您当前的位置: 首页 > 网页快照
Carbon monoxide gas produced by a giant impact in the inner region of a young system - Nature
Abstract . Models of terrestrial planet formation predict that the final stages of planetary assembly—lasting tens of millions of years beyond the dispersal of young protoplanetary disks—are dominated by planetary collisions. It is through these giant impacts that planets like the young Earth grow to their final mass and achieve long-term stable orbital configurations 1 . A key prediction is that these impacts produce debris. So far, the most compelling observational evidence for post-impact debris comes from the planetary system around the nearby 23-million-year-old A-type star HD?172555. This system shows large amounts of fine dust with an unusually steep size distribution and atypical dust composition, previously attributed to either a hypervelocity impact 2 , 3 or a massive asteroid belt 4 . Here we report the spectrally resolved detection of a carbon monoxide gas ring co-orbiting with dusty debris around HD?172555 between about six and nine astronomical units—a region analogous to the outer terrestrial planet region of our Solar System. Taken together, the dust and carbon monoxide detections favour a giant impact between large, volatile-rich bodies. This suggests that planetary-scale collisions, analogous to the Moon-forming impact, can release large amounts of gas as well as debris, and that this gas is observable, providing a window into the composition of young planets. You have full access to this article via your institution. Download PDF Download PDF Main . HD?172555 is a (23?±?3)-Myr-old (ref. 5 ) A-type star with a mass of 1.76? M ☉ (ref. 6 ) and a luminosity of 7.7? L ☉ , located 28.5?pc (refs. 7 , 8 ) from Earth within the young β?Pictoris moving group 9 . Its planetary system hosts large amounts of dust in the terrestrial region, producing an infrared (IR) excess best fit by warm (290?K) dust 10 with an IRluminosity 7.2?×?10 ?4 times that of its host star 4 . Spatially resolved mid-IR observations constrain this dust to a disk of material in the inner, less than 10? au region, viewed close to edge-on from Earth 11 , 12 . Rare solid-state emission features detected at 8–9.3?μm (ref. 13 ) indicate the presence of glassy silica (tektites and obsidian) and solid silicon monoxide. These require high-temperature processing and vapour condensation, supporting a hypervelocity (>10?km?s ?1 ) impact between planetary bodies for the origin of the dust 2 , 3 . Detailed, self-consistent modelling indicates that this impact scenario can explain the spectral features, as well as the overabundance of submicrometre-sized grains and steep size distribution inferred from the Spitzer IR spectrum 3 . We used the Atacama Large Millimeter/submillimeter Array (ALMA) to detect the dust continuum and carbon monoxide (CO) J ?=?2–1 (where J is the total angular momentum quantum number) gas emission from the HD?172555 system (Fig. 1 ). The compact, spatially unresolved dust and gas emission originates from within 15? au of the star, consistent with previous IR and optical observations. The spectrum of CO in Fig. 2a is extracted from the ALMA data cube and is spectrally resolved. The double-peaked profile is expected from gas orbiting in Keplerian rotation around the central star. The centroid of the CO profile is at a heliocentric velocity of 2.3?±?0.2?km?s ?1 , consistent with the stellar velocity 14 , confirming that the gas emission is associated with the HD?172555 system. Fig. 1: Cleaned emission maps of the HD?172555 system. Left: 4 σ detection of the dust continuum emission?(surface brightness I ν ). Right: the moment-0 (spectrally integrated surface brightness I ) map of the?CO J ?=?2–1 transition, with 9 σ detection. Contour levels are at 2 σ intervals, with σ being 0.029?mJy per beam and 15?mJy?km?s ?1 per beam for the continuum and CO moment-0 maps, respectively. The beam size is denoted in lower left corner of each panel. Note that the 2 σ peaks are background noise and not significant. Full size image Fig. 2: Mass–temperature degeneracy of the CO data. a , CO J ?=?2–1 spectrum of HD?172555 (grey solid line), compared with the output of two radiative transfer models: a low-mass, warm-temperature CO model (red dashed line), and a high-mass, cold-temperature CO model (blue dotted line). b , χ 2 map (where χ 2 is relative to the best-fit model) showing a clear CO temperature–mass degeneracy for models that are good fits to the data (darker on the colour scale). The red and light blue dots correspond to the red (warm) and blue (cold) CO models shown in a . Dotted contours enclose models that are consistent with the data at the >[1,2,3,4] σ confidence level. Full size image We modelled the CO emission as a vertically thin ring of optically thin gas with radially and azimuthally uniform surface density between an inner and outer boundary, in Keplerian rotation around the central star. We find that the CO is confined to a ring of radius about 7.5? au and width about 3.3? au . The data are consistent with a symmetric disk model; we find no strong evidence of asymmetry between the blue-shifted and red-shifted sides of the CO spectral line profile. We convert the integrated flux of 120?mJy?km?s ?1 to a CO gas mass considering non-local thermodynamic equilibrium excitation, finding masses in the range between 0.45?×?10 ?5 M ? and 1.21?×?10 ?5 M ? for temperatures between 100?K and 250?K. The total millimetre dust mass, as calculated from the observed continuum emission, is (1.8?±?0.6)?×?10 ?4 M ? for an expected equilibrium blackbody temperature of 169?K at 7.5? au . To test our assumption that the CO J ?=?2–1 emission is optically thin, we create a full three-dimensional radiative transfer model of the ring with the spatial morphology described above. For a gas temperature comparable to, or higher than, the expected blackbody temperature of 169?K at 7.5? au , we find low optical depths, which validates our optically thin assumption. This corresponds to total CO masses of (1.4?±?0.3)?×?10 ?5 M ? at 169?K. However, we find that colder (<<100?K) temperatures would require higher optical depth and much higher CO masses to reproduce our observed CO line emission (Fig. 2b ). This temperature–mass degeneracy can be resolved by the detailed spectral shape of the emission as models of optically thin, warm gas constrained to a narrow ring produce sharper peaks compared with the optically thick, cold models (Fig. 2a ). We find that the data are better fit by less massive, warmer models (lower χ 2 ). The observed CO gas in the circumstellar environment around HD?172555 will be subject to photodissociation from the stellar and interstellar ultraviolet (UV) field. The former dominates at the 7.5? au location of the CO ring and causes CO (if unshielded) to be rapidly destroyed on a timescale of around one day. However, the lifetime of CO is extended when taking into account shielding effects from atomic carbon 15 (C, as produced by CO photodissociation), molecular hydrogen (H 2 ) and CO itself (self-shielding), which prevent UV photons from penetrating into the ring. Self-shielding alone, in the warm-temperature, low-CO-mass scenario, extends the gas lifetime by around two to four years. Given the lack of observational constraints on C and H 2 column densities, we cannot definitively estimate the shielding and hence the CO destruction timescale. However, CO could have easily survived over the system age (23?Myr) both in a scenario where C and CO dominate the gas mass (C/CO ratio of >0.16–0.43; Methods) and in a H 2 -dominated scenario (H 2 /CO ratio of around >10 5 ). Therefore, CO could either be the result of a very recent production and/or replenishment mechanism, or have survived for a large fraction of the system age. The CO detection constrains the gas to be located in the roughly 6–9? au region of the HD?172555 planetary system, co-located with the dust 12 . Accounting for the difference in luminosity between the Sun and HD?172555, this region corresponds to the same thermal conditions as at about 2.1–3.3? au —the asteroid belt region—in our Solar System. This makes the presence of CO, a highly volatile gas with sublimation temperatures as low as 20?K, extremely surprising in a system with age 23?Myr. Some physical mechanism must explain the presence of gas and dust in the outer terrestrial planet-forming region of HD?172555. We test four hypotheses for the origin of the debris, and examine them in light of existing evidence from both dust observations and from the ALMA detection of CO in the system. These scenarios are: (1) leftover gas and dust from a primordial, protoplanetary disk; (2) collisional production within an extrasolar asteroid belt; (3) inward transport of material from an external reservoir; and (4) release in the aftermath of a giant impact between planetary-sized bodies. Young A-type stars are born surrounded by protoplanetary disks of primordial gas and dust, but only 2–3% survive beyond the first 3?Myr of a star’s lifetime 16 , 17 . Even if the CO observed around HD?172555 were primordial, with its lifetime extended through shielding, the system would remain a remarkable outlier not only in age (at 23?Myr?old) but also in dust mass (orders of magnitude lower than protoplanetary disks 18 ) and in radial distribution of CO and dust (constrained within 10? au , and with a CO cavity, in contrast to protoplanetary disks typically extending out to tens/hundreds of astronomical units 19 , 20 ). The extreme depletion in dust mass would require efficient dust removal, either through accretion onto the star or grain growth 18 . At the same time, the confined radial extent would require interior/exterior truncation, for example, by as-yet-undiscovered companions. Finally, the peculiar dust mineralogy of HD?172555 (requiring energetic processing) is seldom seen in young protoplanetary disks, although it may have arisen in nebular shocks analogous to those that could have led to chondrule formation in the Solar System 21 . In conclusion, a primordial scenario would make HD?172555 an extreme outlier among other protoplanetary disks, favouring instead second-generation production. In the second-generation case, a steady-state collisional cascade within an asteroid belt can explain the dust mass detected in the system 22 , but not its abnormally steep particle size distribution 3 , or its mineralogy requiring high-energy collisions at velocities higher than expected within typical belts 2 , 3 , 13 . Although CO gas is commonly observed within collisional cascades in colder extrasolar Kuiper belt analogues at tens of astronomical units 23 , there its presence can be explained by the release of CO gas initially trapped within icy bodies, or by desorption of CO 2 ice followed by rapid photodissociation of CO 2 gas 24 , 25 , 26 , 27 , 28 . This picture is not plausible at 7.5? au ; studies of ice-bearing asteroids in the Solar System show that although retaining water ice in the deep subsurface is possible—explaining the presence of outgassing main-belt comets in the asteroid belt—CO and CO 2 cannot remain trapped at these temperatures 29 , 30 . In addition, any bodies forming at this location would be too warm to have formed with a substantial reservoir of CO or CO 2 . Alternatively, the observed dust and CO might originate from inward scattering of small bodies (exocomets) from an outer, cold reservoir akin to the Solar System’s Edgeworth–Kuiper belt 31 . We first explore the possibility of gas and dust production through sublimation of Solar-System-like, 10-km-sized exocomets. We find that this scenario can reproduce the observations only if exocomets can be delivered on low-eccentricity orbits, finding that the exocomet replenishment rate could potentially be reconciled with the non-detection of an outer belt (Methods). However, this scenario cannot explain the relatively narrow radial distribution of the CO ring, or the mineralogy of the observed dust, and is thus not likely. However, it is possible that a single, cold, massive icy body is scattered inwards from an undetected outer belt. To produce the observed, largely axisymmetric distribution of material and dust mineralogy, the icy body would need to undergo a giant impact and release its CO or CO 2 content. Assuming a 25% CO?+?CO 2 ice mass fraction, and complete ice release at impact, we find that a dwarf planet (at least half the size of Pluto) would be needed to produce the (lower limit on the) CO gas mass observed. Finally, CO gas and dust could be produced by a giant impact between planetary bodies formed in situ at around 7.5? au . The epoch of terrestrial planet formation, lasting from about 10?Myr to 100?Myr, is expected to be dominated by giant impacts. Within the Solar System, there is abundant evidence for the occurrence of giant impacts; the iron enrichment of Mercury 32 , 33 , the formation of the Moon 34 , 35 , 36 , the Martian hemispheric dichotomy 37 , 38 and the retrograde rotation of Venus 39 are all hypothesized to have their origins in giant impacts. The HD?172555 system, at 23?Myr, is at the expected age that terrestrial planet formation proceeds through giant impacts; the dust and gas observed in the system are located in a region that is analogous to the terrestrial zone in our own Solar System. Studies of post-impact dynamics allow us to set constraints on progenitor masses and time since impact by considering the current spatial distribution and mass of dust (Methods). The dust in the system is axisymmetric within observational uncertainties, implying that the time since impact is at least the debris symmetrization timescale, which at 7.5? au is of the order of about 0.2?Myr (ref. 40 ). The width of the dust debris, as resolved at shorter wavelengths 11 , 12 , suggests a progenitor mass of the order of about 8? M ? , although the exact value is dependent on the radial width of the dust debris (Methods). Impacts between such bodies produce debris that would survive encounters with leftover planets, on a timescale longer than symmetrization; this indicates that impacts of such planets could be responsible for the observed, long-lived debris field. Further supporting this scenario, we find that the optically thin CO gas mass detected in the system (0.45?×?10 ?5 M ? –1.21?×?10 ?5 M ? , or about ten times the mass of Earth’s atmosphere) is consistent with post-impact release from a planetary atmosphere. We note that any CO 2 in a planetary atmosphere will rapidly be converted into observable CO if liberated from the atmosphere, as CO 2 cannot be substantially shielded by C in the same way as CO because its photodissociation bands extend further towards the optical 41 . Simulations show that up to 60% of a modestly sized, heavy atmosphere can be stripped in the initial shock of an impact 42 , 43 . We find that the observed CO mass as well as the C needed to shield the CO for at least the symmetrization timescale require the release of an amount of CO 2 corresponding to just 9–23% of the total present in the Venusian atmosphere 44 . More (less) massive planets with similar heavy atmospheres would require a smaller (larger) fraction of the atmosphere removed, or a smaller (larger) abundance of CO and/or CO 2 . For lighter, H 2 -dominated atmospheres, longer-term thermal effects can result in the stripping of the entire atmospheric envelope 45 , in which case, once again, an amount of CO and/or CO 2 consistent with the observations could plausibly be liberated. The detection and morphology of CO gas, combined with previous evidence from dust imaging and spectroscopy, supports a picture where a giant impact took place at least 0.2?Myr?ago in the outer terrestrial planet-forming region of the 23-Myr-old HD?172555 system. Planetary-scale impacts are predicted to be commonplace in the latest stages of planet formation; the discovery of CO gas in the terrestrial planet-forming region, in amounts consistent with the expectation from atmospheric stripping, suggests that giant impacts may release not only copious, observable dust but also detectable amounts of gas. Furthermore, this discovery reveals the importance of gas release in post-impact dynamics, and highlights the potential of using gas as a tool to search for giant impacts in nearby planetary systems, while providing a unique window into the composition of young planets and their atmospheres. Methods . ALMA observations . We analysed archival data from the ALMA telescope taken during cycle 1 in band 6 (project code 2012.1.00437.S). The observations were performed with the 12-m array in a compact antenna configuration. The on-source time was 76?min. The spectral setup included four spectral windows, of which three were set up in time-division mode for continuum observations and centred at 213?GHz, 215?GHz and 228?GHz. One window was set in frequency-division mode (with a high spectral resolution of 488.29?kHz) to target the 12 CO J ?=?2–1?line (rest frequency 230.538?GHz). We calibrated the visibility data using scripts provided by the ALMA observatory. The CASA?(Common Astronomy Software Applications) software (version 5.6.1, https://casa.nrao.edu/casadocs/casa-5.6.0 ) was used for visibility imaging. For both the continuum and the CO line emission, we removed data from antennas 7, 19 and 25. These data were taken in a hybrid configuration, with the three flagged antennas far from the compact group of antennas; removing these data substantially improves the imaging. The CLEAN algorithm 46 was used to image both the continuum and CO line emission through the CASA ‘tclean’ task. To achieve maximum sensitivity, we used natural weighting for both datasets, resulting in a synthesized beam size of 1.16″?×?0.75″, corresponding to 32.9? au ?×?21.3? au at the system distance (28.5?pc) and a position angle of 81°. Before line imaging, we subtracted the continuum (measured in line-free regions of the spectrum) in visibility space using the CASA ‘uvcontsub’ task. We then imaged the CO to produce a cube with frequencies within ±25?MHz of the rest frequency, corresponding to velocities within ±30?km?s ?1 of the radial velocity of the star. The data were imaged at the native spectral resolution (twice the original channel width) of 488.29?kHz. To obtain the CO moment-0 map, shown in Fig. 1 , we integrated along the velocity axis between ±15?km?s ?1 of the stellar velocity. The emission is spatially unresolved in both the continuum and CO images, which have root-mean-square noise levels of 0.029?mJy per beam and 26?mJy?km?s ?1 per beam, respectively. This yields a peak detection at a signal-to-noise ratio per beam of 4 and 9 for the continuum and CO J ?=?2–1 spectrally integrated emission, and total fluxes of 0.12?±?0.03?mJy and 170?±?30?mJy?km?s ?1 , respectively. Note that the flux calibration is expected to be accurate within 10%; this uncertainty was added in quadrature to obtain the quoted errors ( https://almascience.nrao.edu/documents-and-tools/cycle-1/alma-ot-reference-manual ). Both the CO gas and continuum dust emission centroids are consistent with the Gaia proper-motion-corrected position of the system 7 , 8 . The literature 11 values for the dust location in astronomical units have been corrected to account for the new Gaia distance to the star compared with the previous Hipparcos distance. To extract a one-dimensional spectrum from the data cube, we integrate within a circular mask of radius 2.5 σ beam (with σ beam being the standard deviation of the two-dimensional Gaussian beam, averaged between the major axis and the minor axis). Millimetre dust mass . To calculate the dust mass, we assume that the dust grains act as blackbodies and emit according to their Planck function at a temperature of 169?K. Assuming the dust is optically thin, the total dust mass can be calculated from the observed flux density,? F ν ?=?0.12?±?0.03?mJy, when accounting for the stellar contribution. This contribution is estimated to be 0.035?mJy at 230?GHz from our best fit PHOENIX stellar model fitted to optical and near-IR photometry (see ‘Optically thin CO mass calculation’ below). PHOENIX models have been shown to be a good fit to millimetre-wavelength photometry of Sirius A 47 , which has a similar spectral type to HD?172555. Subtracting this stellar contribution to the detected millimetre emission yields a contribution due to dust of 85?±?30??Jy. The dust grain opacity is assumed to be 10?cm 2 ?g ?1 at 1,000?GHz and scaled to the frequency of the observation with an opacity power-law index of β ?=?1 (ref. 48 ). These assumptions yield a dust mass of (1.8?±?0.6)?×?10 ?4 M ? . Optically thin CO ring modelling . To model the velocity spectrum expected from a circular orbiting ring or disk of material, we calculate Keplerian velocities assuming a stellar mass of 1.76? M ☉ (ref. 6 ). Two-dimensional orbital velocity vectors are calculated for a radial and azimuthal grid, assuming a vertically thin ring/disk of gas that has radially and azimuthally uniform surface density between an inner boundary and an outer boundary. They are transformed to the sky plane using the ring/disk inclination to obtain radial velocities along the line of sight. These velocities in the reference frame of the star are then added to the radial velocity of the star in the barycentric frame (left as a free parameter) to obtain barycentric velocities as observed by ALMA. A histogram of these velocities, with the same binning as the observed data, serves as a model spectrum. We normalize the unitless spectrum such that the integral of the spectrum is equal to the integrated flux of the line, a free parameter in the fit. This spectrum is then convolved with a Gaussian of full-width at half-maximum equal to twice the channel width to reproduce the spectral response of the instrument due to Hanning smoothing ( https://safe.nrao.edu/wiki/pub/Main/ALMAWindowFunctions/Note_on_Spectral_Response.pdf ). In addition to these parameters, the model fits the inclination, radial location of the midpoint and width of the ring/disk. A Markov chain Monte Carlo (MCMC) approach was used to determine the best fit to the data. We used the Python package emcee 49 . The uncertainty in each velocity bin was assumed to be equal to the root mean square measured in the region of the spectrum outside the detected emission. Flat priors were applied for the radial location, width, integrated line flux and stellar velocity. A Gaussian prior was applied to the inclination, assuming the gas shares the same inclination as the dust disk, as determined from previous resolved imaging 12 . We carried out an additional run where a flat prior was applied to the inclination, to confirm the assumption of shared inclination. Extended Data Fig. 1 shows the posterior probability distributions of the parameters obtained from our MCMC runs and Extended Data Table 1 indicates the best-fit values, obtained as the 50th?±?34th percentiles of the posterior distributions of each parameter, marginalized over all other parameters. Best-fit values from both runs assuming a Gaussian or flat prior on the inclination are included. Optically thin CO mass calculation . To derive a CO mass from the best-fit spectrally integrated line flux, we begin by assuming that the line is optically thin and considering the excitation conditions the gas may be subject to, which affect this conversion. We follow an existing framework 24 , which considers that the energy levels of a CO molecule may be populated by collisions with other species (or by one dominant species, the main collisional partner), or by absorption and emission of radiation, giving rise to two limiting regimes, a radiation-dominated regime (at low gas densities) and a collision-dominated regime (local thermodynamic equilibrium (LTE) at high gas densities). The choice or density of collisional partners (in our case, electrons) does not affect the level populations in these two limiting regimes, and therefore the range of CO masses derived. To account for the full range of excitation conditions, we therefore use a non-LTE code 24 to solve the statistical equilibrium equations and calculate the level populations. This includes the effect of fluorescence induced by stellar UV and IR radiation as seen by a CO molecule at 7.5? au from the central star 50 . For the star, we adopt a PHOENIX 51 model spectrum fitted to multiband optical and near-IR photometry using the MultiNest code 52 , as described in previous studies 53 , 54 . This yields a best-fit stellar effective temperature of T eff ?=?7,840?±30?K and luminosity of \({L}_{\star }\) ?=?7.7?±0.1? L ☉ . The non-LTE calculation yields a formal range of possible CO masses between 0.45?×?10 ?5 M ? and 1.25?×?10 ?5 M ? for kinetic temperatures between 100?K and 250?K, encompassing the blackbody temperature of 169?K at 7.5? au . Three-dimensional radiative transfer modelling . We use the RADMC-3D ( http://www.ita.uni-heidelberg.de/~dullemond/software/radmc-3d/ ) radiative transfer code to check the impact of optical depth in more detail. We use the same ring geometry as obtained from optically thin fitting (Extended Data Table 1 , Gaussian prior). We vary the input CO mass and kinetic temperature over a two-dimensional grid (Fig. 2b ), and connect the latter to the vertical aspect ratio by assuming a vertically isothermal gas disk with a mean molecular weight of 14 (that is, the gas mass is dominated by atomic C and O, as expected in a second-generation scenario). To sample the ring well spatially and spectrally, we create cubes of J ?=?2–1 emission with a pixel size of 6?mas (corresponding to a physical scale of 0.18? au ), and the same native channel width as our data. We then spectrally convolve with a Gaussian to reproduce the spectral response of the instrument and extract a one-dimensional model spectrum by spatially integrating the model emission. We then compared this model spectrum to the data (as shown in Fig. 2a ) and calculated χ 2 for every mass and temperature in our grid, to obtain a χ 2 map (Fig. 2b ). CO survival lifetime against photodissociation . The observed CO gas in the circumstellar environment around HD?172555 will be subject to photodissociation from the stellar and interstellar UV field. The photodissociation rate in s ?1 of a molecule in a radiation field I ( λ ) is \(k=\int \sigma (\lambda )I(\lambda ){\rm{d}}\lambda \) , where?λ is the wavelength, and? σ ( λ ) is the photodissociation cross-section in cm 2 . We use the CO photodissociation cross-sections from the Leiden database 41 ( https://home.strw.leidenuniv.nl/~ewine/photo/display_co_42983b05e2f2cc22822e30beb7bdd668.html ). We adopt the stellar spectrum from the optically thin mass section above, scaled to the centre of the gas ring (7.5? au ) to obtain the stellar radiation field. We find that the star dominates over the interstellar radiation field, and that the CO photodissociation timescale (1/ k ) at the ring’s radial location is approximately one day. The shielding effects are estimated using pre-computed shielding constants 41 ; we interpolate the constants for stars of 4,000?K and 10,000?K to a stellar temperature of 8,000?K, closest to the effective temperature of HD?172555. The CO column density is calculated from the centre of the ring along the line of sight to the star, using our best-fit uniform ring model parameters. To find the H 2 and C column densities that provide sufficient shielding, we interpolate the shielding constants along the column-density axis and find the column density required. Delivery from an outer belt . Replenishment requirement . We consider a scenario where the dust grains and CO gas are produced from sublimation-driven release by Solar-System-like comets entering the inner region of the HD?172555 system. This requires replenishment of the observed total dust mass (in grains up to centimetres in size) on timescales comparable to their removal (assuming steady state). We assume that their removal is dominated by collisions, setting up a cascade down to the smallest grains (of size approximately 3.5?μm for a grain density of 2,700?kg?m ?3 ) that are then removed by radiation pressure from the central star, to derive a mass-loss rate of 2.2?×?10 ?2 M ? ?Myr ?1 (using equation (21) in ref. 26 ). We neglect the effect of gas drag, assuming that the larger, centimetre grains are unaffected. We use the results of thermochemical modelling 55 calibrated on Solar System comets to estimate the mass-loss rate per unit surface area of an exocomet at 7.5? au around 7.7? L ☉ to be 5.6?×?10 ?6 ?kg?m ?2 ?s ?1 of dust in grains up to centimetres in size. Dividing by an assumed bulk density of 560?kg?m ?3 , and assuming a 1:1 dust/ice ratio, this corresponds to an erosion rate of 0.62?m?yr ?1 . Therefore, a comet with a 10-km radius will be emitting dust at a rate of 7,040?kg?s ?1 (3.72?×?10 ?8 M ? ?Myr ?1 ) and will survive (if continuously sublimating at this rate) for about 16?kyr. High-eccentricity exocomet population . In the first case, we assume exocomets arise from a yet-undetected belt at 100? au (a typical location of cold exocometary belts around A stars 56 ) and approach the inner regions on eccentric orbits. The velocity distribution from the observed CO emission implies line-of-sight velocities for the gas of about 14.5?km?s ?1 , corresponding to 7.5? au for circular orbits around HD?172555. However, these velocities will be achieved at larger radii for comets with non-zero eccentricities, random pericentre directions and orbiting in the same plane as the inner system of HD?172555. The pericentre distance corresponding to pericentre velocities of about 14.5?km?s ?1 will increase with the apocentre (and therefore the eccentricity) of the orbits; for an apocentre of 100? au , we derive a pericentre distance of 13.2? au . Therefore, we expect gas released from exocomets on these orbits to be located at and beyond about 13.2? au from the central star. For comparison, this is 80% larger than the radial location derived for circular orbits (7.5? au ). However, eccentric exocomets with the observed velocities would produce a CO ring with diameter about 26? au (0.9″), which is comparable to the resolution element of our observations, implying that the ring would have been marginally resolved, which is not the case. In addition, scattered light observations strongly constrain dust emission to less than 0.9″ (3 σ ) diameter, with a sharp outer edge. The fact that the expected radius of a CO ring would significantly exceed the observed value makes the eccentric exocomet scenario inconsistent with the available observations. Note that reasonably changing the apocentre location, and therefore the radius of the hypothetical cold belt where the exocomets originate, does not alter our conclusions significantly. For example, assuming an outer belt location of 20? au instead of 100? au implies exocomet pericentres at 10? au . This would produce CO emission with a minimum expected diameter of 20? au (0.7″) and extending much beyond it, which probably would have been spatially resolved in the ALMA data. Low-eccentricity exocomet population . In the second case, we assume exocomets are being continuously scattered inward from an outer belt by a chain of low-mass planets, undergoing multiple scatterings and producing a low-eccentricity comet population at about 7.5? au . We here assume circular orbits for simplicity. The dust observed requires replenishment at a rate of 2.2?×?10 ?2 M ? ?Myr ?1 . Assuming exocomets of 10?km in size with an exocometary dust release rate of 3.72?×10 ?8 M ? ?Myr ?1 , 5.9?×?10 5 exocomets are required to be sublimating at around about 7.5? au at any point in time. For a bulk density of 560?kg?m ?3 , this corresponds to 2.3?×?10 ?4 M ? in 10-km exocomets. While sublimating at this rate, such an exocomet would survive for about 16?kyr, so comets would need to be resupplied by inward scattering to the inner planetary system at a rate of 1.4?×?10 ?8 M ? ?yr ?1 . Inward scattering is an inefficient process; simulations maximizing inward scattering by chains of low-mass planets indicate that only a few per cent of comets encountering an outermost planet make it into the inner regions, as the vast majority are ejected 57 . Therefore, higher supply rates of the order of about 10 ?6 M ? ?yr ?1 from a putative outer belt are probably needed, which would imply that a currently undetected outer belt would have resupplied 23? M ? in 10-km exocomets into the inner regions over the 23?Myr age of the system. We can compare this to the upper limit on the presence of an outer belt at 100? au from our ALMA data. Assuming blackbody temperatures, and the same dust opacity as used for dust in the inner regions, we derive an upper limit of less than 1.8? × ?10 ?3 M ? (3 σ ) on the mass of solids of sizes up to centimetre sizes. Extrapolating from centimetre-sized grains up to 10-km exocomets (assuming a size distribution with a constant power-law slope of ?3.5), we obtain an upper limit on the total mass in 10-km-sized exocomets of less than 1.8? M ? (3 σ ). Therefore, the current mass of the outer belt would be at least a factor of about ten smaller than the mass that has been removed from the belt over its lifetime, which is by itself not impossible. However, it is likely that exocomets reaching their inner regions would retain some eccentricity, which would prolong their survival (lower sublimation rates) and also potentially make the gas velocity distribution inconsistent with the observations. In addition to that, the narrow CO radial distribution is in tension with production rates of CO as a function of heliocentric distance in sublimating Solar System comets 58 , which show no clear enhancement in production rates at radii whose thermal conditions are comparable to that of HD?172555. Although inward scattered, sublimating exocomets could produce a steeper-than-collisional size distribution of grains, down to sizes smaller than the radiation pressure blow-out limit, they would not be able to reproduce the peculiar dust mineralogy indicative of the energetic processing typical of hypervelocity impacts, unless exocometary impacts in the inner regions contribute, and/or are the source of the dust and gas observed. In conclusion, sublimating exocomets could be replenished in the inner system at high-enough rates to explain the observed dust emission, but are unlikely to explain the relatively narrow distribution of CO, and the dust mineralogy observed around HD?172555. In situ giant impact constraints from axisymmetry and width of debris . Observations of both CO and dust 11 , 12 are consistent with the CO and dust distribution being axisymmetric. Because of this, the time since impact must be at least the symmetrization timescale, which is on the order of a few tens of thousands of orbits 40 (about 0.2?Myr at 7.5? au ). Constraints on the planet mass can be derived from the width of the debris, as a proxy for the velocity dispersion of released material. The width, d r , is given by d r ?=?2 re p , where e p is the proper eccentricity of the orbiting debris 59 . This eccentricity is related to the velocity dispersion, \({\sigma }_{v}\) , through \({\sigma }_{v}\approx \sqrt{1.5}{e}_{{\rm{p}}}{v}_{{\rm{k}}}\) , where \({v}_{{\rm{k}}}\) is the Keplerian velocity at the given semimajor axis. We assume the velocity dispersion is related to the escape velocity? v esc ?at the surface, ?of the colliding bodies 40 by \({\sigma }_{v}\approx 0.46{v}_{{\rm{esc}}}\approx 0.46\sqrt{\frac{2G{M}_{{\rm{p}}}}{{R}_{{\rm{p}}}}}\) , where G is the universal gravitational constant, and R p is the?planetary radius. Thus, for an observed debris width d r , a planet of mass \({M}_{{\rm{p}}}\approx 103\frac{{M}_{\ast }^{3/2}}{{\rho }_{p}^{1/2}}{\left(\frac{{\rm{d}}r}{{r}^{3/2}}\right)}^{3}\) in M ? is expected, where \({\rho }_{{\rm{p}}}\) is the planet’s bulk density in g?cm ?3 , \({M}_{\ast }\) is the stellar mass in M ? , and both r and d r are in astronomical units. If the planetary bodies colliding have rocky, Earth-like compositions, they will have bulk densities of about 5.5?g?cm ?3 . We assume that the solid debris is confined to the same radial width as the dust width derived from mid-IR Q-band imaging 11 , which finds d r dust / r ?≈?1.2, where d r dust is the width of the dust ring. These assumptions yield a planetary mass on the order of about 8? M ? . We note that mid-IR imaging is sensitive to small grains whose width might have been broadened due to radiation pressure from the central star; if radiation pressure inflated the width compared with that expected from the velocity dispersion, the planetary mass involved would be reduced. If instead the bulk planetary density is lower, the mass of planet involved in the collision would be larger. Constraints from debris survival . Progenitor mass constraints must also be considered in the context of debris survival; the observed debris must survive encounters with a leftover planet in the time since impact. We conservatively assume that the mass of the largest product of the impact is on the order of the mass of the impact progenitors, and consider the outcome of encounters between debris and a surviving planet 60 . Given an assumed planetary density similar to that of the Earth, and a semimajor axis of 7.5? au , we first note that for a leftover planet with mass below about 1.8? M ? , the predominant (eventual) outcome of an encounter will be accretion, whereas for masses above that threshold, the predominant (eventual) outcome will be ejection. However, timescales of accretion and ejection have a role in debris survival. These timescales must be at least as long as the symmetrization timescale (0.2?Myr) to ensure that debris produced in an impact could be seen in the symmetrical structure observed today. For progenitor masses in the range derived from the debris width constraints above, the reaccretion and ejection timescales are much longer than the symmetrization timescale. Therefore, giant impact debris at 7.5? au is expected to survive encounters with leftover planets with masses on the order of 8? M ? , as large as those of the progenitors, on a timescale longer than the symmetrization timescale. Data availability . The ALMA programme number for the presented data is 2012.1.00437.S and data can be found in the online ALMA archive. The cleaned .fits files are available upon request from the corresponding author. Code availability . RADMC-3D is available at https://github.com/dullemond/radmc3d-2.0 and emcee is available at https://emcee.readthedocs.io/en/stable/ . Custom code, including the ring model and non-LTE code, is available at https://github.com/tmschneiderman/hd172555_CO_2021 . References . 1. Morbidelli, A., Lunine, J. I., O’Brien, D. P., Raymond, S. N. & Walsh, K. J. Building terrestrial planets. Annu. Rev. Earth Planet. Sci. 40 , 251–275 (2012). ADS ? CAS ? Article ? Google Scholar ? 2. Lisse, C. M. et al. Abundant circumstellar silica dust and SiO gas created by a giant hypervelocity collision in the 12 Myr HD 172555 system. Astrophys. J. 701 , 2019–2032 (2009). ADS ? CAS ? Article ? Google Scholar ? 3. Johnson, B. C. et al. A self-consistent model of the circumstellar debris created by a giant hypervelocity impact in the HD 172555 system. Astrophys. J . 761 , 45 (2012). 4. Wyatt, M. C. et al. Steady state evolution of debris disks around A stars. Astrophys. J. 663 , 365–382 (2007). ADS ? CAS ? Article ? Google Scholar ? 5. Mamajek, E. E. & Bell, C. P. M. On the age of the β Pictoris moving group. Mon. Not. R. Astron. Soc. 445 , 2169–2180 (2014). ADS ? Article ? Google Scholar ? 6. David, T. J. & Hillenbrand, L. A. The ages of early-type stars: Str?mgren photometric methods calibrated, validated, tested, and applied to hosts and prospective hosts of directly imaged exoplanets. Astrophys. J. 804 , 146 (2015). ADS ? Article ? Google Scholar ? 7. Prusti, T. et al. The Gaia mission. Astron. Astrophys. 595 , A1 (2016). Article ? Google Scholar ? 8. Gaia Collaboration et al. Gaia Data Release 2: summary of the contents and survey properties. Astron. Astrophys. 616 , A1 (2018). 9. Zuckerman, B., Song, I., Bessell, M. S. & Webb, R. A. The β Pictoris moving group. Astrophys. J. 562 , L87–L90 (2001). ADS ? Article ? Google Scholar ? 10. Cote, J. B and A type stars with unexpectedly large colour excesses at IRAS wavelengths. Astron. Astrophys. 181 , 77–84 (1987). ADS ? CAS ? Google Scholar ? 11. Smith, R., Wyatt, M. C. & Haniff, C. A. Resolving the terrestrial planet forming regions of HD113766 and HD172555 with MIDI. Mon. Not. R. Astron. Soc. 422 , 2560–2580 (2012). ADS ? Article ? Google Scholar ? 12. Engler, N., Schmid, H. M., Quanz, S. P., Avenhaus, H. & Bazzon, A. Detection of scattered light from the hot dust in HD 172555. Astron. Astrophys. 618 , 1–15 (2018). 13. Chen, C. H. et al. Spitzer IRS spectroscopy of IRAS‐discovered debris disks. Astrophys. J. Suppl. Ser. 166 , 351–377 (2006). ADS ? CAS ? Article ? Google Scholar ? 14. Gontcharov, G. A. Pulkovo compilation of radial velocities for 35 495 Hipparcos stars in a common system. Astron. Lett. 32 , 759–771 (2006). ADS ? Article ? Google Scholar ? 15. Kral, Q., Marino, S., Wyatt, M. C., Kama, M. & Matrà, L. Imaging [C I ] around HD 131835: reinterpreting young debris discs with protoplanetary disc levels of CO gas as shielded secondary discs. Mon. Not. R. Astron. Soc. 489 , 3670–3691 (2019). ADS ? Google Scholar ? 16. Kennedy, G. M. & Kenyon, S. J. Stellar mass dependent disk dispersal. Astrophys. J. 695 , 1210–1226 (2009). ADS ? CAS ? Article ? Google Scholar ? 17. Ribas, ?., Bouy, H. & Merín, B. Protoplanetary disk lifetimes vs. stellar mass and possible implications for giant planet populations. Astron. Astrophys. 576 , A52 (2015). ADS ? Article ? Google Scholar ? 18. Wyatt, M. C., Pani?, O., Kennedy, G. M. & Matrà, L. Five steps in the evolution from protoplanetary to debris disk. Astrophys. Space Sci. 357 , 103 (2015). ADS ? Article ? Google Scholar ? 19. Isella, A. et al. The Disk Substructures at High Angular Resolution Project (DSHARP) - IX: a high definition study of the HD 163296 planet forming disk. Astrophys. J. Lett. 869 , L49 (2018). 20. Boehler, Y. et al. The complex morphology of the young disk MWC 758: spirals and dust clumps around a large cavity. Astrophys. J. 853 , 162 (2017). 21. Morris, M. A. & Desch, S. J. Thermal histories of chondrules in solar nebula shocks. Astrophys. J. 722 , 1474–1494 (2010). ADS ? CAS ? Article ? Google Scholar ? 22. Su, K. Y. L. et al. Mid-infrared studies of HD 113766 and HD 172555: assessing variability in the terrestrial zone of young exoplanetary systems. Astrophys. J. 898 , 21 (2020). 23. Matrà, L., ?berg, K. I., Wilner, D. J., Olofsson, J. & Bayo, A. On the ubiquity and stellar luminosity dependence of exocometary CO gas: detection around M dwarf TWA 7. Astron. J. 157 , 117 (2019). 24. Matrà, L., Pani?, O., Wyatt, M. C. & Dent, W. R. F. CO mass upper limits in the Fomalhaut ring—the importance of NLTE excitation in debris discs and future prospects with ALMA. Mon. Not. R. Astron. Soc. 447 , 3936–3947 (2015). ADS ? Article ? CAS ? Google Scholar ? 25. Marino, S. et al. Exocometary gas in the HD 181327 debris ring. Mon. Not. R. Astron. Soc. 460 , 2933–2944 (2016). ADS ? CAS ? Article ? Google Scholar ? 26. Matrà, L. et al. Detection of exocometary CO within the 440 Myr-old Fomalhaut belt: a similar CO+CO 2 ice abundance in exocomets and Solar System comets. Astrophys. J. 842 , 9 (2017). 27. Kral, Q., Matrà, L., Wyatt, M. C. & Kennedy, G. M. Predictions for the secondary CO, C and O gas content of debris discs from the destruction of volatile-rich planetesimals. Mon. Not. R. Astron. Soc. 469 , 521–550 (2017). ADS ? CAS ? Article ? Google Scholar ? 28. Zuckerman, B. & Song, I. A 40Myr old gaseous circumstellar disk at 49 CETI: massive CO-rich comet clouds at young A-type stars. Astrophys. J. 758 , 77 (2012). ADS ? Article ? CAS ? Google Scholar ? 29. Prialnik, D. & Rosenberg, E. D. Can ice survive in main-belt comets? Long-term evolution models of comet 133P/Elst-Pizarro. Mon. Not. R. Astron. Soc. Lett. 399 , L79–L83 (2009). ADS ? Article ? Google Scholar ? 30. Snodgrass, C. et al. The Main Belt comets and ice in the Solar System. Astron. Astrophys . Rev. 25 , 5 (2017). Google Scholar ? 31. Marino, S. et al. ALMA observations of the η Corvi debris disc: inward scattering of CO-rich exocomets by a chain of 3-30 M ? planets? Mon. Not. R. Astron. Soc. 465 , 2595–2615 (2017). ADS ? CAS ? Article ? Google Scholar ? 32. Benz, W., Anic, A., Horner, J. & Whitby, J. A. The origin of Mercury. Space Sci. Rev. 132 , 189–202 (2007). ADS ? CAS ? Article ? Google Scholar ? 33. Chau, A., Reinhardt, C., Helled, R. & Stadel, J. Forming Mercury by giant impacts. Astrophys. J. 865 , 35 (2018). ADS ? Article ? CAS ? Google Scholar ? 34. Canup, R. M. Simulations of a late lunar-forming impact. Icarus 168 , 433–456 (2004). ADS ? CAS ? Article ? Google Scholar ? 35. Canup, R. M. Forming a moon with an Earth-like composition via a giant impact. Science 338 , 1052-1055 (2012). ADS ? CAS ? PubMed ? PubMed Central ? Article ? Google Scholar ? 36. Mastrobuono-Battisti, A., Perets, H. B. & Raymond, S. N. A primordial origin for the compositional similarity between the Earth and the Moon. Nature 520 , 212–215 (2015). ADS ? CAS ? PubMed ? Article ? Google Scholar ? 37. Marinova, M. M., Aharonson, O. & Asphaug, E. Mega-impact formation of the Mars hemispheric dichotomy. Nature 453 , 1216–1219 (2008). ADS ? CAS ? PubMed ? Article ? Google Scholar ? 38. Nimmo, F., Hart, S. D., Korycansky, D. G. & Agnor, C. B. Implications of an impact origin for the Martian hemispheric dichotomy. Nature 453 , 1220–1223 (2008). ADS ? CAS ? PubMed ? Article ? Google Scholar ? 39. Davies, J. H. Did a mega-collision dry Venus’ interior? Earth Planet. Sci. Lett. 268 , 376–383 (2008). ADS ? CAS ? Article ? Google Scholar ? 40. Wyatt, M. C. & Jackson, A. P. Insights into planet formation from debris disks: II. giant impacts in extrasolar planetary systems. Space Sci. Rev. 205 , 231–265 (2016). ADS ? Article ? Google Scholar ? 41. Heays, A. N., Bosman, A. D. & van Dishoeck, E. F. Photodissociation and photoionisation of atoms and molecules of astrophysical interest. Astron. Astrophys. 602 , A105 (2017). 42. Schlichting, H. E. & Mukhopadhyay, S. Atmosphere impact losses. Space Sci. Rev. 214 , 34 (2018). ADS ? Article ? Google Scholar ? 43. Kegerreis, J. A. et al. Atmospheric erosion by giant impacts onto terrestrial planets: a scaling law for any speed, angle, mass, and density. Astrophys. J. Lett. 901 , L31 (2020). ADS ? CAS ? Article ? Google Scholar ? 44. Lammer, H. et al. Origin and evolution of the atmospheres of early Venus, Earth and Mars. Astron. Astrophys. Rev. 26 , 2 (2018). ADS ? Article ? Google Scholar ? 45. Biersteker, J. B. & Schlichting, H. E. Atmospheric mass-loss due to giant impacts: the importance of the thermal component for hydrogen–helium envelopes. Mon. Not. R. Astron. Soc. 485 , 4454–4463 (2019). ADS ? CAS ? PubMed ? PubMed Central ? Article ? Google Scholar ? 46. H?gbom, J. & Cornwell, T. Aperture synthesis with a non-regular distribution of interferometer baselines. Commentary. Astron. Astrophys. 500 , 55–66 (2009). ADS ? Google Scholar ? 47. White, J. A. et al. The MESAS Project: long-wavelength follow-up observations of Sirius A. Astrophys. J. 875 , 55 (2019). ADS ? CAS ? Article ? Google Scholar ? 48. Beckwith, S. V. W., Sargent, A. I., Chini, R. S. & Guesten, R. A survey for circumstellar disks around young stellar objects. Astron. J. 99 , 924 (1990). ADS ? Article ? Google Scholar ? 49. Foreman-Mackey, D., Hogg, D. W., Lang, D. & Goodman, J. emcee: the MCMC hammer. Publ. Astron. Soc. Pac. 125 , 306–312 (2013). ADS ? Article ? Google Scholar ? 50. Matrà, L. et al. Molecular reconnaissance of the β Pictoris gas disk with the SMA: a low HCN/(CO+CO 2 ) outgassing ratio and predictions for future surveys. Astrophys. J. 853 , 147 (2018). 51. Husser, T. O. et al. A new extensive library of PHOENIX stellar atmospheres and synthetic spectra. Astron. Astrophys. 553 , A6 (2013). Article ? CAS ? Google Scholar ? 52. Feroz, F., Hobson, M. P. & Bridges, M. MultiNest: an efficient and robust Bayesian inference tool for cosmology and particle physics. Mon. Not. R. Astron. Soc. 398 , 1601–1614 (2009). ADS ? Article ? Google Scholar ? 53. Sepulveda, A. G. et al. The REASONS Survey: resolved millimeter observations of a large debris disk around the nearby F star HD 170773. Astrophys. J. 881 , 84 (2019). 54. Yelverton, B., Kennedy, G. M., Su, K. Y. L. & Wyatt, M. C. A statistically significant lack of debris discs in medium separation binary systems. Mon. Not. R. Astron. Soc. 488 , 3588–3606 (2019). ADS ? Article ? Google Scholar ? 55. Marboeuf, U., Bonsor, A. & Augereau, J. C. Extrasolar comets: the origin of dust in exozodiacal disks? Planet. Space Sci. 133 , 47–62 (2016). ADS ? Article ? Google Scholar ? 56. Matrà, L. et al. An empirical planetesimal belt radius–stellar luminosity relation. Astrophys. J. 859 , 72 (2018). 57. Marino, S., Bonsor, A., Wyatt, M. C. & Kral, Q. Scattering of exocomets by a planet chain: exozodi levels and the delivery of cometary material to inner planets. Mon. Not. R. Astron. Soc. 479 , 1651–1671 (2018). ADS ? CAS ? Article ? Google Scholar ? 58. Biver, N. et al. The 1995–2002 long-term monitoring of comet C/1995 O1 (Hale-Bopp) at radio wavelength. Earth Moon Planets 90 , 5–14 (2002). ADS ? CAS ? Article ? Google Scholar ? 59. Wyatt, M. C. et al. How observations of circumstellar disk asymmetries can reveal hidden planets: pericenter glow and its application to the HR 4796 disk. Astrophys. J. 527 , 918–944 (1999). ADS ? Article ? Google Scholar ? 60. Wyatt, M. C., Bonsor, A., Jackson, A. P., Marino, S. & Shannon, A. How to design a planetary system for different scattering outcomes: giant impact sweet spot, maximizing exocomets, scattered discs. Mon. Not. R. Astron. Soc. 464 , 3385–3407 (2017). ADS ? Article ? Google Scholar ? Download references Acknowledgements . We are grateful to J. Biersteker for discussions on the liberation of atmospheres in the aftermath of giant impacts. This paper makes use of ALMA data ADS/JAO.ALMA\#2012.1.00437.S. ALMA is a partnership of ESO (representing its member states), NSF (USA) and NINS (Japan), together with NRC (Canada), NSC and ASIAA (Taiwan) and KASI (Republic of Korea), in cooperation with the Republic of Chile. The Joint ALMA Observatory is operated by ESO, AUI/NRAO and NAOJ. K.I.?. acknowledges support from the Simons Foundation (SCOL \#321183). G.M.K. is supported by the Royal Society as a Royal Society University Research Fellow. Author information . Affiliations . Department of Earth, Atmospheric and Planetary Sciences, Massachusetts Institute of Technology, Cambridge, MA, USA Tajana Schneiderman Centre for Astronomy, School of Physics, National University of Ireland Galway, Galway, Ireland Luca Matrà Centre for Planetary Sciences, University of Toronto at Scarborough, Toronto, Ontario, Canada Alan P. Jackson School of Earth and Space Exploration, Arizona State University, Tempe, AZ, USA Alan P. Jackson Department of Physics, University of Warwick, Coventry, UK Grant M. Kennedy Centre for Exoplanets and Habitability, University of Warwick, Coventry, UK Grant M. Kennedy LESIA, Observatoire de Paris, Université PSL, CNRS, Sorbonne Université, Université Paris Diderot, Sorbonne Paris Cité, Meudon, France Quentin Kral Institute of Astronomy, University of Cambridge, Cambridge, UK Sebastián Marino?&?Mark C. Wyatt Jesus College, University of Cambridge, Cambridge, UK Sebastián Marino Center for Astrophysics Harvard & Smithsonian, Cambridge, MA, USA Karin I. ?berg?&?David J. Wilner Steward Observatory, University of Arizona, Tucson, AZ, USA Kate Y. L. Su Authors Tajana Schneiderman View author publications You can also search for this author in PubMed ? Google Scholar Luca Matrà View author publications You can also search for this author in PubMed ? Google Scholar Alan P. Jackson View author publications You can also search for this author in PubMed ? Google Scholar Grant M. Kennedy View author publications You can also search for this author in PubMed ? Google Scholar Quentin Kral View author publications You can also search for this author in PubMed ? Google Scholar Sebastián Marino View author publications You can also search for this author in PubMed ? Google Scholar Karin I. ?berg View author publications You can also search for this author in PubMed ? Google Scholar Kate Y. L. Su View author publications You can also search for this author in PubMed ? Google Scholar David J. Wilner View author publications You can also search for this author in PubMed ? Google Scholar Mark C. Wyatt View author publications You can also search for this author in PubMed ? Google Scholar Contributions . T.S. led the optically thin modelling and discussion. L.M. led the radiative transfer modelling. T.S. and L.M. were involved in data reduction, processing and writing of the manuscript. All authors contributed to discussions of the results and commented on the manuscript. Corresponding author . Correspondence to Tajana Schneiderman . Ethics declarations . Competing interests . The authors declare no competing interests. Additional information . Peer review information Nature thanks Inseok Song and the other, anonymous, reviewer(s) for their contribution to the peer review of this work. Peer reviewer reports are available. Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Extended data figures and tables . Extended Data Fig. 1 Posterior probability distributions for the model parameters obtained from the emcee fitting process. . All parameters are well constrained, with best-fit values listed in Extended Data Table 1 . This model was fitted to spectral data retaining original channel widths, assumed a Gaussian prior on the inclination, and assumed a stellar mass of 1.76 M ☉ . Extended Data Table 1 Best fit values (50 ± 34 percentile) to the optically thin model of gas emission. Left column indicates values derived from the MCMC run where a Gaussian prior was applied to the inclination. Right column indicates values derived from the MCMC run where flat priors were applied to all model parameters Full size table Supplementary information . Peer Review File . Rights and permissions . Reprints and Permissions About this article . Cite this article . Schneiderman, T., Matrà, L., Jackson, A.P. et al. Carbon monoxide gas produced by a giant impact in the inner region of a young system. Nature 598, 425–428 (2021). https://doi.org/10.1038/s41586-021-03872-x Download citation Received : 12 April 2021 Accepted : 01 August 2021 Published : 20 October 2021 Issue Date : 21 October 2021 DOI : https://doi.org/10.1038/s41586-021-03872-x Share this article . Anyone you share the following link with will be able to read this content: Get shareable link Sorry, a shareable link is not currently available for this article. Copy to clipboard Provided by the Springer Nature SharedIt content-sharing initiative Comments . By submitting a comment you agree to abide by our Terms and Community Guidelines . If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate. .
From:
监测目标主题     
(1)  
(1)  
(1)  
系统抽取主题     
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)  
(1)