The scattered disk population as a source of Oort cloud comets: evaluation of its current and past role in populating the Oort cloud

We have integrated the orbits of the 76 scattered disk objects (SDOs), discovered through the end of 2002, plus 399 clones for 5 Gyr to study their dynamical evolution and the probability of falling in one of the following end states: reaching Jupiter's influence zone, hyperbolic ejection, or transfer to the Oort cloud. We find that nearly 50% of the SDOs are transferred to the Oort cloud (i.e., they reach heliocentric distances greater than 20,000 AU in a barycentric elliptical orbit), from which about 60% have their perihelia beyond Neptune's orbit (31 AU<q<36 AU) at the moment of reaching the Oort cloud. This shows that Neptune acts as a dynamical barrier, scattering most of the bodies to near-parabolic orbits before they can approach or cross Neptune's orbit in non-resonant orbits (that may allow their transfer to the planetary region as Centaurs via close encounters with Neptune). Consequently, Neptune's dynamical barrier greatly favors insertion in the Oort cloud at the expense of the other end states mentioned above. We found that the current rate of SDOs with radii R>1 km incorporated into the Oort cloud is about 5 yr⁻¹, which might be a non-negligible fraction of comet losses from the Oort cloud (probably around or even above 10%). Therefore, we conclude that the Oort cloud may have experienced and may be even experiencing a significant renovation of its population, and that the trans-neptunian belt—via the scattered disk—may be the main feeding source.     

The dust trail of Comet 67P/Churyumov-Gerasimenko

We report the detection of Comet 67P/Churyumov-Gerasimenko's dust trail and nucleus in 24 μm Spitzer Space Telescope images taken February 2004. The dust trail is not found in optical Palomar images taken June 2003. Both the optical and infrared images show a distinct neck-line tail structure, offset from the projected orbit of the comet. We compare our observations to simulated images using a Monte Carlo approach and a dynamical model for comet dust. We estimate the trail to be at least one orbit old (6.6 years) and consist of particles of size ?100 μm. The neck-line is composed of similar sized particles, but younger in age. Together, our observations and simulations suggest grains 100 μm and larger in size dominate the total mass ejected from the comet. The radiometric effective radius of the nucleus is 1.87±0.08 km, derived from the Spitzer observation. The Rosetta spacecraft is expected to arrive at and orbit this comet in 2014. Assuming the trail is comprised solely of 1 mm radius grains, we compute a low probability (∼¹⁰⁻³) of a trail grain impacting with Rosetta during approach and orbit insertion.     

Stardust interstellar dust calibration: Hydrocode modeling of impacts on Al‐1100 foil at velocities up to 300 km s−1 and validation with experimental data

Abstract– We present initial results from hydrocode modeling of impacts on Al‐1100 foils, undertaken to aid the interstellar preliminary examination (ISPE) phase for the NASA Stardust mission interstellar dust collector tray. We used Ansys’ AUTODYN to model impacts of micrometer‐scale, and smaller projectiles onto Stardust foil (100 μm thick Al‐1100) at velocities up to 300 km s^?1. It is thought that impacts onto the interstellar dust collector foils may have been made by a combination of interstellar dust particles (ISP), interplanetary dust particles (IDP) on comet, and asteroid derived orbits, β micrometeoroids, nanometer dust in the solar wind, and spacecraft derived secondary ejecta. The characteristic velocity of the potential impactors thus ranges from <<1 to a few km s^?1 (secondary ejecta), approximately 4–25 km s^?1 for ISP and IDP, up to hundreds of km s^?1 for the nanoscale dust reported by Meyer‐Vernet et al. (2009) . There are currently no extensive experimental calibrations for the higher velocity conditions, and the main focus of this work was therefore to use hydrocode models to investigate the morphometry of impact craters, as a means to determine an approximate impactor speed, and thus origin. The model was validated against existing experimental data for impact speeds up to approximately 30 km s^?1 for particles ranging in density from 2.4 kg m^?3 (glass) to 7.8 kg m^?3 (iron). Interpolation equations are given to predict the crater depth and diameter for a solid impactor with any diameter between 100 nm and 4 μm and density between 2.4 and 7.8 kg m^?3.     

Ejecta size-velocity relation derived from the distribution of the secondary craters of kilometer-sized craters on Mars

The relation between the size and velocity of impact crater ejecta has been studied by both laboratory experiments and numerical modeling. An alternative method, used here, is to analyze the record of past impact events, such as the distribution of secondary craters on planetary surfaces, as described by Vickery (Icarus 67 (1986) 224; Geophys. Res. Lett. 14 (1987) 726). We first applied the method to lunar images taken by the CLEMENTINE mission, which revealed that the size-velocity relations of ejecta from craters 32 and 40 km in diameter were similar to those derived by Vickery for a crater 39 km in diameter. Next, we studied the distribution of small craters in the vicinity of kilometer-sized craters on three images from the Mars Orbiter Camera (MOC) on board the Mars Global Surveyor (MGS). If these small craters are assumed to be secondaries ejected from the kilometer-sized crater in each image, the ejection velocities are of hundreds of meters per second. These data fill a gap between the previous results of Vickery and those of laboratory studies.     

Shallow and deep fresh impact craters in Hesperia Planum,Mars

The depths of 109 impact craters 2–16 km in diameter, located on the ridged plains materials of Hesperia Planum, Mars, have been measured from their shadow lengths using digital Viking Orbiter images (orbit numbers 417S–419S) and the PICS computer software. On the basis of their pristine morphology (very fresh lobate ejecta blankets, well preserved rim crests, and lack of superposed impact craters), 57 of these craters have been selected for detailed analysis of their spatial distribution and geometry. We find that south of 30°S, craters <6.0 km in diameter are markedly shallower than similar-sized craters equatorward of this latitude. No comparable relationship is observed for morphologically fresh craters >6.0 km diameter. We also find that two populations exist for older craters <6.0 km diameter. When craters that lack ejecta blankets are grouped on the basis of depth/diameter ratio, the deeper craters also typically lie equatorward of 30° S. We interpret the spatial variation in crater depth/diameter ratios as most likely due to a poleward increase in volatiles within the top 400 m of the surface at the times these craters were formed.     

Structural uplift and ejecta thickness of lunar mare craters: New insights into the formation of complex crater rims

We investigate the elevated crater rims of lunar craters. The two main contributors to this elevation are a structural uplift of the preimpact bedrock and the emplacement of ejecta on top of the crater rim. Here, we focus on five lunar complex mare craters with diameters ranging between 16 and 45 km: Bessel, Euler, Kepler, Harpalus, and Bürg. We performed 5281 measurements to calculate precise values for the structural rim uplift and the ejecta thickness at the elevated crater rim. The average structural rim uplift for these five craters amounts to S_RU = 70.6 ± 1.8%, whereas the ejecta thickness amounts to E_T = 29.4 ± 1.8% of the total crater rim elevation. Erosion is capable of modifying the ratio of ejecta thickness to structural rim uplift. However, to minimize the impact of erosion, the five investigated craters are young, pristine craters with mostly preserved ejecta blankets. To quantify how strongly craters were enlarged by crater modification processes, we reconstructed the dimensions of the transient crater. The difference between the transient crater diameter and the final crater diameter can extend up to 11 km. We propose reverse faulting and thrusting at the final crater rim to be one of the main contributing factors of forming the elevated crater rim.     

Simulations of impact ejecta and regolith accumulation on Asteroid Eros

We have carried out a set of Monte Carlo simulations of the placement of impact ejecta on Asteroid 433 Eros, with the aim of understanding the distribution and accumulation of regolith. The simulations consisted of two stages: (1) random distribution of primary impact sites derived from a uniform isotropic flux of impactors, and (2) integration of the orbits of test particle ejecta launched from primary impact points until their re-impact or escape. We integrated the orbits of a large number of test particles (typically 10⁶ per individual case). For those particles that did not escape we collected the location of their re-impact points to build up a distribution on the asteroid surface. We find that secondary impact density is mostly controlled by the overall topography of the asteroid. A gray-scale image of the density of secondary ejecta impact points looks, in general, like a reduced-scale negative of the topography of the asteroid's surface. In other words, regolith migration tends to fill in the topography of Eros over time, whereas topographic highs are denuded of free material. Thus, the irregular shape of Eros is not a steady-state configuration, but the result of larger stochastic events.     

Visible and near-infrared spectrophotometry of the Deep Impact ejecta of Comet 9P/Tempel 1

We have obtained optical spectrophotometry of the evolution of Comet 9P/Tempel 1 after the impact of the Deep Impact probe, using the Supernova Integral Field Spectrograph (SNIFS) at the UH 2.2-m telescope, as well as simultaneous optical and infrared spectra using the Lick Visible-to-Near-Infrared Imaging Spectrograph (VNIRIS). The spatial distribution and temporal evolution of the “violet band” CN (0-0) emission and of the 630 nm [OI] emission was studied. We found that CN emission centered on the nucleus increased in the 2 h after impact, but that this CN emission was delayed compared to the light curve of dust-scattered sunlight. The CN emission also expanded faster than the cloud of scattering dust. The emission of [OI] at 630 nm rose similarly to the scattered light, but then remained nearly constant for several hours after impact. On the day following the impact, both CN and [OI] emission concentrated on the comet nucleus had returned nearly to pre-impact levels. We have also searched for differences in the scattering properties of the dust ejected by the impact compared to the dust released under normal conditions. Compared to the pre-impact state of the comet, we find evidence that the color of the comet was slightly bluer during the post-impact rise in brightness. Long after the impact, in the following nights, the comet colors returned to their pre-impact values. This can be explained by postulating a change to a smaller particle size distribution in the ejecta cloud, in agreement with the findings from mid-infrared observations, or by postulating a large fraction of clean ice particles, or by a combination of these two.     

Comet nucleus size distributions from HST and Keck telescopes

The Wide Field Camera (WFC) on the Hubble Space Telescope and the Low Resolution Imaging Spectrograph (LRIS) on the Keck II telescope have been used to image 21 distant dynamically new, long-period (LP) and short-period (SP) Jupiter-family (JF) comet nuclei (near aphelion), as part of a long-term program to search for physical differences between short-period comets and Oort cloud comets. WFC data were obtained on Comets C/1987 H1 (Shoemaker) and C/1984 K1 (Shoemaker) during Cycle 5 (1995 December) and on C/1988 B1 (Shoemaker), C/1987 F1 (Torres), and C/1983 O1 (?ernis) during Cycle 6 (1997 April, May, and June). The HST comets were at heliocentric distances 20.4 < r[AU] < 29.5. Each comet observation was allocated 7 orbits, for ≈3.6 hrs of integration. The most difficult part of the image reduction was the removal of cosmic rays. We present our scheme for cosmic ray removal. None of the HST comet nuclei was detected to the 3-σ level at m_R∼27. The inferred upper limits to the nucleus radii are . The SP comets range in radius between , with a median value of R_N∼1.61 km. The LP comets ranged in size between <4.0-56 km. Over a range of radii between 1-10 km, the nuclei can be fit with a cumulative distribution N(>R_N)∝R_N^−α with α=1.45±0.05, and for nuclei in the range 2-5 km, α=1.91±0.06. Statistical analysis and modeling shows that the slopes of the observed TNO and JF comet distributions are not compatible, suggesting that the intrinsic distribution of JF comet nuclei is a differential a^−3.5 power law truncated at small nucleus radii between 0.3 and 2.0 km.     

Petrography and geochemistry of ejecta from the Sudbury impact event

Ejecta from the Connors Creek site in Michigan (500 km from the Sudbury Igneous Complex [SIC]), the Pine River site in western Ontario (650 km from the SIC), and the Coleraine site in Minnesota (980 km from the SIC) were petrographically and geochemically analyzed. Connors Creek was found to have approximately 2 m of ejecta, including shocked quartz, melt droplets, and accretionary lapilli; Pine River has similar deposits about 1 m in thickness, although with smaller lapilli; Coleraine contains only impact spherules in a 20 cm‐thick layer (impact spherules being similar to microkrystites or microtektites). The ejecta transition from chaotic deposits of massively bedded impactoclastic material with locally derived detritus at Connors Creek to a deposit with apparently very little detrital material that is primarily composed of melt droplets at Pine River to a deposit that is almost entirely composed of melt spherules at Coleraine. The major and trace element compositions of the ejecta confirm the previously observed similarity of the ejecta deposits to the Onaping Formation in the SIC. Platinum‐group element (PGE) concentrations from each of the sites were also measured, revealing significantly elevated PGE contents in the spherule samples compared with background values. PGE abundances in samples from the Pine River site can be reproduced by addition of approximately 0.2 wt% CI chondrite to the background composition of the underlying sediments in the core. PGE interelement ratios indicate that the Sudbury impact event was probably caused by a chondritic impactor.     

The impact cratering record of Fennoscandia

The current database of craterform structures in Fennoscandia contains 22 structures of impact origin and about fifty other structures which lack sufficient evidence for impact. The discovery rate of new structures has been one or two per year during the past ten years. The proven impact structures are located in southern Fennoscandia and the majority have been found in Proterozoic target rocks. The age of the structures varies from prehistoric to ≤ 1000 Ma and their diameters (D) from 0.04 km to 55 km. Nine of the structures contain impact melt. A characteristic feature of the Fennoscandian impact record is a relatively large number of small (≤ 5 km) but old (> 200 Ma) structures: this is a result of success of geophysical methods to discover small but old impact structures in an eroded shield covered with relatively thin overburden. Some of the large circular structures in satellite images and/or in geophysical maps may represent deeply eroded scars of very old impacts, but due to the lack of shock metamorphic features, impact-generated rocks or identified ejecta layers, they cannot yet be classified as impact sites. Two huge structures are proposed here as possible impact sites on the basis of circular satellite images and distinct geophysical anomalies: the Lycksele structure in northern Sweden (D ~ 120 km, see also Witschard, 1984) and the Valga structure in Latvia/Estonia (D ~ 180 km). However, endogeneous explanations, like buried granites, basement domings, or fault-bounded blocks are also possible for these structures. Hints, such as distal ejecta layers or impact produced breccia dykes, of an Archaean or Early Proterozoic impact structure have not been found in Fennoscandia so far. New ways of searching for these structures are proposed with particular emphasis on high-resolution integrated geophysical methods. The impact cratering rate in Fennoscandia is ~ 2.0 · 10^?14 km^?2 a^?1 (for craters with D > 3 km) corresponding to about two events per every 100 Ma for the last 700 Ma. Due to erosion, this is a minimal estimate but is higher than the global rate probably due to strong research activity for finding impact structures in Fennoscandia.     

Shocked quartz in polymict impact breccia from the Upper Cretaceous Yallalie impact structure in Western Australia

Yallalie is a ~12 km diameter circular structure located ~200 km north of Perth, Australia. Previous studies have proposed that the buried structure is a complex impact crater based on geophysical data. Allochthonous breccia exposed near the structure has previously been interpreted as proximal impact ejecta; however, no diagnostic indicators of shock metamorphism have been found. Here we report multiple (27) shocked quartz grains containing planar fractures (PFs) and planar deformation features (PDFs) in the breccia. The PFs occur in up to five sets per grain, while the PDFs occur in up to four sets per grain. Universal stage measurements of all 27 shocked quartz grains confirms that the planar microstructures occur in known crystallographic orientations in quartz corresponding to shock compression from 5 to 20 GPa. Proximity to the buried structure (~4 km) and occurrence of shocked quartz indicates that the breccia represents either primary or reworked ejecta. Ejecta distribution simulated using iSALE hydrocode predicts the same distribution of shock levels at the site as those found in the breccia, which supports a primary ejecta interpretation, although local reworking cannot be excluded. The Yallalie impact event is stratigraphically constrained to have occurred in the interval from 89.8 to 83.6 Ma based on the occurrence of Coniacian clasts in the breccia and undisturbed overlying Santonian to Campanian sedimentary rocks. Yallalie is thus the first confirmed Upper Cretaceous impact structure in Australia.     

Pantasma: Evidence for a Pleistocene circa 14 km diameter impact crater in Nicaragua

The circa 14 km diameter Pantasma circular structure in Oligocene volcanic rocks in Nicaragua is here studied for the first time to understand its origin. Geomorphology, field mapping, and petrographic and geochemical investigations all are consistent with an impact origin for the Pantasma structure. Observations supporting an impact origin include outward‐dipping volcanic flows, the presence of former melt‐bearing polymict breccia, impact glass (with lechatelierite and low H₂O, <300 ppm), and also a possible ejecta layer containing Paleozoic rocks which originated from hundreds of meters below the surface. Diagnostic evidence for impact is provided by detection in impact glass of the former presence of reidite in granular zircon as well as coesite, and extraterrestrial ε⁵⁴Cr value in polymict breccia. Two ⁴⁰Ar/³⁹Ar plateau ages with a combined weighted mean age of 815 ± 11 ka (2 σ; P = 0.17) were obtained on impact glass. This age is consistent with geomorphological data and erosion modeling, which all suggest a rather young crater. Pantasma is only the fourth exposed crater >10 km found in the Americas south of N30 latitude, and provides further evidence that a significant number of impact craters may remain to be discovered in Central and South America.     

The nature and emplacement of distal aqueous-rich ejecta deposits from Hale crater,Mars

Hale crater formed in the Early to Middle Amazonian and is one of the best preserved large craters on Mars. We focus on the emplacement of previously mapped distal continuous ejecta and newly recognized discontinuous distal ejecta deposits reaching up to 450 km northeast of Hale. The distal continuous ejecta deposits are typically tens of meters thick, likely water-rich, and subsequent dewatering of some resulted in flow along gradients of 10 m km^-1 for distances of tens of kilometers. The discontinuous distal ejecta are typically <10 m thick with volumes generally <0.5 km³ and embay Hale secondaries, which occur up to ~600 km from Hale. Both continuous and discontinuous distal ejecta deposits are typically smooth at scales of tens to hundreds of meters, relatively dark-toned, devoid of eolian bedforms, inferred to be mostly fine-grained, and were likely emplaced within hours to 1–2 days after impact. The occurrence of well-preserved discontinuous distal ejecta at Hale is unusual compared to other large Martian craters and could be due to impact into an ice-rich substrate that enabled their formation and (or) their survival after minimal postimpact degradation relative to older craters. The pristine nature of distal continuous and discontinuous distal deposits at Hale and the preservation of associated secondaries imply (1) low erosion rates after the Hale impact, comparable to those estimated elsewhere during the Amazonian; (2) the impact did not significantly influence long-term global or regional scale geomorphic activity or climate; and (3) the Hale impact occurred after late alluvial fan activity in Margaritifer Terra.     

Capture of comets from the Oort cloud into Halley-type and Jupiter-family orbits

This paper analyzes the capture of comets into Halley-type and Jupiter-family orbits from the nearparabolic flux of the Oort cloud. Two types of capture into Halley-type orbits are found. The first type is the evolution of near-parabolic orbits into short-period orbits (with heliocentric orbital periods P < 200 years) as a result of close encounters with giant planets. This process is followed by a very slow drift of cometary orbits into the inner part of the Solar System. Only those comets may pass from short-period orbits into Halley-type and Jupiter-family orbits, which move in orbits with perihelion distances q < 13 au. In the second type of capture, the perihelion distances of cometary orbits become rather small (< 1.5 au) during the first stage of dynamic evolution under the action of perturbations from the Galaxy, and then their semimajor axes decrease as a result of diffusion. The capture takes place, on average, in 500 revolutions of the comet about the Sun, whereas in the first case, the comet is captured, on average, after 12500 revolutions. The region of initial orbital perihelion distances q > 4 au is found to be at least as important a source of Halley-type comets as the region of perihelion distances q < 4 au. More than half of the Halley-type comets are captured from the nearly parabolic flux with q > 4 au. The analysis of the dynamic evolution of objects moving in short-period orbits shows that the distribution of Centaurs orbits agrees well with the observed distribution corrected for observational selection effects. Hence, the hypothesis associating the origin of Centaurs with the Edgeworth-Kuiper belt and the trans-Neptunian region exclusively should be rejected.     

Crater geometry and ejecta thickness of the Martian impact crater Tooting

Abstract— We use Mars Orbiter Laser Altimeter (MOLA) topographic data and Thermal Emission Imaging System (THEMIS) visible (VIS) images to study the cavity and the ejecta blanket of a very fresh Martian impact crater ?29 km in diameter, with the provisional International Astronomical Union (IAU) name Tooting crater. This crater is very young, as demonstrated by the large depth/diameter ratio (0.065), impact melt preserved on the walls and floor, an extensive secondary crater field, and only 13 superposed impact craters (all 54 to 234 meters in diameter) on the ?8120 km² ejecta blanket. Because the pre‐impact terrain was essentially flat, we can measure the volume of the crater cavity and ejecta deposits. Tooting crater has a rim height that has >500 m variation around the rim crest and a very large central peak (1052 m high and >9 km wide). Crater cavity volume (i.e., volume below the pre‐impact terrain) is ?380 km³ the volume of materials above the pre‐impact terrain is ?425 km³. The ejecta thickness is often very thin (<20 m) throughout much of the ejecta blanket. There is a pronounced asymmetry in the ejecta blanket, suggestive of an oblique impact, which has resulted in up to ?100 m of additional ejecta thickness being deposited down‐range compared to the up‐range value at the same radial distance from the rim crest. Distal ramparts are 60 to 125 m high, comparable to the heights of ramparts measured at other multi‐layered ejecta craters. Tooting crater serves as a fresh end‐member for the large impact craters on Mars formed in volcanic materials, and as such may be useful for comparison to fresh craters in other target materials.     

Observations of Martian layered ejecta craters and constraints on their formation mechanisms

The formation mechanism of layered ejecta craters on Mars has remained a topic of intense debate since their discovery. In this study, we perform a global morphological analysis of Martian layered ejecta craters using Thermal Emission Imaging System (THEMIS) images and Mars Orbiter Laser Altimeter (MOLA) data. The study focuses on the ejecta morphologies and well‐defined distal rampart characteristics associated with 9945 layered ejecta craters with a diameter greater than 1.5 km distributed across the entire Martian surface. Data analysis based on the new database provides new information on the distribution and morphological details of the three major layered ejecta morphologies (single layer ejecta [SLE], double layer ejecta [DLE], and multiple layer ejecta [MLE]). Global analysis is applied to the latitudinal distribution of characteristic parameters, including the ejecta mobility, lobateness values, and onset diameter. Our survey of the distribution and characteristics of layered ejecta craters reveals that strong correlations exist between ejecta mobility and latitude, and there is a latitudinal dependence of onset diameter. Our study of Martian layered ejecta craters provides more detailed information and insights of a connection between the layered ejecta morphologies and the subsurface volatiles.     

Long-period comet flux in the planetary region: Dynamical evolution from the Oort cloud

This study analyzes the evolution of 2 × 10⁵ orbits with initial parameters corresponding to the orbits of comets of the Oort cloud under the action of planetary, galactic, and stellar perturbations over 2 × 10⁹ years. The dynamical evolution of comets of the outer (orbital semimajor axes a > 10⁴ AU) and inner (5 × 10³ < a (AU) < 10⁴) parts of the comet cloud is analyzed separately. The estimates of the flux of “new” and long-period comets for all perihelion distances q in the planetary region are reported. The flux of comets with a > 10⁴ AU in the interval 15 AU < q < 31 AU is several times higher than the flux of comets in the region q < 15 AU. We point out the increased concentration of the perihelia of orbits of comets from the outer cloud, which have passed several times through the planetary system, in the Saturn-Uranus region. The maxima in the distribution of the perihelia of the orbits of comets of the inner Oort cloud are located in the Uranus-Neptune region. “New” comets moving in orbits with a < 2 × 10⁴ AU and arriving at the outside of the planetary system (q > 25 AU) subsequently have a greater number of returns to the region q < 35 AU. The perihelia of the orbits of these comets gradually drift toward the interior of the Solar System and accumulate beyond the orbit of Saturn. The distribution of the perihelia of long-period comets beyond the orbit of Saturn exhibits a peak. We discuss the problem of replenishing the outer Oort cloud by comets from the inner part and their subsequent dynamical evolution. The annual rate of passages of comets of the inner cloud, which replenish the outer cloud, in the region q < 1 AU in orbits with a > 10⁴ AU (～ 5.0 × 10^?14 yr^?1) is one order of magnitude lower than the rate of passage of comets from the outer Oort cloud (～ 9.1 × 10^?13 yr^?1).     

Generation and emplacement of fine-grained ejecta in planetary impacts

We report here on a survey of distal fine-grained ejecta deposits on the Moon, Mars, and Venus. On all three planets, fine-grained ejecta form circular haloes that extend beyond the continuous ejecta and other types of distal deposits such as run-out lobes or ramparts. Using Earth-based radar images, we find that lunar fine-grained ejecta haloes represent meters-thick deposits with abrupt margins, and are depleted in rocks ?1 cm in diameter. Martian haloes show low nighttime thermal IR temperatures and thermal inertia, indicating the presence of fine particles estimated to range from ∼10 μm to 10 mm. Using the large sample sizes afforded by global datasets for Venus and Mars, and a complete nearside radar map for the Moon, we establish statistically robust scaling relationships between crater radius R and fine-grained ejecta run-out r^* for all three planets. On the Moon, r^* ∼ R^−0.18 for craters 5-640 km in diameter. For Venus, radar-dark haloes are larger than those on the Moon, but scale as r^* ∼ R^−0.49, consistent with ejecta entrainment in Venus’ dense atmosphere. On Mars, fine-ejecta haloes are larger than lunar haloes for a given crater size, indicating entrainment of ejecta by the atmosphere or vaporized subsurface volatiles, but scale as R^−0.13, similar to the ballistic lunar scaling. Ejecta suspension in vortices generated by passage of the ejecta curtain is predicted to result in ejecta run-out that scales with crater size as R^1/2, and the wind speeds so generated may be insufficient to transport particles at the larger end of the calculated range. The observed scaling and morphology of the low-temperature haloes leads us rather to favor winds generated by early-stage vapor plume expansion as the emplacement mechanism for low-temperature halo materials.