Physical Observations of Very Young Dynamical Families of Asteroids

A Proposal submitted to NASA's "Planetary Astronomy" Program

2 June 2006

Principal Investigator: Clark R. Chapman () Southwest Research Institute (SwRI)

Suite 400, 1050 Walnut St.

Boulder, CO 80302

Co-Investigators:William J. Merline (SwRI)

David Nesvorný (SwRI)

Eliot F. Young (SwRI)

I. INTRODUCTION

Synopsis

Extraordinary discoveries by Co-I Nesvorny (Nesvorny et al. 2002a, 2003a, 2006a, Nesvorny & Vokrouhlicky 2006) present a new, stunning opportunity to addresssome fundamental issues in planetary science, including the rates of processes that physically affect asteroids. Here, we propose to continue our carefully coordinated program of astronomical observations of members of 7(up from 3 a couple years ago!) special families of asteroids. Perhaps the most vital but difficult-to-obtain kind of planetary fact is the absolute age of an event. Until now, the only precise dates were from radioisotopic dating techniques (e.g. of moon rocks, meteorites, terrestrial rocks). However, many such ages are compromised since we don't know where the rocks come from -- which unit on Mars, which asteroid, etc. Therefore, most insights about absolute chronologies of planetary bodies have been from indirect techniques, like crater-counting, which provide very crude ages (e.g.± factors of several for the age of Ganymede’s surface). With no good absolute ages available, rates of basic planetary processes remain poorly known.

The brilliant discovery of Nesvorny et al., which we propose to exploit, is a robust, dynamical dating of the formation of asteroid families to a fractional precision generally of only ~1-10%! They find that the Karin cluster (part of the large Koronis family) formed in a catastrophic collision just 5.75 ±0.05 My ago. Also, the Veritas family formed in such a collision 8.3 ±0.1 My ago. And a third cluster, Iannini, is younger than either of them. Just this year, Nesvorny et al. (2006a) and Nesvorny & Vokrouhlicky (2006) identified four new families with even younger ages: between 50 and 600 ky! The Datura, Emilkowalski, 1992YC2, and Lucascavin clusters are 450±50 thousand years (ky), 220±30 ky, 50-250 ky, and 200-600 ky old, respectively (the precision of the last two ages should improve dramatically as new astrometric data become available next year). They are the most recent asteroid breakups ever discovered in the main belt.

Not only are these ages unexpectedly recent (just ~0.005-0.2% of the age of the solar system) but they are known with unprecedented precision. This is the first time that robust, precise, absolute chronological ages can be unquestionably assigned to particular asteroids -- in this case, to members of seven asteroid families. The extraordinary opportunity that we are exploiting is to use these youthful age-determinations to study the rates of fundamental asteroidal processes through a suite of astronomical observations of these objects. For slow processes that have rates much longer than millions-of-years (e.g. orbital evolution of asteroid satellites, evolution of asteroid spins), we have, at last, the opportunity to study initial conditions. For rapid processes that have rates of the order of a few hundred thousands to millions of years or less (e.g. space weathering, loss or covering of volatiles), we have a chance to study the early rates of processes that have gone to completion for most asteroids.

Our research team has extensive experience in observational astronomy of small bodies including (a) noted asteroid authority Clark Chapman (with decades of experience in asteroid spectrophotometry and photometric geodesy, as well as a specialist on space weathering and asteroid families); (b) asteroid satellite expert Bill Merline (who has discovered many asteroid satellites using adaptive optics [AO] on the CFHT, Keck, VLT, and Gemini telescopes and HST); (c) small-body spectroscopist Eliot Young (a frequent observer at IRTF and Keck),and (d) dynamicist David Nesvorný (who made the remarkable theoretical/numerical discoveries that underpin our proposed observing program). We are also assisted by Peter Tamblyn (a specialist in computational astronomy,an IR spectroscopist, and Spitzer P.I.). For many representative members of the seven recently-formed families, we propose coordinated observations of low-resolution visibleand IR spectrophotometry and radiometry (using SpeX and MIRSI at IRTF and LRIS at Keck), medium-resolution spectroscopy (using NIRSPEC at Keck), AO imaging (using VLT, Gemini, or Keck), and lightcurve photometry (using various telescopes in Chileand the American Southwest). Our goal is to search for both predicted and serendipitous indications of youthfulness of these asteroids by obtaining observations that enable us to contrast them with similar data previously published for normal, older main belt asteroids.

Objectives and Expected Significance: Relevance to NASA Strategic Goals3C and PAST

(Consolidation of the required “expected significance” and “relevance to NASA” elements.)

A major element of NASA's Planetary Science Program is to explore and understand small Solar System bodies, and asteroids in particular. They are among the most primitive materials preserved from accretionary epochs (including the random samples known as meteorites). By impacts, they also have been primary modifiers of terrestrial planetary surfaces over the past 3.5 Gy. Since asteroids are numerous, small, widely separated, and amazingly heterogeneous, exploration of their diversity requires major but affordable groundbased observing programs, which supplement occasional Earth-orbital observations, enabling extrapolation of spacecraft results from a very few asteroids to the overall population. Indeed, many recent, underway, or prospective deep space missions address asteroids and other small bodies; but many vital issues -- such as understanding asteroid families (which, by definition, are spread-out groups of numerous bodies) -- require study by Earth-based astronomy.

Our proposed observations address Sub-goals 3C.1 (“learning how...minor bodies originated and evolved”), 3C.2 (“understanding the processes that determine the history and future of habitability in the solar system”), and 3C.4 (“discover potential hazards and...search for resources”) of the 2006 NASA Strategic Plan. We address 6 of the 19 "Research Focus Areas" of the 2003 S.S.E. Roadmap We primarily "study how the processes that determine the characteristics of Solar System bodies operate and interact" (#3) and "determine the physical characteristics of comets and asteroids" (#11); our observations (especially of the Veritas family) also address "the nature, history, and distribution of volatile and organic compounds in the Solar System" (#6), "the initial stages of...satellite formation" pertaining to asteroidal satellites (#1), the "dynamics of bodies that may pose an impact hazard to Earth" (#10), and, by evaluating and thus being able to strip away overprinting by collisions and space weathering, getting closer to understanding the "original characteristics of Solar System bodies" (#2).

Making such observations is central to NASA's PAST Program: "...ground-based astronomical observations...that contribute to the understanding of the general properties and evolution of ... asteroids and comets." Due to funding limitations in PAST, we propose only first-order analysis, interpretation, and synthesis here; we will seek additional funding elsewhere for thorough analysis, after data are obtained and reduced. We expect our observing program to yield exciting, unexpected results for PAST and for the planetary astronomy community. Indeed, our first two years of observing have already yielded an important result (discussed in Sect. V;Fig. 1): widely reported spectral variations on different parts of the young-family member 832 Karin are false – it is spectrally homogeneous!

Background

The fundamental process affecting asteroids in recent aeons is mutual catastrophic collisions. They (a) create asteroid families (clusters of asteroids with similar orbital elements); (b) produce the fragmental size distribution; (c) rearrange large "rubble pile" assemblages; (d) provide delta-v's that help move fragments into chaotic zones, which can radically change their orbits and remove them from the asteroid belt, sometimes into Earth-approaching paths; and (e) establish the distribution of spins, shapes, and configurations (e.g. presence of satellites). Collisional processes also shock, de-gas, or even melt the constituent minerals in asteroidal rocks (meteorites).

Observations of families critically constrain asteroid collisional/dynamical evolution (cf. Arnold 1969; Gradie, Chapman & Williams 1979; Chapman et al. 1989; and chapters in Sect. 5.1 in Asteroids III [Bottke et al., 2002], Nesvorny et al. 2006b). A decade ago we realized that: (a) families are not numerous; (b) most families are very old, hundreds of millions to billions of years; and (c) families generally represent the break-up of homogeneous precursor bodies (spectral properties of family members are usually similar and often distinct from background asteroids). Family attributes (a) & (b) appear compatible with a developing appreciation from hydrocode modeling of asteroid collisions (cf. Benz & Asphaug 1999) that it is difficult to catastrophically fragment, disrupt, and disperse large asteroids, so few families have formed in recent aeons. More recently, we have realized that families are subject to slow spreading and ultimate dispersal due to often-minor resonances in the belt and the Yarkovsky Effect. Therefore, family dimensions and shapes do not reflect outcomes of cratering events or catastrophic disruptions, but rather this dynamical diffusion, often etched by chaotic zones at the borders of, or even within, the families. Therefore, it is vital to study young, just-formed, not-yet-dispersed families to compare with predicted outcomes of hydrocode simulations of collisions.

Now, one of us (Nesvorný) has led research yielding unexpected, startling, but robust results that date seven extremely recent collisional breakups. Now we can observationally test for time-dependent asteroidal phenomena and behavior, which should have evolved little since these families formed. In our current PAST program, we are makinga suite of observations to determine specific characteristics (e.g. spin periods, presence of volatiles, radiometric fluxes) of Karin, Veritas, and Iannini family/cluster members. Among the phenomena with observable attributes or consequences that evolve or mature over time scales of tens of thousands to billions of years, but which might be incomplete after just a few million years, are: space weathering of asteroid surfaces, tidal evolution and collisional destruction of asteroidal satellites, sublimation of large ice pockets that may have survived within the parent asteroid but are now at or near the surfaces of disrupted fragments, and Yarkovsky dispersal of family members. We can also study traits of members of a truly freshly-formed family (e.g. shapes, spins, bias-corrected size distribution) to compare with predicted outcomes of hydrocode simulations of collisional disruption and family formation. Our preliminary inferences from our observations of Karin and Iannini members (see below) are that they represent intermediate cases between pristine and space-weathered surfaces, suggesting that space weathering may progress on even shorter timescales than several My. It is vital to extend spectroscopic observations to the even younger families to explore changes over several 100 ky timescales, so we are now motivated to extend our program to the even younger asteroids in theDatura, Emilkowalski, 1992YC2, and Lucascavin clusters.

Our proposed observations will provide diagnostic tests of issues comparing very-young vs. young vs. old asteroid families, employing visible and infrared spectrophotometry (at both low- and moderate-resolution), radiometry, AO imaging, and extensive lightcurve photometry. The broad array of techniques that we intend to employ may also lead to serendipitous discoveries of unexpected attributes of exceptionally youthful asteroids. We propose to conclude our observations of members of the three 1 – 10 My old clusters (Karin, Veritas, and Iannini, including members discovered in the last few years [there are now 90 known Karin members, up from 39 when we first proposed our observing program]) and observe available targets among the 16 known members of the newly discovered, even younger clusters. We plan to observe perhaps ~150 asteroids in all (obviously fewer for techniques restricted to brighter objects, like AO imaging), including controls. We will usually observe brighter members to reduce integration times, but will spend some time sampling fainter, smaller family members with larger telescopes at adequate S/N to study potential size-dependencies. Although spectroscopic observations of faint asteroids in the new families are challenging, we show below that a reasonable S/N can be obtained with the larger telescopes that we plan to use.

II. TECHNICAL BACKGROUND

A few tens of asteroid families have been found (Milani and Knezevic 1994)from clusters of asteroid positions in the 3-dimensional space of the socalled proper orbital elements (which average over the rapidly varying osculating elements): proper semimajor axis aP, eccentricity eP, and inclination iP. Proper elements are much more constant over time than instantaneous orbital elements; thus a cluster in proper-element space suggests common ancestry. As asteroid discoveries rapidly increase, family identifications have also multiplied. We now realize that members of families gradually drift apart in (aP,eP,iP) space (Bottke et al. 2001, Nesvorný et al. 2002b), so the youngest families are expected to be those that are still compact. Co-I Nesvorný et al. (2002a, hereafter NBDL02) applied the Hierarchical Clustering Method (HCM, Zappalà et al. 1994) to a new, stateoftheart proper element database (Knezevicet al. 2002, to search for compact families with just 1/10th of the typical dispersions of most families. They found the prominent, compact Karin cluster within the Koronis family. The precursor body was evidently near its perihelion when a catastrophic collision and break-up occurred, launching fragments away at ~15 m s-1which then circled the sun as a group. But within ~1000 years, they drifted away from each other around their nearly common orbits. Over longer durations, their orbital orientations (longitude of ascending node, argument of perihelion) also drifted apart, around 360º, due to planetary perturbations. (For still younger clusters, the latter elements do not fully spread.) After only a few million years, the once-clustered asteroids had dispersed into a toroid around the Sun, like the IRAS dust bands (only aP,eP,iP, remain tightly clustered for long durations). By numerically integrating the orbits back in time to an instant when all the orbital elements are clustered, NBDL02 could find the absolute time when the Karin cluster formed: ~5.75 My ago. (The chance that such a convergence could happen by chance during the full age of the solar system is <10-6!) This first-ever dated event must not be confused with the event that created the Koronis family. The age of the Koronis family, based on cratering studies of Ida (a Koronis member) and collisional/dynamical evolution studies, is 2-3 Gy (Chapman 2002, Marzari et al. 1995, Bottke et al. 2001), roughly 400 times older than the Karin-cluster event, which is a very recent secondary disruption of an asteroid within the Koronis family (see also Nesvorny et al. 2006c). Similar searches identified the Veritas family and Iannini cluster as also having very young ages.

Very recently, Nesvorny et al. (2006a) and Nesvorny & Vokrouhlicky (2006) identified four extremely-young asteroid families by using an approach that differs from the traditional family-identification methods. Instead of using a catalog of proper orbital elements they used directly the osculating orbital elements. Because the osculating elements are not constant over My time intervals, they did not expect to find asteroid families older than ~1 My. But extremely young families could show up as clusters in a 5-dimensional space of osculating orbital elements (not includingmean anomaly which becomes dispersed by Keplerian shear within ~10-1000 yr.). Examining orbits of 264,403 asteroids, they identified four new, extremely-young asteroid clusters. Named after the largest members, these are Datura, Emilkowalski, 1992YC2 and Lucascavin clusters (Fig. 2 and Table 1). To determine the formation ages, they propagated the orbits of cluster members backward in time. This method was similar to that used for the Karin cluster in NDBL02 but given the small size of some of these asteroids (most members are 1-3-km in diameter), the computer simulations explicitly accounted for radiation forces like the Yarkovsky Effect. These are the most recent asteroid breakups ever discovered in the main belt.

III. TECHNICAL GOALS AND OBJECTIVES

Because we have determined the absolute ages of seven very recent asteroid collisions, we have the vital chance to study timedependent phenomena that affect asteroids by contrasting observations of these clustered asteroids with more ancient ones. We have designed our observing program to test several predictions or expectations concerning very young asteroids, in addition to looking for serendipitous attributes of these bodies.

Space weathering

Our spectroscopic observations in 2005-6 show that Karin and Iannini cluster members are S-like asteroids (Figs. 1 & 3). The Veritas family members have been known to be C-type asteroids. The taxonomic types of asteroidsin the Datura, Emilkowalski, 1992YC2 and Lucascavin clusters are not known, but could beS-types. Space weathering is the phenomenon, possibly due to micrometeorite impact and solar wind impingement, that modifies spectral reflectances from that of the inherent mineral assemblages. Many S-types, may be ordinary chondritic (OC) in composition, but the diagnostic spectral shape and depth of absorption bands are space weathered from those typical of OCs into a typical S-type spectrum, characterized by a reddish continuum and shallower absorption bands (cf. Chapman 2004). Binzel et al. (1996) observed a full range of spectra among small Near Earth Asteroids (NEAs), from OC to traditional S-type, suggesting that there is a range of collisional ages, with recently "freshened" surfaces looking like unmodified OC meteorites (from lab spectra) while progressively older asteroids are more maturely space weathered (reddened slope, shallower 0.9 and 2.0 μm bands).

Published estimates of space-weathering timescales range widely from 50 ky to 100 My (Hapke, 2001; Sasaki et al. 2001), so it is vital to understanding space weathering processes to determine the degree to which spectral evolution has matured on members of these families, ranging from ~200 ky to 5.8 My old. By studying members within a single family of different sizes, we can even test whether space weathering rates are a function of body size, as expected if the processes are mediated by regolith processes, which in turn depend on a body's gravity. Such processes must affect C-type asteroids, as well, although spectral effects may be smaller,or even opposite (Nesvorny et al. 2005), than for S-types (cf. Sect. 7.2 of Rivkin et al. 2002); so we will continue to study C-type Veritas family members. Indeed, this is a perfect opportunity to understand how space-weathering differently affects C- and S-types, calibrating the early changes before maturity is reached. Space weathering processes are believed to be material-dependent, so we are fortunate to have compositionally distinct young families to study. Our studies will augment the partial understanding of space weathering already achieved from spacecraft investigations of Ida and Eros (Chapman 1996, Clark et al. 2002); e.g., we may be able to date the formation of the fresh crater Azzura on Eros from its color.