An Essay...Last updated 25 August 2000
NEAR'S NEW VIEW OF EROS:
IMPACT PROCESSES AND SPACE WEATHERING
25 August 2000
Meteoritical Society, Chicago, August 2000
Clark R. Chapman
Southwest Research Institute, Boulder CO
The NEAR MSI/NIS Team
Abstract. Eros has been found to resemble Ida in almost all respects (e.g. saturated cratering, S(IV) spectral type, bulk density, and probable ordinary chondritic composition) except for one: it lacks the prominent color differences seen on Galileo's multispectral images of Ida. The latter have been ascribed to recent, fresh units (impact penetration to bedrock and irregular distribution of fresh ejecta around Ida) which contrast with the redder coloration of most of Ida's surface, explained by an as-yet-unidentified "space weathering" process that reddens Ida's surface with time. Eros exhibits the same reddened coloration as Ida, presumably due to space-weathering inasmuch as NEAR's X-ray compositional measurements indicate that Eros has an L- or LL-chondrite-like composition which should not appear to be so red. Here we attribute the lack of any fresh units having ordinary chondrite-like colors to a plausible ~50 Myr hiatus in cratering of Eros following its dynamical removal from the impact environment of the main asteroid belt, during which space weathering has reached maturity everywhere on the body. Other possible implications of such a cratering hiatus are also discussed, including implications for the origin of the numerous boulders on Eros and the possible role of usually minor geological processes in shaping the geomorphology of Eros.
Some fundamental issues in planetary science depend on identifying meteorite parent bodies. There has been a wide presumption that ordinary chondritic and carbonaceous chondritic material represents the predominant material in the inner solar system at the time of planetary accretion. In addition, the differentiated meteorites show that an uncertain fraction of primordial material was heated to the point of melting at early times. So it remains uncertain whether non-differentiated chondritic material (as represented in our meteorite collections) is representative of the materials from which the planets accreted.
One of the longest standing issues in asteroid science has been the nature of S-type asteroids, the second most common type of asteroid, and their relationship to meteorites. In particular, there has been debate about whether (a) meteorites are a highly selective, non-representative sample of inner-main-belt asteroids and that the apparent spectral dissimilarity between S-types and the common ordinary chondrites (OC's) means that the common S-type asteroids are parents of rare or non-sampled meteorites (like the differentiated stony-irons and rare achondrites) or (b) the sampling is roughly representative and the spectral dissimilarity between the common S-types and common OC's means that non-differentiated asteroidal surfaces are subject to some kind of "space weathering" process that modifies the OC spectral traits observed in the laboratory to appear S-like.
Abundant evidence assembled in the last five years has swung the pendulum strongly in favor of the second option -- that some or many S-types are space-weathered OC's -- including analysis of spatial variations in the reflectance spectrum of Ida (Chapman, 1996), analysis of the systematics of spectral data of Earth-approaching S-types (Binzel et al., 1996), and laboratory simulations of plausible space weathering processes (Moroz et al., 1996; Yamada et al., 1999). Nevertheless, the arguments have been indirect and remain debatable.
The NEAR-Shoemaker mission to Eros has provided the first opportunity to take the analysis of an S-type's composition to the next level. NEAR-Shoemaker's complement of instruments was designed to measure an independent suite of remotely sensed indicators of mineralogical and chemical composition with sufficient precision to distinguish between the major meteorite types. The measurement techniques are X-ray, γ-ray, near-IR spectroscopy, and multi-spectral imaging, in addition to constraints provided for bulk density from the radio science, laser-ranging, and imaging experiments. In addition, NEAR-Shoemaker was planned to achieve unprecedented spatial resolution for multispectral imaging and near-IR spectroscopy by orbiting Eros at much closer distances than were reached by the distant Galileo flybys of the S-type asteroids Gaspra and Ida. Compositional variations across the surface of an asteroid can be especially diagnostic, as first demonstrated by Gaffey (1984) from spectral differences resolved only hemispherically; Chapman (1996) later studied compositional variations on Ida at Galileo's much higher resolution.
Needless to say, NEAR-Shoemaker's results for Eros represent only a single asteroid within the very large and heterogeneous S-type class and cannot be blithely applied to all other S-types. However, Eros represents a particularly pertinent test of the inferences previously made from groundbased studies. On the one hand, the mean groundbased reflectance spectrum of Eros appears to be in the S(IV) subclass of Gaffey et al. (1993), thought at the time to be more likely (or least unlikely) to be OC's than any other S-type subclass. On the other hand, reinterpretation of groundbased spectra of Eros by Murchie & Pieters (1996) revealed possible hemispheric variations that were considerably greater than would be expected for a homogeneous, non-differentiated parent body (i.e. outside of the bounds of such known variations within one type of OC as the range of petrologic class).
Partway through NEAR-Shoemaker's orbital mission, it has become fairly clear what the final picture of Eros' mineralogy will be. Its evident spatial homogeneity in spectral properties, as observed by both the Near-Infrared Spectrometer (NIS) and especially by the Multispectral Imager (MSI), as well as preliminary chemical data from the X-ray detector (thanks to energetic solar flares accompanying solar maximum) both point away from any differentiated composition and toward an OC composition (in particular, Trombka  stated that Eros resembles either an L or LL chondrite in the major-element chemistry to which X-ray fluorescence is sensitive). (Some depletions, relative to chondritic, in sulfur and possibly other elements have been noted, perhaps indicating partial melting, but global differentiation remains ruled out.) The OC interpretation for the composition of Eros is consistent with the 2.67 gm/cm^3 bulk density of Eros and with preliminary evidence that the density is homogeneous, given L- or LL-like densities and a modest fraction of void space.
Even if one accepts, at this preliminary stage, the evidence suggesting that Eros is a low-iron OC, or a strongly heated L- or LL-like assemblage with some partial melting, there remains a striking puzzle. To date, neither the NIS data nor the MSI images indicate any significant (to within 2 or 3 percent) color variations across the surface of Eros that can be ascribed to intrinsic mineralogical differences. This statement appears to be true down to the spatial resolutions that had been achieved by mid-May when the NIS went out of service -- roughly a couple of km for NIS and many tens of meters for MSI. (Some very slight color differences have been seen, but most probably relate to photometric effects. There are hints of correlations between geology and very minor color variations.)
Such spatial uniformity, however, would not have been expected on the basis of earlier observations of Eros and, especially, based on the experience of Ida. It is plausible that the apparent hemispheric color differences seen in groundbased data (Murchie & Pieters 1996) were not real inasmuch as they were near the systematic error limits. However, the tens-of-percent color variations on Ida presented a reasonable picture (described by Chapman, 1996) of an OC body, subject both to space weathering and impacts, which would seem to be applicable to Eros. Recent impacts would excavate fresh OC material (with observed spectral traits approaching OC), which would then be subject to space weathering with time, gradually maturing into S-type (reddened) colors. A goal of this essay is to analyze why this is not evidently true for Eros. If Eros is truly of OC composition, then its S-type (rather than OC-like) colors must reflect some generic form of space weathering. Yet it shows no evidence for such in situ evolution of OC-like colors in recently formed units to space-weathered colors for older geological units. What's going on?
IMPACT AND SPACE-WEATHERING ENVIRONMENT OF EROS
Ida presents the most reasonable analog for Eros, rather than Gaspra, because Galileo's observations of Ida were more comprehensive, and at better resolution, than for Gaspra. Moreover, Gaspra's appearance (geologically and spectrally) is dissimilar from Eros in that it has a low crater density and its spectrum is too olivine-rich to be OC. Indeed, Chapman (1997) concluded that Gaspra was probably a metal-rich, differentiated object. Despite the fact that Gaspra exhibits modest color differences across its surface, it is therefore unlikely that it would be a good analog for Eros.
Ida, on the other hand, is also an S(IV) with a bulk density very similar to that of Eros. If Chapman's (1996) conclusion is correct that Ida is OC-like in composition, then it is a very good analog for Eros. In the case of Ida, major color variations across its surface have been modelled (Geissler et al., 1996) as ejecta from the largest (~7 km diameter) recent impact crater on Ida, named Azzurra. The spectral reflectances of both Azzurra and its irregularly distributed ejecta are much more OC-like than most of Ida (Chapman, 1996); Sullivan et al. (1996) and Chapman (1996) have also pointed out that some other small, fresh craters on Ida exhibit spectra less than half-way evolved from OC to S-like. The hypothesis is, therefore, that recent impacts have penetrated the surficial, gardened regolith to bedrock, have excavated OC-like materials, and have ballistically distributed those materials in patches around the body; they have not had time to reach spectral maturity.
The age of Azzurra is uncertain, of course. Chapman et al. (1996a) estimate Ida's age as ~2 x 10^9 years. Azzurra is about the 6th or 7th largest crater recognized on Ida and is probably the morphologically freshest of those. It also may be fresher than any other crater on the well-imaged side of Ida down to 1.5 to 2 km diameter. Although Azzurra could have formed very recently, it is more reasonable to expect that it has an age of roughly 100 Myr, a small fraction of Ida's average age. A rather small percentage (perhaps <5%) of prominent, small craters on the well-imaged side of Ida exhibit the prominent color differences that indicate that space weathering has not yet gone to completion, again implying a timescale for space weathering of somewhat less than 100 Myr.
My hypothesis for Eros is that its spectral uniformity reflects a recent history in which space weathering has continued but during which there has been a hiatus in large-scale cratering. While "space weathering" is a generic term not implicating any particular physical process for the inferred modification with time of the spectral properties of asteroidal bedrock to S-type characteristics, there exist several hypotheses about physical mechanisms. The most promising source of energy to modify the surface grains is instantaneous zapping by micrometeorites and, perhaps, by solar wind particles. What it is, precisely, that changes the chemistry and optical properties of affected minerals like olivine is the subject of continuing research. In any case, such processes ought to be widespread in the solar system, both in the asteroid belt and in regions closer to the Sun where Eros now orbits. Even asteroidal dust, generated in the asteroid belt, spreads inward in the solar system due to radiation forces. Indeed, according to some concepts of space weathering processes, they should be more efficient closer to the Sun, where there is a greater flux of solar wind particles and where impact velocities are higher. Those may be two factors (among others, including the repetitive gardening of the lunar regolith due to the Moon's higher gravity) that result in the much more extreme space weathering effects on the Moon; space weathering has long been held to be responsible for the extreme differences between laboratory spectra of lunar rocks and regionally-averaged spectra of the lunar surface.
Macroscopic cratering, on the other hand, operates at a radically greater rate for bodies within the main asteroid belt than for those that orbit wholly interior to the belt. The cratering rate is down roughly three orders of magnitude in the latter case. So my simple idea is that Eros has been decoupled from asteroidal cratering for the last several percent of its lifetime (say 10 to 100 Myr) so that space weathering has had a chance to mature on even the most recently formed geological units that are resolvable by the NEAR-Shoemaker spectral instruments.
It is not possible to deterministically trace the past orbital history of Eros because of chaotic dynamics. Presently, Eros' aphelion distance of 1.78 AU restricts it from interacting with almost any main-belt asteroid, excepting only those few near the inner edge of the belt that have unusually high eccentricities. But in the past, the orbit of Eros has surely undergone major variations since Eros was first derived from the asteroid belt. Numerical simulations of sixteen Eros clones by Michel et al. (1998) illustrate the most likely orbital behaviors for Eros; Michel et al. also addressed, in advance of NEAR-Shoemaker's arrival at Eros, the possible impact history of Eros.
Let us consider the plausible dynamical history of Eros, paying special attention to the implied cratering history. First, it should be noted (as is obvious from all recent NEAR-Shoemaker images of Eros) that Eros is saturated with craters of diameters >100 m, with a crater frequency distribution very similar to that on Ida (see figure). This suggests that Eros was cratered in the asteroid belt (not in an orbit mostly decoupled from the belt). Moreover, it must have been exposed to cratering, subsequent to its "creation" by catastrophic fragmentation of a parent body, for a time appreciably longer than the 200 Myr estimated cratering age of undersaturated Gaspra (Chapman et al., 1996b), perhaps 1 Gyr.
Zappalà et al. (1997) proposed that the catastrophic break-up of a sizeable parent asteroid that created the Maria family might be the origin for the large near-Earth asteroids Eros and Ganymed. In their scenario, however, Eros would have been placed in the 3:1 resonance immediately and begun its orbital evolution quickly. Indeed, it probably would have evolved in a few million years to crash into the Sun (90% die within 11 Myr, as discussed by Gladman et al, 1997, and Migliorini et al., 1998). Such quick evolution seems to be incompatible with the very long times expected between such major family-producing collisions. If we were lucky enough that the Maria family formed less than a few tens of Myr ago so that we can be observing Eros during its brief history as an Earth-approacher, we would not expect it to be heavily cratered at all.
The simulation of sixteen Eros clones by Michel et al. (1998) presents a very different picture of Eros' likely origin. Eros is in a type of orbit that is unusually long-lived for Earth-approaching asteroids, with a typical lifetime of 50 - 100 Myr before solar crash, ejection from the solar system by Jupiter, or less-likely impacts with the terrestrial planets. Rather than being derived from the immediate vicinity of a resonance, objects like Eros are more likely derived from more typical parts of the main asteroid belt; then they slowly diffuse (due to numerous minor resonances) to become Mars-crossing (Migliorini et al., 1998). Thus Eros is expected by Michel et al. to have been formed "long before its eventual insertion into the current Mars-crossing orbit." The bulk of its cratering presumably occurred before it became Mars-crossing, yet there is the possibility that its orbital aphelion continued to remain in the asteroid belt, in which case it might have continued to be cratered during the last several tens of Myr.
Michel et al. (1998) have calculated the average intrinsic collision probabilities with main-belt asteroids of their sixteen clones during the 5 Myr integrations and also the mean impact velocities. What they find is that individual clones range from having impact histories comparable to that of an average main-belt asteroid (3 of the 16 clones) to having cratering rates down by two orders of magnitude (half of the clones).
If Eros, like half of the clones, has had a history in which its cratering rate has been depressed by a factor of ~100 during the last many tens of millions of years, then we have a natural explanation for its spectral homogeneity. If Eros has existed for 1 Gyr and has been in something resembling its present orbit for 50 Myr, then it was able to receive only 0.05% of its cumulative cratering during the last 50 Myr. Such a history would virtually rule out the formation of any Azzurra-like crater on its surface. Moreover, the largest crater expected to have been formed on Eros during the last 50 Myr would be under 0.5 km diameter, which might well fail to penetrate regolith and excavate bedrock, thus explaining the lack of freshly excavated materials like those observed on Ida.
Above I estimated that the timescale for space-weathering on Ida is less than 100 Myr. Assuming that the rate is at least that rapid for Eros, which is closer to the Sun and in a higher-velocity micrometeorite collisional environment, then it is very plausible that the most recent excavations of fresh material just prior to Eros' orbital removal from the main-belt collisional environment would have had time to become maturely space-weathered by now.
DISCUSSION AND CONCLUSIONS
We normally think of asteroids as being modified almost exclusively by ongoing collisions and cratering. We think of most features of asteroids as being created and destroyed in a quasi-steady-state fashion, occasionally spiked by the stochastic effects of a rare, very large impact that may suddenly generate ejecta blankets, boulders, cracks, etc. Eros, however, may well have been in a virtual hiatus in cratering activity for the last tens of millions of years.
Therefore, we see a tableau of Eros as it existed tens of Myr ago when, by slow dynamical processes, it was gradually extricated from the collisional environment in which it had existed for perhaps 95% of its existence. A corollary of such a recent hiatus, however, is that other processes that are normally overwhelmed by the usual cratering rates suddenly are augmented in relative efficacy by two orders of magnitude. We need to think about what such processes might be.
One example has been discussed at length in this essay: space-weathering processes, caused perhaps by inner solar system micrometeoroids and solar wind particles impacting the surface, have a chance to mature with virtually none of the usual competing regolith gardening processes due to macroscopic impactors.
Mass wasting and readjustment processes, usually overwhelmed by cratering, might become manifest during a long cratering hiatus. For example, thermal cycling continues as Eros rotates every five hours and the small expansions and contractions might yield accumulated downslope movement (thermal creep) and other manifestations that we usually don't see or think about. If Eros undergoes polar wandering, as has been suggested due to its nearly identical b and c moment of inertias, changing magnitudes and directions of stress within the body might even yield observable tectonic features that would cross-cut virtually all impact craters.
Such enhanced endogenic geological processes might explain an unexpected observation concerning craters on Eros (see Figure). Before NEAR-Shoemaker descended to low orbits, modest resolution images revealed a crater size-frequency distribution very much like that seen on Ida and on the lunar surface. When craters much smaller than 100 meters diameter were imaged, however, it became apparent that such smaller craters are increasingly rare -- an order-of-magnitude less than empirical saturation at 20 meter diameters. (Indeed boulders become more frequent than craters at diameters smaller than 20 meters!) This is extraordinarily unexpected, and radically different from what is observed on the Moon. The highly degraded morphologies of most of these smaller craters suggest an erosional or blanketing process -- uncoupled from cratering -- has degraded and destroyed smaller craters. The enhanced relative magnitude of endogenic processes, described above, during a cratering hiatus could explain these unexpected data. Alternatively, a major impact event could have generated blanketing ejecta shortly before the cratering hiatus and there has been little or no re-cratering since.
Conceivably a small fraction of crater ejecta blocks wind up in temporary orbits around asteroids, later leaking into heliocentric orbits or re-impacting the asteroid. In the usual situation, such re-impacts compete at a low level with the continued cratering of the surface by asteroidal projectiles and by low-velocity crater ejecta. However, during the hypothesized hiatus in cratering, the decay and re-impact of blocks orbiting around Eros might continue for an appreciable time (depending on the timescale for decay) and could preferentially contribute to the most recent features on the surface of Eros. For instance, it is possible that a significant fraction of the numerous blocks found on Eros might be reimpacted satellites which have not, as would normally be the case, been covered over or destroyed by subsequent impacts and regolith evolution. (The thoughts in this paragraph are being developed in collaboration with Bill Merline.)
Insofar as issues of space weathering and the composition of Eros and S-type asteroids are concerned, this essay primarily demonstrates that there is a natural explanation for what otherwise would seem to be a disconnect between Eros and our experience with Ida. Now, it can be understood that Ida and Eros might be very similar bodies in all respects, excepting only that Eros has been removed from continued cratering very recently in its history. Quantification of these ideas might even lead to estimates of the relative rates of space weathering on Eros and Ida (which could even lead to clues about what the predominant space weathering process is); and it might lead to constraints on the type of dynamical orbit (among the classes defined by Milani et al, 1989 and discussed by Michel et al., 1998) that Eros has been in; that, in turn, might shed further light on Eros' place of origin within the main asteroid belt.
Binzel, R.P., S.J. Bus, T.H. Burbine & J.M. Sunshine 1996. Spectral properties of near-Earth asteroids: Evidence for sources of ordinary chondrite meteorites. Science 273, 946-948.
Chapman, C.R. 1996. S-type asteroids, ordinary chondrites, and space weathering: The evidence from Galileo's fly-bys of Gaspra and Ida. Meteoritics and Planet. Sci. 31, 699-725.
Chapman, C.R. 1997. Gaspra and Ida: Implications of spacecraft reconnaissance for NEO issues. In Near-Earth Objects: The United Nations Conference (ed. J. Remo, Ann. N.Y. Acad. Sci. 822), 227-235.
Chapman, C.R., E.V. Ryan, W.J. Merline, G. Neukum, R. Wagner, P.C. Thomas, J. Veverka & R.J. Sullivan 1996a. Cratering on Ida. Icarus 120, 77-86.
Chapman, C.R., J. Veverka, M.J.S. Belton, G. Neukum & D. Morrison 1996b. Cratering on Gaspra. Icarus 120, 231-245.
Gaffey, M.J. 1984. Rotational spectral variations of asteroid (8) Flora: Implications for the nature of the S-type asteroids and for the parent bodies of the ordinary chondrites. Icarus 60, 83-114.
Gaffey, M.J., J.F. Bell, R.H. Brown, T.H. Burbine, J.L. Piatek, K. Reed & D.A. Chaky 1993. Mineralogical variations within the S-type asteroid class. Icarus 106, 573-602.
Geissler, P., J-M. Petit, D.D. Durda, R. Greenberg, W. Bottke & M. Nolan 1996. Erosion and ejecta reaccretion on 243 Ida and its moon. Icarus 120, 140-157.
Gladman, B.J., F. Migliorini, A. Morbidelli, V. Zappalà, P. Michel, A. Cellino, Ch. Froeschlé, H.F. Levison, M. Bailey & M. Duncan 1997. Dynamical lifetimes of objects injected into asteroid belt resonances. Science 277, 197-201.
Michel, P., P. Farinella & Ch. Froeschlé 1998. Dynamics of Eros. Astron. J. 116, 2023-2031.
Migliorini, F., P. Michel, A. Morbidelli, D. Nesvorný & V. Zappalà 1998. Origin of multikilometer Earth- and Mars-crossing asteroids: A quantitative simulation. Science 281, 2022-2024.
Milani, A., M. Carpino, G. Hahn, A.M. Nobili 1989. Dynamics of planet-crossing asteroids: Classes of orbital behavior. Icarus 78, 212-269.
Moroz, L.V., A.V. Fisenko, L.F. Semjonova, C.M. Pieters & N.N. Korotaeva 1996. Optical effects of regolith processes on S-asteroids as simulated by laser shots on ordinary chondrite and other mafic materials. Icarus 122, 366-382.
Murchie, S.L. & C.M. Pieters 1996. Spectral properties and rotational spectral heterogeneity of 433 Eros. J. Geophys. Res--Planets 101, 2201-2214.
Sullivan, R., R. Greeley, R. Pappalardo, E. Asphaug, J.M. Moore, D. Morrison, M.J.S. Belton, M. Carr, C.R. Chapman, P. Geissler, R. Greenberg, J. Granahan, J.W. Head III, R. Kirk, A. McEwen, P. Lee, P.C. Thomas & J. Veverka 1996. Geology of 243 Ida. Icarus 120, 119-139.
Trombka, J. 2000. Presentation at American Geophys. Union meeting, Washington D.C., May 2000.
Yamada, M., S. Sasaki, H. Nagahara, A. Fujiwara, S. Hasegawa, H. Yano, T. Hiroi, H. Ohashi & H. Otake 1999. Simulation of space weathering of planet-forming materials: Nanosecond pulse laser irradiation and proton implantation on olivine and pyroxene samples. Earth Planets Space 51, 1255-1265.
Zappalà, V., A. Cellino, M. DiMartino, F. Migliorini & P. Paolicchi 1997. Maria's family: Physical structure and possible implications for the origin of giant NEAs. Icarus 129, 1-20.
FIGURE CAPTION. This is an R-plot (differential size-frequencies divided by D^-3) showing the spatial densities of craters and boulders on Eros larger than 10 meters diameter. Craters larger than 200 meters diameter resemble the empirically saturated frequencies of craters on Ida; but smaller craters are deficient. On the other hand, small boulders are very numerous.
"Space weathering," as used in this essay, refers to an unknown process or processes that modifies the optical reflectance properties (especially spectral reflectance) of surface materials with time. That such spectral changes with time happen has long been well-documented for the Moon, and more recently for asteroids. While the process/es are believed to be due to exposure of the surface materials to space, and hypotheses of specific processes have been proposed, the details of what causes space weathering is of secondary concern for this essay so long as it occurs for bodies in near-Mars and near-Earth space.
Back to NEAR/Eros Page.
Available Abstracts, Preprints, Articles.
Clark R. Chapman's Publications.
Clark R. Chapman's Homepage.