This article has been accepted for publication in Astronomy & Geophysics ©: 2015 Bill Napier, David Asher, Mark Bailey and Duncan Steel. Centaurs Published by Oxford University Press on behalf of the Royal Astronomical Society. All rights reserved. http://astrogeo.oxfordjournals.org/content/56/6/6.24.abstract Centaurs as a hazard to civilization

Assessments of the risk posed small bodies from the main asteroid belt. asteroids predates the discovery of the vast The timescale for NEO supply from such a trans-Neptunian and centaur populations, by near-Earth objects ignore source is tens of millions of years, leading and the finding that centaurs, in chaotic the possibility of a giant comet to an expectation that the population of orbits, leak from the outer solar system into entering the inner solar system. NEOs does not fluctuate substantially over the inner planetary region at a significant Bill Napier, David Asher, such intervals. Thus crater-counting on rate, largely via the Jupiter family of comets. planetary surfaces, and the understanding A new, self-consistent picture has emerged Mark Bailey and Duncan Steel of what such counts imply, is based on an in which there is a gradual pass-me-down

examine the likelihood and assumption that such craters are produced of substantial objects (50–100 km across and Downloaded from potential consequences of the at a constant rate over geological/astro- above) into the inner solar system (Horner nomical time; but this core assumption is et al. 2004ab, Steel 2014, Napier 2014, 2015). appearance of such a centaur. contraindicated by the terrestrial cratering Napier (2015) discussed how the occa- record (the only craters for sional arrival of a massive

o we live in dangerous times? The which we have individual “The idea that impacts centaur will dominate the http://astrogeo.oxfordjournals.org/ risks to civilization from impacts datings), which shows that can affect our planet’s supply of Earth-crossing Dby asteroids and comets have been these impacts on our own climate and life is now debris. The size distribution appreciated only in the past few decades. planet have been largely established” of centaurs is shallow (Bauer Programmes such as NASA’s Spaceguard episodic in nature. That is, et al. 2013), i.e. more of the observations seek to map near-Earth it seems that the population of NEOs has mass is in the larger objects. These being objects (NEOs) as a way to quantify the varied substantially across time. The ques- comet-like bodies, we may assume they are risk to Earth. But does the current count tion arises whether the presently observed subject to wholesale fragmentation. This of NEOs reflect the population over time? NEOs, at all sizes, are characteristic of a leads to an expectation that when a centaur We argue that the population is variable long-term average, or whether we are living arrives in an orbit with perihelion within

and that assessments of the extraterrestrial in dangerous times, at a time of significant Jupiter’s heliocentric distance there will be by guest on November 20, 2015 impact risk based solely on near-Earth enhancement in the NEO population. a temporarily enhanced mass of material asteroid counts underestimate its nature in the inner solar system, spread across and magnitude for timescales of order Perpetual vigilance all sizes from dust at <100 microns to tens 10 000 years, i.e. the interval of interest It is clear that no survey of Earth-approach- of kilometres across. The transition rates and concern to our civilization. A more ing asteroids lasting just 10–20 years can into the near-Earth environment indicate variable but significant threat comes from possibly lead to the identification of over that fluctuations in the mass of near-Earth centaurs, giant comets derived from the 90% of all Earth-approaching asteroids interplanetary material may amount to trans-Neptunian region that reach the if there is a significant fraction in larger, two orders of magnitude over a timescale inner solar system generally via short-term, intermediate- or long-period (i.e. Jupiter 30–100 000 yr. dynamically unstable residence periods in and Saturn-crossing) orbits, because many While the terrestrial record of impacts is the outer planetary region. The disintegra- of these will not come close enough to the incomplete and does not support the notion tion of such giant comets would produce Sun during that interval so as to become that all mass extinctions arise from asteroid intermittent but prolonged periods of bom- discoverable. To discover and track all such impacts, the idea that impacts can affect our bardment lasting up to 100 000 years. Mass asteroids would require a survey lasting planet’s climate and life is now established extinction/geological boundary events on longer than the maximum orbital periods (see box, “Impacts and extraterrestrial bod- Earth show such a pattern, as do levels of deemed to be of interest. This links to the ies: the story so far”). These new astronomi- dust and meteoroids in the upper atmos- problem of long-period or near-parabolic cal and terrestrial lines of evidence merit a phere. Over the past 10 000 years, Earth has comets: perpetual vigilance is necessary reappraisal of the celestial hazard faced by been experiencing the intermittent arrival to protect civilization (Marsden & Steel contemporary civilization. Here we concen- of dust, meteoroids and comet fragments 1994). The claim that the NASA Spaceguard trate on the implications over timescales of from the disintegration of comet 2P/Encke, programme has (so far) led to the discovery order 10 000 yr. trapped within the orbit of Jupiter. and orbit determination of 93 or 95% of The assessments by Chapman & NEOs larger than 1 km in size is justifiable, The centaur flux Morrison (1994) and others of the major so long as one recognizes that the term (5145) Pholus, the second-discovered impact hazard posed to modern civiliza- “NEOs” as used there implies only aster- centaur after (2060) Chiron, and

tion are based on the assumption, usually oids in short-period (i.e. cis-jovian) orbits, (15760) 1992 QB1, the second trans- unspoken, that the contemporary NEO and not those asteroids in larger orbits – Neptunian object (TNO) after Pluto, were population is in a steady state, in terms of and no comets at all. both discovered as recently as 1992. Discov- the populations at different sizes, being The basic model of NEO supply by ery rates have increased such that known maintained primarily through leakage of gravitational perturbations upon main-belt centaurs now number in the hundreds and

6.24 A&G • December 2015 • Vol. 56 • www.astrongeo.com This article has been accepted for publication in Astronomy & Geophysics ©: 2015 Bill Napier, David Asher, Mark Bailey and Duncan Steel. Published by Oxford University Press on behalf of the Royal Astronomical Society. All rights reserved. Centaurs http://astrogeo.oxfordjournals.org/content/56/6/6.24.abstract

Impacts and extraterrestrial bodies: the story so far

The idea that the geological and biological 90 km diameter Popigai crater, are associated 1 Correlations between evolution of the Earth proceeds in isolation with regional, not global, effects. The evidence from its surroundings, developed by Darwin, of iridium excess, shocked quartz and the volcanism, extinctions Hutton, Lyell and others, was the default like at extinction boundaries has either been and bombardment position in the Earth sciences until the 1970s. ambiguous or absent (MacLeod 2013). Further, This perspective began to change when some of the mass extinctions, such as that VEB E = BV = BV = E terrestrial craters such as Odessa (Texas), in the Late Devonian, 365 Myr ago, appear to 2 1 Barringer (Arizona) and Henbury (Australia) have occurred in steps (McGhee 1996), while ● were associated with iron meteorites; when others coincide in time with massive volcanic/ 15–17 spacecraft missions in the 1960s indicated that flood basalt activity (see table 1). 29–31 planetary surfaces are generally pockmarked The fact that the main asteroid belt 34 35 ● with craters as predicted by Öpik (1951); when between Mars and Jupiter fails by an order of 49 returned Apollo samples were shown to magnitude to supply the required transient 55–61 display evidence of shock metamorphism, population of large near-Earth objects (NEOs), 65–67 66 65 ● ● ● demonstrating the well-known lunar craters to or to account for the number of large impact 87–90 be of impact (rather than volcanic) origin; and craters (Menichella et al. 1996, Minton & 94 when wide-field telescopic surveys unveiled Malhotra 2010) is another complication. 109 –119

a significant population of small (~1 km) This perspective raises the question of Downloaded from 122–122 122 ● near-Earth asteroids (Helin & Shoemaker 1979). whether, and to what extent, cosmic impacts 132–134 Impact cratering is now considered a common in the present epoch (the Holocene) might 146 144 and continuing geologic process. have deleterious effects limited to restricted ● 166 Napier & Clube (1979) indicated that impact areas, as with the 1908 Tunguska impact or the 180–183 183 ●

rates were high enough to be associated with 2013 Chelyabinsk event, or globally. Chapman http://astrogeo.oxfordjournals.org/ the mass extinctions seen in the geological & Morrison (1994) considered that the arrival 199–202 200 200 ● ● ● record and they proposed possible extinction of a ~2 km asteroid would result in the death 250–253 251 255 (●)(●) ● agencies: collapse of food chains as a result of a large fraction of the human population, 257–260 260 ● of impact-generated fine dust in the upper and that there is a probability of about 1% 322 atmosphere blocking out sunlight; ozone-layer of such an event in the next 10 000 yr. The 360 destruction; straightforward blast; and others. combination of small probability and large 364–367 The concept of food-chain collapse caused by consequences (for substantial NEOs) has led 370–370 dust was not new: Hoyle & Wickramasinghe to systematic searches for large near-Earth 375 377 ● (1978) had suggested that such dusting might asteroids (>1 km) in cis-jovian orbits and their 385 be caused by an encounter with the coma of discovery appears to be above 93% complete by guest on November 20, 2015 445 445 ( ) a comet. Earlier suggestions of links between (cf. Harris & D’Abramo 2015), although ● 488 asteroid impacts and geological boundary attention has now been shifted to smaller (i.e. mass extinction) events include Nininger NEOs around 50 m across. 495 (1942) (see Steel 1995). There are emerging claims from some Earth 499 The discovery of iridium – an element scientists that the terrestrial disturbance of Table Correlations over the last ~500 Ma common in meteorites but sparse in the ~12 800 BP, the sudden onset of 1300 years between peak activity of large igneous provinces Earth’s crust – at the Cretaceous–Palaeogene of cooling, a global collapse of large animal (V; Bond & Wignall 2014), marine genera mass boundary (Smit & Hertogen 1980, Alvarez et populations and other terrestrial phenomena extinctions (E; Bambach 2006) and bombardment al. 1980) cemented the idea that the great (e.g. Wittke et al. 2013, Kinzie et al. 2014, Kennett episodes (B; Napier 2015). Dates are in Ma. Tem- mass extinctions of the biological record could et al. 2015) had a cosmic or even cometary poral coincidences within measurement errors originate in highly energetic asteroid impacts. origin, although this remains controversial. are marked by ● or by (●) where one impact has Early enthusiasm for a simple link – giant A similar claim has also been made for the been identified. There is an empirical three-way asteroid impacts = mass extinctions – has not simultaneous collapse of civilizations, global correlation between bombardment episodes, generally been supported by palaeontological cooling and drought, at near 2350 BC (Courty the creation of some large igneous provinces and evidence. Some of the major extinctions et al. 2008). With current collision rates, the mass extinctions, implying that some massive may be associated with relatively small largest asteroid impact energy expected volcanisms may have been induced or exacer- craters, perhaps incapable of explaining the during the Neolithic is 300–600 megatons, bated by such episodes. Richards et al. (2015) have biological devastation recorded in the strata. which is incapable of causing the global shown that the Chicxulub crater formed within Conversely, several large craters, such as the upsets that appear to have taken place. 0.1 Myr of a major outburst of Deccan volcanism. known TNOs over a thousand. Although (2013) use 5 < q < 28 and a < 1000 au, where q most being weeded out along the way by formally classed as being “minor plan- is perihelion distance and a semimajor axis the influences of the giant planets. ets” (i.e. asteroids), centaurs are cometary (see Steel 2014). Centaurs comprise a transi- Neglecting physical destruction and bodies, composed of volatile ices as well tional population deriving from cometary non-gravitational effects, about one in as silicates, which cross or approach the reservoirs beyond Neptune. Some reach ten centaurs in Chiron-like orbits become orbits of one or more giant planets and are Earth-crossing orbits (Hahn & Bailey 1990, Earth-crossers, often repeatedly (Horner therefore dynamically unstable. The exact Asher & Steel 1993), even though only a et al. 2004ab, Napier 2015): an example definition of what entails a “centaur” varies minor fraction of the objects that begin is shown in figure 1. The rate of arrival between researchers: Emel’yanenko et al. such an inward passage actually get here, of ≥100 km centaurs into short-period,

A&G • December 2015 • Vol. 56 • www.astrongeo.com 6.25 This article has been accepted for publication in Astronomy & Geophysics ©: 2015 Bill Napier, David Asher, Mark Bailey and Duncan Steel. Centaurs Published by Oxford University Press on behalf of the Royal Astronomical Society. All rights reserved. http://astrogeo.oxfordjournals.org/content/56/6/6.24.abstract

1 The evolution of 1.8 45 perihelion distance 40 (left) for a modelled 1.6 centaur with integrated 35 1.4 Earth-crossing lifetime 30 ~22 000 years. Over the 1.2 25 bulk of this interval, the

q au Earth semimajor axis is ~3 au, 1 20 inclination corresponding to an 15 0.8 orbital period ~5 years. Venus 10 The orbit has many 0.6 low-inclination phases, 5 providing opportunities 0.4 0 for terrestrial encounters 240000 250000 260000 270000 280000 240000 250000 260000 270000 280000 with its debris trail. time (yr) time (yr)

Earth-crossing orbits is of order one per that there is indeed a strong tendency for These near-synchronicities are difficult 30 000 yr (Napier 2015). Integrating the impacts to occur in tight clusters (Napier to reconcile with the impactors as frag- orbits of 100 Chiron clones, the mean total 2015). There are 27 craters of this dataset ments arriving from the disintegration of dynamical lifetime in an Earth-crossing with differential ages dt ≤ 2 Myr, and 22 with a main-belt asteroid (lasting ~5–100 Myr), configuration was found to be ~17 400 yr, dt ≤ 1.5 Myr, comparable with the RMS sum or as independent comets, even within a Downloaded from with a highly skewed distribution, the of their quoted radiometric age uncertain- comet shower (duration ~2 Myr). They are median lifetime being 6700 yr (figure 2). ties. In all, nine impact episodes have been most easily understood in terms of the Substantial physical disintegration is likely identified (figure 3). Paradoxically, the bom- arrival and disintegration of a Chiron- over such an interval (Di Sisto et al. 2009). bardment episodes emerge more strongly sized Earth-crossing comet, in a Jupiter

A prime form of disintegration, as where the dataset is sparse, namely for the family or Encke-like orbit, into hundreds http://astrogeo.oxfordjournals.org/ opposed to sublimation of volatiles and 16 impact craters older than or thousands of short-lived complete outgassing, appears to be gross 100 Myr. Their mean age “Large comets may form cometary fragments. Di Sisto fragmentation, either through internal separation is 25 Myr and yet by the aggregation of et al. (2009) showed that the processes such as thermal stresses or, on 12 of the 16 belong to crater building blocks from perihelion distribution q of occasion, passage through the Roche limit pairs with dt < 1.6 Myr, a result various locations” the current Jupiter family of Jupiter or the Sun. A large centaur may which has chance probability population requires strong hold 102–103 times the mass of the current of ~0.7%. With the more thorough erasure erosion at small q to match the observa- near-Earth asteroid system, and an episode of impact craters at these more remote tions, along with a high rate of cometary of bombardment may be expected whose epochs, most impact episodes will go unre- splittings and outbursts. Such evolution duration is defined by the dynamical or corded and only the strongest are likely to is to be expected for the Jupiter family by guest on November 20, 2015 physical lifetime of the comet and its frag- have surviving multiple craters. throughout geological history. ments, whichever is the shorter. The Popigai/Chesapeake Bay impactors One consequence of the hypothesis Stratigraphic studies had different chemical compositions, H argued herein, that large comets arrive Relative dating through stratigraphic and L chondrite respectively, which is not sporadically, then disintegrate and pro- studies has the potential to provide a much easily reconciled with their provenance duce enhanced terrestrial bombardment tighter constraint on the duration of impact from the break-up of a single asteroid. The episodes, is that impact cratering should episodes. It is available for only two pairs near-synchronicity of two large, unrelated be episodic. Such episodes may be distin- of craters: the 150 km Chicxulub crater in asteroids is dynamically unlikely even as guished from those arising from comet the Gulf of Mexico and the 24 km Boltysh part of a break-up event in the main belt (and showers (Bailey et al. 1990) or groups of crater in Ukraine, at the end-Cretaceous; no asteroid family at that date is known). main belt asteroids (Zappalà et al. 1998) by and the 90 km Popigai crater in Siberia The distinct compositions may be consistent their duration. Short-period comet disin- and the 40 km Chesapeake Bay crater in with the break-up of a contact binary comet tegration may occur over ~0.01–0.1 Myr; the eastern USA, in the Late Eocene. The formed from a gentle merger, as appears to a shower of comets from the Öpik–Oort end-Cretaceous craters are separated in age be the case with 67P/Churyumov-Gerasi- cloud will be spread over ~2 Myr; and that by ≲2500 yr (Jolley et al. 2010), and the Late menko (Rickman et al. 2015), a morphologi- from the formation of an asteroid family Eocene ones by 10–20 000 yr (Koeberl 2009). cally diverse comet (Thomas et al. 2015). The lasts ~5–100 Myr. The enhancement in flux The latter pair may be joined by two other elemental abundance of material recovered of ⪆1 km asteroids on Earth-approaching craters of similar age (Napier 2015); this from Comet 81P/Wild 2 by the Stardust orbits following a large main-belt asteroid episode is associated with a peak in cosmic spacecraft was that of CI chondrites, with break-up is small in relation to the back- dust input at 35.8 Myr ago, determined the presence of minerals formed at high ground (Napier & Asher 2009) because of from helium-3 laid down in ocean sedi- temperatures indicating large-scale radial the spread in time over which such objects ments (Farley 2009). mixing of material at its time of formation. are injected onto inner solar system paths. Chicxulub is the largest crater in the This opens up the prospect that large comets We tested this conjecture with the avail- dataset of 41 impactors employed. Monte may form by the aggregation of building able terrestrial impact database (Earth Carlo trials reveal that the odds of its near- blocks from a variety of locations, but their Impact Database 2014). We used 41 impact synchronicity with a crater at least as large provenance, chemical and isotope composi- structures over 3 km in diameter from the as Boltysh are 2500:1 against. The odds tions are too uncertain for strong statements. past 500 Myr (i.e. the Phanaerozoic), radio- that Popigai, the second largest, coincides A group of large near-Earth asteroids, in metrically dated to precision 2s ≤ 5 Myr. in time with a crater as large as that at orbits similar to that of 2P/Encke, have spec- Frequentist and Bayesian analyses confirm Chesapeake Bay are around 1000:1 against. tra matching those of ordinary chondrites

6.26 A&G • December 2015 • Vol. 56 • www.astrongeo.com This article has been accepted for publication in Astronomy & Geophysics ©: 2015 Bill Napier, David Asher, Mark Bailey and Duncan Steel. Published by Oxford University Press on behalf of the Royal Astronomical Society. All rights reserved. Centaurs http://astrogeo.oxfordjournals.org/content/56/6/6.24.abstract

100000 3 (Right) Probable bombardment 1 = current epoch episodes identified in Napier (2015). 2 = Eocene–Oligocene In addition, two single craters, 3 = unidentified with major transition 10000 marked by asterisks, coincide 4 = Cretaceous–Palaeogene 5 = unidentified with major transition within dating uncertainties with 6 = Jurassic–Cretaceous 1000 the Permo–Triassic (250 Ma) and 7 = middle–late Jurassic 8 = Triassic–Jurassic

duration (years) Ordovician–Silurian extinctions. 9 = Late Devonian

100 The curves are cubic fits to the data; there is no unique best fit. The decline in crater numbers as one 10 1 2 3 4 5 6 7 8 9 clone number goes into the past is presumably due to subduction, erosion and 2 The integrated durations of the q < 1 au phases overlaying by sedimentation. for nine clones based on the orbit of (2060) Chiron.

(Popescu et al. 2014). Five bright Taurid large fluctuations, following the arrival (Babadzhanov et al. 1990, Steel et al. 1991, fireballs observed between 2010 and 2012, and disintegration of large bodies, and that Porubcˇan & KornoŠ 2002) is apparent in of undoubted cometary provenance, had the remnants of a large comet may still be accurate, double station TC meteor orbits. spectra compatible with a chondritic compo- orbiting among the overall NEO popula- Babadzhanov et al. computed that up to sition (Madiedo et al. 2014). tion (Clube & Napier 1984), which includes 18 kyr is needed to induce this relative Downloaded from Figure 3 reveals an apparent gap in the not only near-Earth asteroids, but also longitude precession. This may be taken impact record between 5 and 35 Myr ago. boulder- and meteoroid-sized objects down as a lower age limit for a system. The TC is This dearth extends to the 184 impact to submillimetre dimensions. Such appears embedded in the broader and more mas- structures in the Earth Impact Database: to be the case, as we indicate below; we also sive (Štohl) sporadic stream, although if we

only the 24 km Ries crater, from an achon- discuss the implications over timescales regard the TC as broad enough to contain http://astrogeo.oxfordjournals.org/ dritic impactor, is securely placed in this relevant to civilization. 12 showers over three w cycles then there time interval. Models involving random is no precise dividing line, the terms TC vs arrival times suggest that about 50 impacts The Taurid Complex Štohl simply describing a more-structured capable of forming a crater >20 km across The Taurid meteoroid stream has long been versus somewhat more-dispersed stream. should have taken place over this period, 17 recognized as an ancient, widely dispersed Overall we can associate a timeframe of of them on land (Bland & Artemieva 2006). system through which the Earth takes ~20 kyr for the formation of this broad For such a recent interval, most craters several weeks to pass each year, in contrast complex of meteoroids on cis-jovian, mod- should have been identified. The prob- to most meteor showers, which persist erate eccentricity (e ~ 0.8) orbits. This is also ability is ~0.001 that such a gap would exist for only a few days. It can be separated consistent with the 10–20 kyr required for by chance anywhere during the interval observationally into northern and southern dynamical evolution into small-q orbits so by guest on November 20, 2015 0–100 Myr. Conversely, there seems to have branches, radiants being a few degrees as to explain the observed excess calcium been an enhancement in impact cratering above and below the ecliptic (e.g. Wright emission in Mercury’s exosphere as being in the past 5 Myr relative to the longer-term & Whipple 1950, Štohl & Porubcˇan 1992). due to the TC (Christou et al. 2015). average. Such dips and surges suggest that There are four branches in all (figure 4), Splitting events and multiple fragmenta- the largest impactors are derived from the post-perihelion encounters producing day- tions are a prime disintegration route for break-up of large comets, because their time meteors: the z Perseids radiate from comets (Sekanina 1997, 2002, Sekanina & mass distribution is top-heavy and their north of the ecliptic, the b Taurids from Chodas 2004, Williams 2011). By studying input is erratic and episodic. To the con- south. Planetary perturbations over mil- the association of dormant comets with trary, Menichella et al. (1996), dynamically lennia cause the argument of perihelion w meteoroid streams, Jenniskens (2008a, modelling the supply from the main belt of of a Taurid-type orbit to cycle through 360° 2008b) argued that discrete break-up/ near-Earth asteroids down to bodies 0.1 km (Whipple 1940), so that all w values are now fragmentation events dominate continuous across, find that the population is unlikely present within the stream, which we term cometary activity in the creation of streams to vary by more than a factor of two over the Taurid Complex (TC). The four shower and indeed comprise the main replenish- timescales of 10 Myr and longer. branches represent the Earth-intersecting ing source for the zodiacal cloud. This is in A 100 km comet has a mass ~100 times cross-sections of the TC. line with the finding of Di Sisto et al. (2009) that of the contemporary NEO population, To develop the range of w precession that the observed structure of the Jupiter and entry of such a body into a short- required to span the four branches takes family of comets is best explained by physi- period, Earth-crossing orbit is anticipated ~5 kyr (Babadzhanov et al. 1990, Babadzh- cal disintegration as well as dynamical with a characteristic timescale ~30 000 yr. anov & Obrubov 1992). In fact precession evolution, with splitting a major process. This timescale estimate is conservative has occurred through more than one With the passage of time, meteoroids – the because non-gravitational forces were complete w cycle, with a full cycle before particles that comprise the streams – lose neglected, these increasing the rate of the main four Taurid branches and a cycle the memory of their origins, and merge into transfer into orbits uncoupled from Jupiter after; correspondingly longer than 5 kyr is the zodiacal cloud, becoming sporadics. by a factor of a few (Asher et al. 2001). The required. The observed night-time showers A long-standing issue is that the zodiacal current clearance time of the zodiacal are known as Piscids (the w cycle before) cloud is substantially overmassive in rela- dust cloud is ~20 000 yr as a result of the and c Orionids (after), both with north- tion to current inputs, which largely come Poynting–Robertson effect and collisional ern and southern branches (Olsson-Steel from short-period comets, in particular 2P/ processes (Grün et al. 1985, Steel & Elford 1987, Babadzhanov et al. 1990, Babadzh- Encke. The idea of a zodiacal cloud imbal- 1986). It is thus reasonable to expect that anov & Obrubov 1992). Further, a cor- ance has a long history, Hughes (1996) con- the zodiacal dust cloud will be subject to relation of semimajor axis with longitude cluding that “at some time in the last 103 to

A&G • December 2015 • Vol. 56 • www.astrongeo.com 6.27 This article has been accepted for publication in Astronomy & Geophysics ©: 2015 Bill Napier, David Asher, Mark Bailey and Duncan Steel. Centaurs Published by Oxford University Press on behalf of the Royal Astronomical Society. All rights reserved. http://astrogeo.oxfordjournals.org/content/56/6/6.24.abstract

105 yr, the cloud has benefited from a large and unusual mass enhancement”, in accord with long-term fluctuations in its mass deriving from a simple model of comet arrival and disintegration (Napier 2001). Radar and optical studies indicate that the current zodiacal cloud has a total mass ~4 × 1019 g, to within about 50%. Around 1020 g of material (cf. Whipple 1967, Hughes 1996) must have been supplied to the zodiacal cloud on 104 yr timescales. The presence of the Štohl stream suggests that the TC progenitor is the cometary source. The latter cannot be supplied by 2P/Encke in its present state, whose mass loss is only about one-hundredth of that required for 4 Four Taurid Complex branches (meteor showers). The separation into branches is not real within steady-state replenishment (Nesvorný et al. the physical stream; it is an artifact of the fact that the Earth’s path in space only intersects orbits with 2011). The dominance and structure of the particular parameters: for a given size and shape of orbit, the argument of perihelion w must take one of TC in the zodiacal cloud suggests that the four values, corresponding to northern and southern branches, before and after perihelion. Part of orbit bulk of it has been supplied by the hierar- above ecliptic drawn thicker, solid; part below ecliptic thinner. chic disintegration of an erstwhile large Downloaded from comet. If assembled into a single body of objects, is 2 × 1019 g (Asher et al. 1994, Napier Concentrations of material within the bulk density 0.4 g cm–3 and an equal mass & Asher 2009). The disintegration of the debris trails persist due to resonances, lead- of lost volatiles is assumed, a progenitor original progenitor comet thus appears to ing to episodes of high-intensity bombard- comet of diameter ~100 km is implied. An have given rise to a hierarchy of fragmen- ment when the Earth runs through such

early estimate of 25–30 km was given by tations, including debris trails as well as material. The most prominent meteoroid http://astrogeo.oxfordjournals.org/ Sekanina (1972), provided the 2P/Encke short-lived comets. In summary, a total swarm impacting the Moon during the progenitor was captured ~20 000 years mass loss, first into the TC and then via years of the Apollo seismic experiment ago. How much material has been lost the Štohl stream into the zodiacal cloud, of occurred in late June 1975 (Oberst & Naka- altogether is unknown; the possibility that 1020 g in the past ~20 kyr may be deduced. mura 1991), coinciding with the b Taurid the progenitor of the TC was rather larger As well as splitting, sublimation is a major shower: as many boulders struck the Moon cannot be excluded. route of cometary disintegration. The zodia- over a five-day period as struck it over the cal cloud is fed by way of other five years of the experi- Interplanetary dust debris trails, which are a com- “A fly-by revealed ment’s data collection. This, Analysis of the ages of the lunar micro- mon feature of short-period that icy fragments of along with evidence of the craters (“zap pits”) on rocks returned in comets. The trails are domi- 15–20 cm were being night-time Taurids having by guest on November 20, 2015 the Apollo programme indicate that the nated by millimetre-sized ejected from its surface” higher activity in specific near-Earth interplanetary dust (IPD) flux particles, although a fly-by of years (Asher & Izumi 1998, has been enhanced by a factor of about 103P/Hartley 2 revealed that icy fragments Beech et al. 2004, Johannink & Miskotte ten over the past ~10 kyr compared to the of 15–20 cm dimensions were being ejected 2006, Dubietis & Arlt 2007), indicates that long-term average, if one assumes that from its surface (A’Hearn et al. 2011), and there are concentrations of material in the the heavy ion solar cosmic ray (SCR) flux metre-sized bodies in bound orbits have TC. The years of this higher observed activ- has been constant. The IPD flux and its been detected close to the nucleus of 67P/ ity are as predicted if a meteoroid swarm time-variation were inferred from the SCR- Churyumov-Gerasimenko (Rotundi et al. exists in the 7:2 jovian resonance (Asher & produced nuclear track densities in the 2015). The maximum size of dust aggregate Clube 1993), a strong dynamical influence melt-glass floors of the microcraters, which that can be lifted from a comet comes from in the TC (Soja et al. 2011). allow them to be dated, Hartung & Storzer a balance between gravity and sublimation The dust trail observed by IRAS in the (1974) suggesting that the abundance of (gas outflow) pressure; in the case of a comet, thermal infrared (Sykes et al. 1986) close such young microcraters might be linked say, 50 km in diameter in an Encke-like orbit, to 2P/Encke’s orbit may be interpreted to the arrival of the progenitor of 2P/Encke. metre-sized fragments may be released in terms of the same resonance (Asher & Measurements of the space-exposure ages around perihelion. These, and smaller (~cm) Clube 1993), implying that the material of IPD collected in the stratosphere (~104 yr: meteoroids, are then collisionally eroded released in the present epoch occupies Bradley et al. 1984) provide supplementary (on timescales of order 104 yr: Steel & Elford orbits well away from Earth intersection. support for this notion. Zook (1978) argued 1986) to produce a major fraction of the In any epoch, the TC’s densest region will to the contrary, postulating that the IPD zodiacal dust cloud. If the zodiacal cloud comprise a concentrated trail close to the flux has been more-or-less constant, and were, say, 100 times as massive as the current parent orbit. Orbital precession, specifically it is the SCR flux that has varied – without one, erosion would be more rapid, and icy the w precession described above, brings support from solar physicists. aggregates up to metre dimensions could this trail or core to Earth inter­section dur- The TC population extends upwards to be lost on that same timescale (Napier 2004). ing particular epochs: the w inter­sections macroscopic (asteroid-sized) bodies (Asher Numerical modelling reveals that, without occur in pairs a couple of centuries apart, et al. 1994, Napier 2010), some of which replenishment, radiative and collisional successive pairs separated by a few millen- have meteor showers associated with them processes will reduce the mass of a substan- nia (Asher et al. 1994, Steel & Asher 1996). (Babadzhanov 1998, 2001, Babadzhanov et tial zodiacal cloud by an order of magnitude Given the presence of such a large, al. 2008). 2P/Encke is a 5 km diameter comet within this 104 yr timescale (Napier 2001), disintegrating comet in our environment (Boehnhardt et al. 2008) and the mass in although particles of high tensile strength throughout the Holocene and earlier, the TC and Štohl stream, including large may persist for longer. we may ask about the probability and

6.28 A&G • December 2015 • Vol. 56 • www.astrongeo.com This article has been accepted for publication in Astronomy & Geophysics ©: 2015 Bill Napier, David Asher, Mark Bailey and Duncan Steel. Published by Oxford University Press on behalf of the Royal Astronomical Society. All rights reserved. Centaurs http://astrogeo.oxfordjournals.org/content/56/6/6.24.abstract

5 Recently released cometary material is not dispersed evenly into the comet’s debris stream, but remains concentrated within dense, narrow trails. Shown is the density profile (cross-section in the ecliptic plane, then along Earth’s assumed path) within a Taurid trail, computed following Asher (2008). Earth is likely to have passed close to the centre of a Taurid trail once or more

during the Holocene. Downloaded from http://astrogeo.oxfordjournals.org/ consequences of passage through a debris nuclear war (Crutzen et al. 1984, Robock and other geochemical anomalies that trail. While trail cross-sectional shapes et al. 2007). Dusting of this intensity is they attribute to an oceanic impact in the vary from stream to stream (Asher 2010, expected to reduce the level of sunlight to southern Indian Ocean. Estimates of the Asher & Steel 2012), the common feature is that of moonlight, and result in a global duration of the cooling vary from a decade a high spatial density of meteoroids if Earth cooling sufficient to destroy commercial to two centuries. encounters a trail’s central region. The agriculture (Robock et al. 2007). Such work It has been proposed that a sudden Leonid meteor storms (Brown 1999) prove has implications for atmospheric dusting global cooling at 12 800 BP, which persisted that meteor activity from a stream need not events of cosmic origin, although there are for 1300 years, led to cultural reorganiza- be uniform, and that substantial fractions significant differences, of course. Hoyle & tions of earlier societies in southwest Asia of the total mass input can arrive during Wickramasinghe (1978) considered that the (Moore et al. 2000) and North America by guest on November 20, 2015 the Earth’s rare, deep encounters with acquisition of ~1014 g of comet dust in the (Anderson et al. 2011), and a substan- dense trails (Kondrat’eva & Reznikov 1985, upper atmosphere would have a substan- tial decline of large mammals in North McNaught & Asher 1999, Lyytinen & Van tial effect on the Earth’s climate. Such an America, Europe and Australia. This event, Flandern 2000). Figure 5 shows a Taurid encounter is a reasonably probable event at the Younger Dryas (YD) boundary, was trail cross-section and an example traverse during the active lifetime of a large, disinte- accompanied by a peak in microspherule through it by the Earth. Overall we should grating comet in an Encke-like orbit (Napier deposition over at least that 10% of the expect enhanced mass influxes during TC 2010). If the meteoric smoke has absorption Earth’s surface so far investigated (Wittke et w-intersection epochs. For a 1020 g comet coefficientk ~ 105 cm2 g–1, comparable with al. 2013), the presence of high-temperature undergoing a hierarchical disintegration that of soot, then passage through ~1014 g of melt glass (>2100 °C) at several locations, across ~10 4 yr, there is an expectation of one dust would yield a veil of substantial opti- a thin carpet of nanodiamonds laid down encounter (or at most a few) with a debris cal depth for as long as the smoke persisted synchronously (to within ~100 yr) with the 14 trail sufficient to dump ≳10 g (cf. Napier (Hoyle & Wickramasinghe 1978). The effect onset of the cooling (Kinzie et al. 2014, Ken- 2010 section 3) into the atmosphere. is comparable with the darkening expected nett et al. 2015), a large platinum spike of from an asteroid land impact of energy non-terrestrial provenance in the GRIP ice Terrestrial consequences 106 megatons, expected at ~2 Myr intervals core (Petaev et al. 2013), and other indicators The effects of running through the debris according to Chapman & Morrison (1994). of cosmic disturbance. There is addition- trail of a large comet are liable to be com- Societal collapses in antiquity often fol- ally an abundance of charcoal and soot plex, and to involve both the deposition of lowed abrupt climatic coolings of unknown at the boundary, likely to be indicative of fine dust into the mesosphere and, poten- origin, persisting for decades and leading extensive wildfires. Sharp ammonium and tially, the arrival of hundreds or thousands to drought and famine (Weiss & Bradley nitrate spikes in Greenland ice cores, the of megaton-level bolides over the space 2001). Two such collapses have been linked, largest in 300 000 years, have been linked of a few hours. Incoming meteoroids and albeit controversially, with cosmic distur- to this boundary; these may have been pro- bolides may be converted to micron-sized bances. Around 2350 BC the earliest civili- duced by extensive biomass burning, atmo- smoke particles (Klekociuk et al. 2005), zations, throughout the near East, northern spheric chemistry induced by incoming which have high scattering efficiencies India and China, collapsed simultaneously, fireballs, or direct deposition of cometary and so the potential to yield a large opti- apparently driven by cooling and drought. material (Melott et al. 2010). Whatever their cal depth from a small mass. Modelling Courty et al. (2007) have found evidence for origin, they confirm that the YD cooling of the climatic effects of dust and smoke the dispersal of hot, viscous impact ejecta was an exceptional event over at least this loading of the atmosphere has focused over hemispheric dimensions (Peru, Spain, timescale. To match the 1300 yr cooling of on the injection of such particulates in a Syria etc) at this time, with nanodiamonds the YD, a quasi-equilibrium perturbation

A&G • December 2015 • Vol. 56 • www.astrongeo.com 6.29 This article has been accepted for publication in Astronomy & Geophysics ©: 2015 Bill Napier, David Asher, Mark Bailey and Duncan Steel. Centaurs Published by Oxford University Press on behalf of the Royal Astronomical Society. All rights reserved. http://astrogeo.oxfordjournals.org/content/56/6/6.24.abstract

results are shown in table 2. The statistics Other existential hazards 2 Bombardment episodes reveal no significant difference between The main existential hazards to be faced the temporal behaviour of large and small over the next few centuries are likely to be diameter (km) in out craters over the size range tested, 3 km anthropogenic (Bostrom 2013), but these >20 15 7 upwards, corresponding to bolides of can in principle be countered. No natural 10–20 4 6 ≳200 m dimensions. On this evidence, large existential risks are known to be imminent, 3–10 6 3 comets arrive accompanied by a flotilla of although solar superflares and cometary Membership of bombardment episodes for fragments, and we do not receive a steady encounters are currently unpredictable. A the 41 accurately dated terrestrial impact rain of small bolides through time. Further centaur arrival carries the risk of inject- craters. There is no evidence of a systematic evidence of current nonequilibrium among ing, into the atmosphere, from above, a trend with diameter. the sub-km meteoroids comes from the mass of dust and smoke comparable to fireball study of Brown et al. (2013), who that assumed in nuclear winter studies. find that the number of impactors with Thus, in terms of magnitude, its ranking of the terrestrial environment induced by diameters of tens of metres may be an order among natural existential risks appears to such comet dusting would be required, of magnitude higher than estimates based be high. Further, according to the analysis as proposed by Hoyle (1981), but no one on telescopic surveys and lunar counts. presented here, such events happen at least mechanism on its own explains the YD They suggest that there is a departure from a hundred times more frequently than cooling (Renssen et al. 1981). equilibrium in the NEO population for globally incinerating asteroid impacts. Apart from their effects on atmospheric objects between 10 and 50 m in diameter. Whether the calamities at 2350 BC and opacity, a swarm of Tunguska-level The current Tunguska-level rate inferred 12 800 BP were in fact due to passages fireballs could yield wildfires over an area from their study is in line with an assess- through cometary debris is a matter for fur- Downloaded from of order 1% of the Earth’s surface (Napier ment based on lunar impact counts (Asher ther enquiry; irrespective of this, there is a 2010). Large comets in Encke-like orbits et al. 2005). All three major 20th-century need for a more thorough exploration of the may sporadically generate a substantial impacts (Tunguska, British Guiana, Curuçà centaur population from both theoretical population of small, albeit short-lived, River) coincided with our passage through and observational points of view, and for a

impactors. To test this idea, the 41 craters major meteoroid streams (respectively the better understanding of the environmental http://astrogeo.oxfordjournals.org/ culled from the Earth Impact Database b Taurids, Geminids and Perseids); the odds consequences of encountering a dense com- were divided by membership of bombard- against this arising by chance are about plex of cosmic dust and meteoroids. ● ment episodes and crater diameter. The 1000:1 (Napier & Asher 2009).

Authors Bailey M E et al. 1990 The Origin of Comets Space Sci. 53 523 Olsson-Steel D 1987 Observatory 107 157 All four authors of this review have had some asso- (Pergamon, Oxford) Hughes D W 1996 Quart J. Roy. Astron. Soc. 37 593 Öpik E J 1951 Proc. R. Irish Acad. 54 165 ciation with the Armagh Observatory (AO). Mark Bambach R K 2006 Annu. Rev. Earth Planet. Sci. Jenniskens P 2008a Icarus 194 13 Petaev M I et al. 2013 Proc. Nat. Acad. Sci. 110 12917 Bailey has served as director there for the past 20 34 127 Jenniskens P 2008b Earth Moon Planets 102 505 Popescu M et al. 2014 Astron. Astrophys. 572 A106 years; David Asher is currently an astronomer at Bauer J M et al. 2013 Astrophys. J. 773 22 Johannink C & Miskotte K 2006 WGN J. Int. Porubčan V & Kornoš L 2002 in Warmbein B (ed.) AO; Bill Napier was formerly an astronomer there; Beech M et al. 2004 Observatory 124 277 Meteor Org. 34 7 Proc. ACM 2002, ESA-SP-500 (ESA, Noordwijk) 177 and Duncan Steel is a visiting astronomer at AO. Bland P A & Artemieva N A 2006 Meteorit. Planet. Jolley D et al. 2010 Geology 38 835 Renssen H et al. 2015 Nature Geoscience 12 by guest on November 20, 2015 Napier and Steel also hold positions as visiting Sci. 41 607 Kennett J P et al. 2015 Proc. Nat. Acad. Sci. July 27 October DOI: 10.1038/NGEO2557 professors of astrobiology at the University of Boehnhardt H et al. 2008 Astron. Astrophys. Kinzie C R et al. 2014 J. Geol. 122 475 Richards M A et al. 2015 GSA Bulletin 30 April Buckingham. 489 1337 Klekociuk A R et al. 2005 Nature 436 1132 Rickman H et al. 2015 Astron. Astrophys. accepted Bond D P G & Wignall P B 2014 in Keller G & Kerr Koeberl C 2009 Geol. Soc. Amer. Spec. Pap. 452 17 Robock A et al. 2007 J. Geophys. Res. 112 D13107 DEDICATION A C eds Geol. Soc. Amer. Spec. Pap. 505 29 Kondrat’eva E D & Reznikov E A 1985 Sol. Syst. Rotundi A et al. 2015 Science 347 3905 This article is dedicated to the memory of Ernst Bostrom N 2013 Global Policy 4 15 Res. 19 96 Sekanina Z 1972 in Chebotarev G A et al. (eds) Öpik (1893–1985), a long-term staff member at Bradley J P et al. 1984 Science 226 1432 Lyytinen E & Van Flandern T 2000 Earth Moon Proc. IAU Symp. 45 (Reidel, Dordrecht) 301 the Armagh Observatory, on the 30th anniversary Brown P 1999 Icarus 138 287 Planets 82 149 Sekanina Z 1997 Astron. Astrophys. 318 L5 of his death. Brown P G et al. 2013 Nature 503 238 MacLeod N 2013 The Great Extinctions (Natural Sekanina Z 2002 Astrophys. J. 576 1085 Chapman C R & Morrison D 1994 Nature 367 33 History Museum, London) Sekanina Z & Chodas P W 2004 Astrophys. J. references Christou A A et al. 2015 Geophys. Res. Lett. accepted Madiedo J M et al. 2014 Icarus 231 356 607 620 A’Hearn M F et al. 2011 Science 332 1396 Clube S V M & Napier W M 1984 Mon. Not. Roy. Marsden B G & Steel D I 1994 in Gehrels (ed.) Smit J & Hertogen J 1980 Nature 285 198 Alvarez L W et al. 1980 Science 208 1095 Astron. Soc. 211 953 Hazards Due to Comets and Asteroids (Univ. Arizona, Soja R H et al. 2011 Mon. Not. Roy. Astron. Soc. Anderson D G et al. 2011 Quaternary Inter­ Courty M-A et al. 2007 Geophys. Res. Abstr. 9 10975 Tucson) 221 414 1059 national 30 1 Courty M-A et al. 2008 in New Trends in Soil Micro­ McGhee G R 1996 The Late Devonian Mass Extinc­ Steel D I 1995 Austral. J. Astr. 6 87 Asher D J 2008 Earth Moon Planets 102 27 morphology (Springer-Verlag, Heidelberg) 211 tion (Columbia University Press) Steel D I 2014 in Keller G & Kerr A C (eds) Geol. Soc. Asher D J 2010 in Andreić Ž & Kac J (eds) Proc. Int. Crutzen P J et al. 1984 Climatic Change 6 323 McNaught R H & Asher D J 1999 WGN J. Int. Amer. Spec. Pap. 505 397 Meteor Conf. 2009 (Int. Meteor Org.) 7 Di Sisto R P et al. 2009 Icarus 20 3 140 Meteor Org. 27 85 Steel D I & Asher D J 1996 in Gustafson B Å S & Asher D J & Clube S V M 1993 Quarterly J. Roy. Dubietis A & Arlt R 2007 Mon. Not. Roy. Astron. Melott A L et al. 2010 Geology 38 355 Hanner M S (eds) Proc. IAU Colloq. 150 (ASP Conf. Astron. Soc. 34 481 Soc. 376 890 Menichella M et al. 1996 Earth Moon Planets Ser. 104) 125 Asher D J & Izumi K 1998 Mon. Not. Roy. Astron. Earth Impact Database 2014 http://www.passc. 72 133 Steel D I & Elford W G 1986 Mon. Not. Roy. Astron. Soc. 297 23 net/EarthImpactDatabase/index.html accessed Minton D A & Malhotra R 2010 Icarus 207 744 Soc. 218 185 Asher D J & Steel D I 1993 Mon. Not. Roy. Astron. 30 September 2014 Moore A M T et al. 2000 Village on the Euphrates Steel D I et al. 1991 Mon. Not. Roy. Astron. Soc. Soc. 263 179 Emel’yanenko V V et al. 2013 Earth Moon Planets (Oxford University Press) 251 632 Asher D J & Steel D I 2012 in Gyssens M & Rogge- 110 105 Napier W M 2001 Mon. Not. Roy. Astron. Soc. Štohl J & Porubčan V 1992 in Ferraz-Mello S (ed.) mans P (eds) Proc. Int. Meteor Conf. 2011 40 Farley K A 2009 in Koeberl C & Montanari A (eds) 321 463 Proc. IAU Symp. 152 (Kluwer, Dordrecht) 315 Asher D J et al. 1994 Vistas Astron. 38 1 Geol. Soc. Amer. Spec. Pap. 452 27 Napier W M 2004 Mon. Not. Roy. Astron. Soc. Sykes M V et al. 1986 Adv. Space Res. 6(7) 67 Asher D J et al. 2001 in Marov M Ya & Rickman Grün E et al. 1985 Icarus 62 244 348 46 Thomas N et al. 2015 Science 347 (6220) aaa0440 H (eds) Collisional Processes in the Solar System Hahn G & Bailey M E 1990 Nature 348 132 Napier W M 2010 Mon. Not. Roy. Astron. Soc. 405 Weiss H & Bradley S 2001 Science 291 609 (Kluwer, Dordrecht) 121 Harris A W & D’Abramo G 2015 Icarus 257 302 1901 Whipple F L 1940 Proc. Amer. Phil. Soc. 83 711 Asher D J et al. 2005 Observatory 125 319 Hartung J B & Storzer D 1974 Proc. Fifth Lunar Sci. Napier W M 2014 in Keller G & Kerr A C (eds) Geol. Whipple F L 1967 in Weinberg J L (ed.) NASA Babadzhanov P B 1998 Cel. Mech. Dyn. Astr. Conf. 3 2527 Soc. Amer. Spec. Pap. 505 383 SP-150 (NASA, Washington DC) 409 69 221 Helin E F & Shoemaker E M 1979 Icarus 40 321 Napier W M 2015 Mon. Not. Roy. Astron. Soc. 448 27 Williams I P 2011 in Cooke W J et al. (eds) Proc. Babadzhanov P B 2001 Astron. Astrophy. 373 329 Horner J et al. 2004a Mon. Not. Roy. Astron. Soc. Napier W M & Asher D J 2009 Astron. Geophys. Meteoroids 2010 (NASA, Huntsville, Alabama) 2 Babadzhanov P B & Obrubov Yu V 1992 Cel. 354 798 50(1) 18 Wittke J H et al. 2013 Proc. Nat. Acad. Sci. 110 E2088 Mech. Dyn. Astr. 54 111 Horner J et al. 2004b Mon. Not. Roy. Astron. Soc. Napier W M & Clube S V M 1979 Nature 282 455 Wright F W & Whipple F L 1950 Harv. Coll. Obs. Babadzhanov P B et al. 1990 Sol. Syst. Res. 24 12 355 321 Nesvorný D et al. 2011 Astrophys. J. 743 129 Tech. Rep. No. 6 Babadzhanov P B et al. 2008 Mon. Not. Roy. Hoyle F 1981 Ice (Hutchinson) Nininger H H 1942 Pop. Astr. 50 270 Zappalà V et al. 1998 Icarus 134 176 Astron. Soc. 386 1436 Hoyle F & Wickramasinghe N C 1978 Astrophys. Oberst J & Nakamura Y 1991 Icarus 91 315 Zook H A 1978 Space Research XVIII 411

6.30 A&G • December 2015 • Vol. 56 • www.astrongeo.com