, such an interstellar comet would have to arrive at >1,000 km/sec. Atmospheric entry dynamics at such speeds are not well (at all) studied, but it seems unlikely to be gentle. The energy of a 4 km diameter object at that speed is about the same as that of a 50 km diameter asteroid or comet at ordinary interplanetary speeds – enough to likely sterilize the planet.

Summary ydb impact hypothesis l.jpg
Summary: YDB Impact Hypothesis otherwise noted

  • Asteroid impact, present flux: 10-3 odds.

  • Comet impact, present flux: 10-5 odds.

  • Asteroid or comet showers: time scale is wrong; if we were in a shower 13,000 years ago, it would still be with us.

  • SN triggers: insufficient energy/momentum, same time problem as showers.

  • SN direct transfer: 10-5 odds, plus arrival velocity, energy.

Slide35 l.jpg

“There is a long tradition of catastrophist ideas, going back to the biblical flood and Plato’s story of Atlantis. Philosophically, many people prefer the idea that humans have not had much effect on the planet, either 13,000 years ago or today – better to blame thunderbolts from the gods than to accept responsibility for our stewardship of Earth.”

Skeptical Inquirer34, No. 3, 14-18, 2010

Slide36 l.jpg

“The fault, dear Brutus, is not in our stars, but in ourselves.”

— Shakespeare’s Julius Caesar (I, ii, 140-141)

Slide38 l.jpg

Lifetime Risk of Death from Various Causes, USA unless otherwise noted

Asteroid risk remaining at present level of survey

Slide39 l.jpg

End of Presentation otherwise noted

Some back-up slides follow.

Cumulative population40 l.jpg
Cumulative Population otherwise noted

The 2003 NASA SDT report used the constant power law model shown here as a straight dashed line. Even then actual population estimates suggested a factor of 2 or so lower population in the tens to hundreds of meter diameter range. My 2007 estimate (blue circles) brings this even lower by as much as another factor of 2. Thus, while the population in the >1 km diameter range is essentially unchanged, the frequency of impacts in the small size range is much less than assumed in the 2003 and 2006 reports

Impact hazard revisited41 l.jpg
Impact Hazard Revisited otherwise noted

Impact hazard revisited42 l.jpg
Impact Hazard Revisited otherwise noted

The impact hazard in perspective l.jpg
The Impact Hazard in Perspective otherwise noted

The table below summarizes the impact hazard by class of risk, and total risk, in units, F, “fatalities per year”, worldwide. We list the intrinsic risk, before any NEOs were discovered, for both the straight-line power law population used in the 2003 SDT report and for the new population. For the latter, we then list the “residual” risk at the current level of completion, and finally the residual risk that should remain after the next generation survey goal is achieved (90% completion to 140 m diameter). For each of these four cases, we list the risk assuming the SDT report “kill curve” (F) and the revised “kill curve” (F’) with tsunami risk reduced by a factor of ten, to correspond to actual expected fatalities, and the extended range of ground damage down to ~26 m diameter.