Красное пятно на полюсе Харона

Красное пятно на полюсе Харона

• 1.
  0 0 MO N T H 2 0 1 6 | V O L 0 0 0 | N A T U R E | 1 LETTER doi:10.1038/nature19340 The formation of Charon’s red poles from seasonally cold-trapped volatiles W. M. Grundy1 , D. P. Cruikshank2 , G. R. Gladstone3 , C. J. A. Howett4 , T. R. Lauer5 , J. R. Spencer4 , M. E. Summers6 , M. W. Buie4 , A. M. Earle7 , K. Ennico2 , J. Wm. Parker4 , S. B. Porter4 , K. N. Singer4 , S. A. Stern4 , A. J. Verbiscer8 , R. A. Beyer2,9 , R. P. Binzel7 , B. J. Buratti10 , J. C. Cook4 , C. M. Dalle Ore2,9 , C. B. Olkin4 , A. H. Parker4 , S. Protopapa11 , E. Quirico12 , K. D. Retherford3 , S. J. Robbins4 , B. Schmitt12 , J. A. Stansberry13 , O. M. Umurhan2 , H. A. Weaver14 , L. A. Young4 , A. M. Zangari4 , V. J. Bray15 , A. F. Cheng14 , W. B. McKinnon16 , R. L. McNutt Jr14 , J. M. Moore2 , F. Nimmo17 , D. C. Reuter18 & P. M. Schenk19 & the New Horizons Science Team* A unique feature of Pluto’s large satellite Charon is its dark red northern polar cap1 . Similar colours on Pluto’s surface have been attributed2 to tholin-like organic macromolecules produced by energetic radiation processing of hydrocarbons. The polar location on Charon implicates the temperature extremes that result from Charon’s high obliquity and long seasons in the production of this material. The escape of Pluto’s atmosphere provides a potential feedstock for a complex chemistry3,4 . Gas from Pluto that is transiently cold-trapped and processed at Charon’s winter pole was proposed1,2 as an explanation for the dark coloration on the basis of an image of Charon’s northern hemisphere, but not modelled quantitatively. Here we report images of the southern hemisphere illuminated by Pluto-shine and also images taken during the approach phase that show the northern polar cap over a range of longitudes. We model the surface thermal environment on Charon and the supply and temporary cold-trapping of material escaping from Pluto, as well as the photolytic processing of this material into more complex and less volatile molecules while cold-trapped. The model results are consistent with the proposed mechanism for producing the observed colour pattern on Charon. The most prominent colour feature in New Horizons images of Charon is its reddish northern polar cap. Figure 1a combines the blue (400–550 nm), red (540–700 nm) and near-infrared (NIR; 780–975 nm) channels from the Multispectral Visible Imaging Camera (MVIC), part of New Horizons’ Ralph remote sensing package5,6 . Figure 1b shows how reflectance in these three filters varies with latitude, averaged over the longitudes shown in Fig. 1a. Charon’s surface gets darker and redder towards higher latitudes. Colour ratios of NIR/red and red/ blue show similar latitude dependences to one another2 , suggesting a single-­pigment material with increasing abundance towards the pole. Additional approach images are shown in Extended Data Fig. 1. Longitudinal variability in the NIR/blue colour ratio is shown in Fig. 1c. At higher latitudes, colours are redder across all of the longitudes observed, although the trend is not perfectly uniform. Deviations may be related to local variations in topography or other parameters. The red coloration is interrupted by a few impact craters with diameters of several kilometres. Impacts that size occur rarely, probably much less frequently than once every million years7 , so their existence implies that the red material must accumulate slowly (see Methods). Infrared spectroscopy also supports slow accumulation. Spectra of Charon’s pole are dominated by H2O ice absorptions similar to spectra of lower-latitude regions2 . The data are consistent with up to 10% tholin mixed with H2O ice at the millimetre depths that are probed by the infrared observations, so tholin deposition must occur slowly enough that H2O resupply or upward mixing of H2O by impact gardening can compete. Charon’s south pole is currently in winter night. New Horizons observed the night-side with the Long Range Reconnaissance Imager (LORRI8 ) approximately 2.6 days after the closest approach, illuminated by Pluto-shine. The images reveal a decreasing brightness towards the pole (Fig. 2 and Extended Data Fig. 2) that cannot be attributed to the declining illumination alone, but requires a decreased albedo of the pole relative to equatorial latitudes (see Methods). The sunlit northern hemisphere in LORRI approach images shows a comparable albedo decline towards that pole (Fig. 2d). Charon’s surface temperature responds to solar forcing on ­diurnal (6.39 Earth days) and annual (248 Earth years) timescales. The high obliquity (currently 119°) causes polar latitudes to experience long ­periods of continuous darkness, during which they become extremely cold. Complicating the situation, the eccentricity of Pluto’s ­heliocentric orbit, currently 0.253, results in a factor of 2.8 difference in the ­intensity of sunlight between perihelion and aphelion. To assess the thermal ­history of Charon’s surface, we ran thermophysical ­models9 that track the diurnal and annual vertical heat flow into and out of Charon’s ­surface at different latitudes, accounting for the present-day orbital parameters10 (see Methods). Figure 3a shows model ­thermal ­histories for four different northern hemisphere latitudes over the past few centuries for a nominal thermal inertia of 10 J m−2 K−1  s−1/2 , which is consistent with estimates of Charon’s diurnal thermal inertia from Herschel Space Telescope observations11 . Models for bracketing ­thermal inertias (2.5 and 40 J m−2 K−1  s−1/2 ) are shown in Extended Data Fig. 3, spanning the range reported from icy satellites12 to Kuiper belt objects13 . Charon’s north pole experienced more than a century of continuous, extremely low temperatures from the late 1800s through to the spring equinox in 1989. Lower latitudes experienced briefer periods of continuous extreme cold, and also reached less extreme minimum temperatures. The latest Charon year shown in Fig. 3a provides an incomplete ­picture of Charon’s long-term thermal history because the pole ­precesses and the longitude of perihelion regresses10,14 , both on 3-mil- lion-year (3-Myr) timescales. For an idea of Charon’s longer-term thermal history, we ran thermal models for previous Charon years, selecting one every 4 ×​ 105  yr over the past 3 Myr. Durations of continu- ous periods colder than a 25 K threshold temperature for cold-trapping 1 Lowell Observatory, Flagstaff, Arizona, USA. 2 NASA Ames Research Center, Moffett Field, California, USA. 3 Southwest Research Institute, San Antonio, Texas, USA. 4 Southwest Research Institute, Boulder, Colorado, USA. 5 National Optical Astronomy Observatory, Tucson, Arizona, USA. 6 George Mason University, Fairfax, Virginia, USA. 7 Massachussetts Institute of Technology, Cambridge, Massachusetts, USA. 8 University of Virginia, Charlotteville, Virginia, USA. 9 Carl Sagan Center, SETI Institute, Mountain View, California, USA. 10 NASA Jet Propulsion Laboratory, La Cañada Flintridge, California, USA. 11 University of Maryland, College Park, Maryland, USA. 12 Université Grenoble Alpes, CNRS, IPAG, Grenoble, France. 13 Space Telescope Science Institute, Baltimore, Maryland, USA. 14 Johns Hopkins University Applied Physics Laboratory, Columbia, Maryland, USA. 15 Lunar and Planetary Laboratory, University of Arizona, Tucson, Arizona, USA. 16 Washington University in St Louis, St Louis, Missouri, USA. 17 University of California, Santa Cruz, California, USA. 18 NASA Goddard Space Flight Center, Greenbelt, Maryland, USA. 19 Lunar and Planetary Institute, Houston, Texas, USA. *A list of participants and their affiliations appears at the end of the paper. © 2016 Macmillan Publishers Limited, part of Springer Nature. All rights reserved.
• 2.
  2 | NA T U R E | V O L 0 0 0 | 0 0 M O N T H 2 0 1 6 LETTERRESEARCH (derived below) were averaged over the modelled epochs and also over the northern and southern hemispheres. These are shown as a function of latitude in Fig. 3b. Higher latitudes experience longer periods below the threshold temperature. CH4 and N2 currently escape from Pluto at rates of 5 ×​ 1025 and 1 ×​ 1023 molecules s−1 , respectively, as estimated from New Horizons data4 , with the Pluto exobase located at around 2.5 Pluto radii (that is, at an altitude of approximately 1,780 km)15 . These are different from conditions assumed in a pre-encounter study3 , which ­investigated the transfer of Pluto’s escaping atmosphere to Charon including ­condensation on the winter pole. This model was based on an N2 escape rate from Pluto of 2.3 ×​ 1027 molecules s−1 , of which 5.7 ×​ 1025 ­molecules s−1 (that is, around 2.5%) encountered Charon, leading to N2 ice being deposited at a rate of 0.2 μ​m per decade on the winter pole. More recent simulations16 report comparable arrival fractions. The long-term temporal variability of the escape rate is not known. Assuming that Charon intercepts 2.5% of the CH4 escape flow ­determined by New Horizons data, that would correspond to a ­globally averaged arrival rate of 2.7 ×​ 1011 molecules m−2  s−1 at Charon. We estimate that most of the CH4 remains at Charon long enough to find its way to the winter pole (see Methods). If it accumulates within 45° of the winter pole (a typical polar size, see Fig. 3b), it would produce a layer of CH4 ice that is approximately 0.3 μ​m thick at each pole of Charon during the winter portion of the Pluto year. The accumulation would presumably vary with latitude, being thicker towards the pole. Condensation on a surface depends on the temperature and vapour pressure17 . The pressure at Charon’s surface depends on the area of the polar cold trap. For a 45° radius pole, it can be estimated as 1 ×​ 10−11 Pa. This pressure corresponds to an equilibrium vapour pressure of CH4 ice at a temperature of 25 K (ref. 18). Where the surface temperature is colder, CH4 will tend to freeze out as ice. Methane escaping from Pluto’s atmosphere is accompanied by other minor species. These include about 0.2% N2, various C2 hydrocarbons (at a few tens of parts per million) and radicals such as CH3 (at around 100 parts per million). N2 is more volatile and thus requires lower ­temperatures to be cold-trapped, so it should freeze onto Charon’s surface over a smaller range of latitudes and times of year than CH4 does. Heavier hydrocarbons can condense anywhere on Charon, but would not produce enough of a deposit to be visible. Our hypothesis requires energetic radiation to process the ­seasonally cold-trapped CH4. It is frozen on Charon’s surface only during the polar winter night, so it must be processed rapidly, on the timescale of a century, and only by radiation impinging on the night side. It need not be fully converted into macromolecular solids such as tholins on such a short timescale, only into molecules that are sufficiently non-­volatile to remain on the surface after the pole re-emerges into ­sunlight and warms back up. Charon’s surface is subject to a variety of ­energetic radiation sources, including ultraviolet photons, solar wind charged ­particles, interstellar pickup ions and galactic cosmic rays19,20 . The most ­important night-side source of energetic radiation appears to be solar ultraviolet Lyman alpha photons (Lyα​, 10.2 eV) that have been scattered by the interplanetary medium, with a photon flux of 3.5 ×​ 1011  m−2  s−1 on the night side. For 2.7 ×​ 1011 molecules m−2  s−1 arriving at Charon, ­concentration by cold-trapping in a 45° cold pole ­(approximately 1/7th of Charon’s surface area) results in the accumulation of around 3 nm of ice per Earth year, of which we estimate 21% is ­photolysed (see Methods). There would probably also be some loss to sputtering21 . N2 ice is unaffected by Lyα​22 but some could be processed by cosmic rays. When the winter pole re-emerges into sunlight, CH4 and N2 ­sublimate away rapidly, but heavier, less volatile products remain behind. Assuming that the mass density of these photolytic prod- ucts is double that of CH4 ice, around 40 nm would accumulate a b c d 0.2 0.4 0.6 0.8 1.0 Reflectance ratio 0 +30 +60 +90 Equator –20° 300° 330° +30° +60° NIR 780–975 nm RED 540–700 nm BLUE 400–550 nm Latitude(°) 180 270 0 90 180 Longitude (° E) 1.0 1.5 2.0 +40° +55° +70° +85° NIR/blueratio 400 500 600 700 800 900 1,000 Wavelength (nm) 0.2 0.4 0.6 0.8 1.0 +50°–60° +70°–80° Reflectanceratio 0° 270° 90° 30° 60° 120° Figure 1 | Charon's red northern pole. a, Polar stereographic projection with Ralph’s BLUE, RED and NIR filter images displayed in blue, green and red colour channels, respectively, relative to a Hapke photometric model (see Methods). b, Latitude dependence of the reflectance relative to the photometric model. c, Longitudinal dependence of the NIR/BLUE colour ratio. d, Wavelength dependence at two latitudes (coloured points) compared with spectral models of a laboratory tholin plus a neutral material (grey curves). The vertical bars indicate the standard deviation within each latitude bin; the horizontal bars indicate the filter widths. © 2016 Macmillan Publishers Limited, part of Springer Nature. All rights reserved.
• 3.
  0 0 MO N T H 2 0 1 6 | V O L 0 0 0 | N A T U R E | 3 LETTER RESEARCH per Pluto winter, or 0.16 mm per million Earth years. They will be exposed to other sources of energetic radiation including ultraviolet and extreme-ultraviolet photon radiation directly from the Sun, driv- ing further photolytic chemistry23 as well as sputtering erosion. At Pluto and Charon’s mean heliocentric distance of 39 au, the solar Lyα​ energy flux is 1.9 ×​ 1013  eV m−2  s−1 , and that of the extreme-ultraviolet (>​12.4 eV) is 8.7 ×​ 1011  eV m−2  s−1 . The charged particle flux from the ambient solar wind, coronal mass ejections and interstellar pickup ions is highly variable, and has energies ranging from a few electronvolts to megaelectronvolts. The production efficiency of tholin is not known for Charon’s circumstances, but if it were 50% that would translate to around 30 cm of tholin produced at Charon’s poles over 4 billion years. Strongly coloured tholins have been made in the laboratory from a variety of CH4 and N2 ice mixtures using diverse radiation sources from charged particles to ultraviolet photons24,25 . For irradiation approx- imately 10×​ greater20,25 , experiments that initially produced a red– orange ­ tholin generally go on to produce colourless and much darker materials, indicating carbonization (graphitization). The fact that Charon’s pole is not completely blackened requires a balance between the production and further processing of tholins and processes such as in-falling H2O ice dust, ejecta or micrometeorite impact gardening that would mix the tholins into the uppermost millimetres to few metres of Charon’s H2O regolith. Two distinct sources could contribute dust. Low-velocity dust from Pluto’s small satellites is estimated to produce a 3 cm coating over the course of 4 billion years (ref. 26) with a com- position that is ­probably dominated by H2O ice. This is an order of magnitude slower than our estimated tholin production, although it could be augmented by ejecta from large impacts elsewhere on Charon. Higher-velocity Kuiper belt debris impact rates are poorly constrained, but models based on lunar rates27,28 imply gardening to centimetre depths on a timescale of 107  yr (see Methods), during which time we estimate about a millimetre of tholin accumulates. By diluting the accu- mulating tholin in local substrate material, this shallow gardening may explain the enduring brightness of a few relatively recent craters where accumulating tholin is mixed into more neutral-coloured H2O-rich Figure 2 | Winter pole in Pluto-shine. a, Stack of 99 images showing a bright, sunlit crescent and fainter reflected light from Pluto. North, defined by the angular momentum vector, is up. b, Photometric model assuming a uniform albedo (see Methods). c, Observation/model ratio, showing that southern high latitudes are dark relative to equatorial latitudes. d, Dependence of the ratio on the absolute value of latitude as indicated by the x axis. The data points are for the Pluto-shine-illuminated southern hemisphere. The horizontal bars indicate the width of the latitude bin and the vertical bars show the standard deviation of the mean within each latitude bin. The dashed curve is for the sunlit northern hemisphere (see Methods). 0° –30° –60° 0° 0° –30° –60° 0° 0° –30° –60° 0° a b c d 0 30 60 90 |Latitude| (°) 0.2 0.4 0.6 0.8 1.0 Reflectanceratio Figure 3 | Thermal environment. a, Model surface temperature history for the equator and three northern latitudes on Charon from 1750–2050 for a thermal inertia of 10 J m−2 K−1  s−1/2 . Envelopes for each latitude indicate diurnal minimum and maximum temperatures. The dashed line is the 25 K threshold below which CH4 is cold-trapped. b, Longest continuous duration below that temperature each Charon year, averaged over thermal models spanning the past 3 Myr. Dotted, solid, and dashed curves are for thermal inertias (in J m−2 K−1  s−1/2 ) of 2.5, 10 and 40, respectively. 1750 1800 1850 1900 1950 2000 2050 Year 10 20 30 40 50 60 Surfacetemperature(K) +30° +60° Equator +90° 0 30 60 90 |Latitude| (°) 0 20 40 60 80 100 120 140 Meancold-traplongevity(yr) a b © 2016 Macmillan Publishers Limited, part of Springer Nature. All rights reserved.
• 4.
  4 | NA T U R E | V O L 0 0 0 | 0 0 M O N T H 2 0 1 6 LETTERRESEARCH ejecta, rather than being mixed into the already-darkened substrate, as elsewhere. The distribution of dark, reddish material around Charon’s ­northern pole is notable for its generally symmetric distribution across ­longitudes and its gradual increase with latitude, although there are local ­irregularities associated with craters, topographic features and perhaps subsurface variations in thermal properties. These ­characteristics, and the existence of an albedo feature around the southern pole with a similar latitude dependence, are consistent with our hypothesis that the combination of Pluto’s escaping atmosphere and Charon’s long, cold winters enables CH4 to be seasonally cold-trapped at high ­latitudes, where some is photolytically processed into heavier molecules that are subsequently converted to reddish tholin-like materials. The ­symmetry argues against recent polar wander on Charon. Could the process occur elsewhere? Nix has a reddish spot29 , but it and the other small ­satellites orbit farther from Pluto and have much lower masses, making the ­process less efficient (see Methods). New Horizons Science Team S. A. Stern1 , F. Bagenal2 , K. Ennico3 , G. R. Gladstone4 , W. M. Grundy5 , W. B. McKinnon6 , J. M. Moore3 , C. B. Olkin1 , J. R. Spencer1 , H. A. Weaver7 , L. A. Young1 , T. Andert8 , O. Barnouin7 , R. A. Beyer3 , R. P. Binzel9 , M. Bird10 , V. J. Bray11 , M. Brozović12 , M. W. Buie1 , B. J. Buratti12 , A. F. Cheng7 , J. C. Cook1 , D. P. Cruikshank3 , C. M. Dalle Ore13,3 , A. M. Earle9 , H. A. Elliott4 , T. K. Greathouse4 , M. Hahn14 , D. P. Hamilton15 , M. E. Hill7 , D. P. Hinson13 , J. Hofgartner12 , M. Horányi2 , A. D. Howard16 , C. J. A. Howett1 , D. E. Jennings17 , J. A. Kammer1 , P. Kollmann7 , T. R. Lauer18 , P. Lavvas19 , I. R. Linscott20 , C. M. Lisse7 , A. W. Lunsford17 , D. J. McComas4 , R. L. McNutt Jr.7 , M. Mutchler21 , F. Nimmo22 , J. I. Nunez7 , M. Paetzold14 , A. H. Parker1 , J. Wm. Parker1 , S. Philippe23 , M. Piquette2 , S. B. Porter1 , S. Protopapa15 , E. Quirico23 , H. J. Reitsema1 , D. C. Reuter17 , S. J. Robbins1 , J. H. Roberts7 , K. Runyon7 , P. M. Schenk24 , E. Schindhelm1 , B. Schmitt23 , M. R. Showalter13 , K. N. Singer1 , J. A. Stansberry21 , A. J. Steffl1 , D. F. Strobel25 , T. Stryk26 , M. E. Summers27 , J. R. Szalay2 , H. B. Throop28 , C. C. C. Tsang1 , G. L. Tyler20 , O. M. Umurhan3 , A. J. Verbiscer16 , M. H. Versteeg4 , G. E. Weigle II4 , O. L. White3 , W. W. Woods20 , E. F. Young1 & A. M. Zangari1 . 1 Southwest Research Institute, Boulder, Colorado, USA. 2 University of Colorado, Boulder, Colorado, USA. 3 NASA Ames Research Center, Moffett Field, California, USA. 4 Southwest Research Institute, San Antonio, Texas, USA. 5 Lowell Observatory, Flagstaff, Arizona, USA. 6 Washington University in St. Louis, St. Louis, Missouri, USA. 7 Johns Hopkins University Applied Physics Laboratory, Columbia, Maryland, USA. 8 Universität der Bundeswehr München, Neubiberg 85577, Germany. 9 Massachussetts Institute of Technology, Cambridge, Massachusetts, USA. 10 University of Bonn, Bonn D-53113, Germany. 11 University of Arizona Lunar and Planetary Laboratory, Tucson ARizona, USA. 12 NASA Jet Propulsion Laboratory, La Cañada Flintridge, California, USA. 13 Carl Sagan Center at the SETI Institute, Mountain View, California, USA. 14 Rheinisches Institut für Umweltforschung an der Universität zu Köln, Cologne 50931, Germany. 15 University of Maryland, College Park, Maryland, USA. 16 University of Virginia, Charlottesville, Virginia, USA. 17 NASA Goddard Space Flight Center, Greenbelt, Maryland, USA. 18 National Optical Astronomy Observatory, Tucson, Arizona, USA. 19 Université de Reims Champagne-Ardenne, 51687 Reims, France. 20 Stanford University, Stanford, California, USA. 21 Space Telescope Science Institute, Baltimore, Maryland, USA. 22 University of California, Santa Cruz, California, USA. 23 Université Grenoble Alpes, CNRS, IPAG, Grenoble, France. 24 Lunar and Planetary Institute, Houston, Texas, USA. 25 Johns Hopkins University, Baltimore, Maryland, USA. 26 Roane State Community College, Jamestown, Tennessee, USA. 27 George Mason University, Fairfax, Virginia, USA. 28 Planetary Science Institute, Tucson, Arizona, USA. Acknowledgements This work was supported by NASA’s New Horizons Project. E.Q., B.S. and S.Phi. acknowledge the Centre National d’Etudes Spatiales (CNES) for its financial support through its ‘Système Solaire’ programme. Author Contributions W.M.G. led the study and wrote the paper, with significant input from G.R.G. and M.E.S. on the escape of Pluto’s atmosphere and its transport to Charon, D.P.C. on radiolytic production of tholins, C.J.A.H. and J.R.S. on thermal models, and T.R.L. on processing of Pluto-shine images. M.W.B., A.J.V., and B.J.B. developed photometric models. A.M.E. and R.P.B. computed insolation as functions of time and location. K.E. processed MVIC approach images. G.R.G., M.E.S., J.Wm.P., and K.D.R. assessed ultraviolet irradiation from various sources. K.N.S., S.B.P., and S.J.R. assessed effects of impacts from dust and larger projectiles. S.A.S. contributed diverse insights and led the overall mission. R.A.B., P.M.S., A.H.P., and A.M.Z. helped with geometric registration and image processing. C.M.D.O., J.C.C., S.Pro, C.B.O., E.Q., B.S., and S.Phi. interpreted infrared spectral data. J.A.S. and O.M.U. contributed insights on cold trapping of volatiles. R.L.M. contributed insights on the charged particle environment. L.A.Y. led development of the observation sequences that produced the data used here, and also checked the thermal models. H.A.W. and A.F.C. played key roles in LORRI instrument development and data processing, and D.C.R., C.B.O., A.H.P., and C.J.A.H. did the same for Ralph. W.B.M., J.M.M., K.N.S., F.N., P.M.S., and V.J.B. contributed insights on geophysics and geology of Charon. Author Information All spacecraft data presented in this paper will be delivered to NASA’s Planetary Data System (https://pds.nasa.gov) in a series of stages in 2016 and 2017 in accordance with the schedule established by NASA and the New Horizons project. Reprints and permissions information is available at www.nature.com/reprints. The authors declare no competing financial interests. Readers are welcome to comment on the online version of the paper. Correspondence and requests for materials should be addressed to W.G. (w.grundy@lowell.edu). Reviewer Information Nature thanks L. Trafton and the other anonymous reviewer(s) for their contribution to the peer review of this work. Online Content Methods, along with any additional Extended Data display items and Source Data, are available in the online version of the paper; references unique to these sections appear only in the online paper. Received 12 April; accepted 13 July 2016. Published online 14 September 2016. 1. Stern, S. A. et al. The Pluto system: initial results from its exploration by New Horizons. Science 350, aad1815 (2015). 2. Grundy, W. M. et al. Surface compositions across Pluto and Charon. Science 351, aad9189 (2016). 3. Tucker, O. J., Johnson, R. E. & Young, L. A. Gas transfer in the Pluto–Charon system: a Charon atmosphere. Icarus 246, 291–297 (2015). 4. Gladstone, G. R. et al. The atmosphere of Pluto as observed by New Horizons. Science 351, aad8866 (2016). 5. Reuter, D. C. et al. Ralph: a visible/infrared imager for the New Horizons Pluto/ Kuiper belt mission. Space Sci. Rev. 140, 129–154 (2008). 6. Howett, C. J. et al. Inflight radiometric calibration of New Horizons’ Multispectral Visible Imaging Camera (MVIC). Preprint at http://arxiv.org/ abs/1603.08940. 7. Moore, J. M. et al. The geology of Pluto and Charon through the eyes of New Horizons. Science 351, 1284–1293 (2016). 8. Cheng, A. F. et al. Long-Range Reconnaissance Imager on New Horizons. Space Sci. Rev. 140, 189–215 (2008). 9. Spencer, J. R., Lebofsky, L. A. & Sykes, M. V. Systematic biases in radiometric diameter determinations. Icarus 78, 337–354 (1989). 10. Earle, A. M. & Binzel, R. P. Pluto’s insolation history: latitudinal variations and effects on atmospheric pressure. Icarus 250, 405–412 (2015). 11. Lellouch, E., et al. The long-wavelength thermal emission of the Pluto–Charon system from Herschel observations: evidence for emissivity effects. Astron. Astrophys. 588, A2 (2016). 12. Howett, C. J. A., Spencer, J. R., Pearl, J. & Segura, M. Thermal inertia and bolometric Bond albedo values for Mimas, Enceladus, Tethys, Dione, Rhea, and Iapetus as derived from Cassini/CIRS measurements. Icarus 206, 573–593 (2010). 13. Lellouch, E. et al. “TNOs are Cool”: A survey of the trans-Neptunian region IX. Thermal properties of Kuiper belt objects and Centaurs from combined Herschel and Spitzer observations. Astron. Astrophys. 557, A60 (2013). 14. Dobrovolskis, A. R. & Peale, S. J & Harris. A. W. in Pluto and Charon (eds Stern, S. A. & Tholen, D. J.) 159–190 (Univ. Arizona Press, 1997). 15. Bagenal, F. et al. Pluto’s interaction: solar wind, energetic particles, dust. Science 351, aad9045 (2016). 16. Hoey, W. A., Yeoh, S. K., Trafton, L. M., Goldstein, D. B. & Varghese, P. L. Rarefied gas dynamic simulation of transfer and escape in the Pluto–Charon system with the DSMC method. In Proc. 47th Lunar and Planetary Science Conference Contribution No. 1903 (Lunar And Planetary Institute, 2016). 17. Shorghofer, N. & Taylor G. J. Subsurface migration of H2O at lunar cold traps. J. Geophys. Res. 112, E02010 (2007). 18. Fray, N. & Schmitt, B. Sublimation of ices of astrophysical interest: a bibliographic review. Planet. Space Sci. 57, 2053–2080 (2009). 19. Johnson, R. E. Energetic Charged-Particle Interactions with Atmospheres and Surfaces (Springer, 1990). 20. Johnson, R. E. Effect of irradiation on the surface of Pluto. Geophys. Res. Lett. 16, 1233–1236 (1989). 21. Mejía, C. et al. Compaction of porous ices rich in water by swift heavy ions. Icarus 250, 222–229 (2015). 22. Cruz-Diaz, G. A., Muñoz Caro, G. M., Chen, Y. J. & Yih, T. S. Vacuum-ultraviolet spectroscopy of interstellar ice analogs II: absorption cross-sections of nonpolar ice molecules. Astron. Astrophys. 562, A120 (2014). 23. Madey, T. E., Johnson, R. E. & Orlando, T. M. Far-out surface science: radiation-induced surfaces processes in the solar system. Surf. Sci. 500, 838–858 (2002). 24. Materese, C. K., Cruikshank, D. P., Sandford, S. A., Imanaka, H. & Nuevo, M. Ice chemistry on outer solar system bodies: electron radiolysis of N2-, CH4-, and CO-containing ices. Astrophys. J. 812, 150 (2015). 25. Thompson, W. R., Murray, B. G. J. P. T., Khare, B. N. & Sagan, C. Coloration and darkening of methane clathrate and other ices by charged particle irradiation: applications to the outer solar system. J. Geophys. Res. 92, 14,933–14,947 (1987). 26. Porter, S. B. & Grundy, W. M. Ejecta transfer in the Pluto system. Icarus 246, 360–368 (2015). 27. Gault, D. E., Hörz, F., Brownlee, D. E. & Hartung, J. B. Mixing of the lunar regolith. Lunar Sci. Conf. Proc. 3, 2365–2386 (1974). 28. Spencer, J. R. & Denk, T. Formation of Iapetus’ extreme albedo dichotomy by exogenically triggered thermal ice migration. Science 327, 432–435 (2010). 29. Weaver, H. A. et al. The small satellites of Pluto as observed by New Horizons. Science 351, aae0030 (2016). © 2016 Macmillan Publishers Limited, part of Springer Nature. All rights reserved.
• 5.
  LETTER RESEARCH METHODS Colour images.The colour image of Charon shown in Fig. 1a was obtained on 2015 July 14 at 10:42 ut. The spacecraft was 73,000 km from Charon, result- ing in an MVIC image scale of 1.4 km per pixel. The Mission Elapsed Time (MET) unique label of this observation was 0299176432. Calibration details are described in ref. 6. The original image is shown in Extended Data Fig. 3c. Figure 1a showsthisimagereprojectedonapolarstereographicprojectionand­photometrically corrected by dividing each pixel by a Hapke model30 computed for the same ­illumination and viewing geometry. That model had uniform parameters across the scene: single-scattering albedo w =​ 0.9, single-scattering phase ­function P(g) =​ 0.8, backscattering amplitude B0 =​ 0.6, porosity parameter h =​ 0.0044 and macroscopic roughness θ =​ 20° (parameters from ref. 31). Figure 1b shows the latitude ­dependence of the image divided by the Hapke model, averaged over the longitudes shown in Fig. 1a and normalized to unity at the equator. A similar latitude dependence in Charon’s normal albedo has been reported from LORRI observations32 . The longitude and wavelength dependence are shown in Fig. 1c, d, respectively. More approach images are shown in Extended Data Fig. 1, without reprojection or photometric correction. Pluto-shine observations. New Horizons obtained 219 LORRI images of Charon’s night side early on 2015 July 17 ut from a distance of approximately 3 ×​ 106  km. The spacecraft’s orientation is controlled by hydrazine thrusters, so full-resolution LORRI exposures must be short to minimize smear as pointing bounces around within a deadband. Longer exposures of 0.2 s were enabled by 4 ×​ 4 pixel on-chip binning, resulting in a total integration of 44 s. The images were grouped into two sets, the first comprising 99 images with MET labels 0299398349 through 0299398716 acquired at a mean observation time of 00:26 ut. The second set had 120 images with MET labels 0299405549 through 0299405916 and a mean time of 02:26 ut. The two sets had image scales of 60 km per pixel and 62 km per pixel, respectively. As LORRI was pointed close to the Sun to observe the night side of Charon, the images were affected by scattered light patterns modified by small pointing ­variations among the images. The scattered light variability in each image set was modelled using principal component analysis on the complementary set, after masking Charon and bright stars from the images. The eigenimages in each set were used to subtract the scattered light contribution from the other set. The images were then co-registered at the sub-pixel level and combined into two stacks, the first of which is shown in Fig. 2a, the second in Extended Data Fig. 2a. We modelled these observations by simulating Charon as a sphere ­approximated by 20,480 triangular facets. The bidirectional reflectance behaviour of each facet was represented with a Hapke model, using the same parameters as above. The Sun was treated as a point source and the distribution of light from Pluto was obtained by assuming it to be a Lambertian sphere with the appropriate size, ­location and illumination geometry. Being larger than Charon, Pluto casts some light onto Charon’s pole, albeit obliquely. Comparisons between data and model are shown for the first of the two image stacks in Fig. 2 and for the other stack in Extended Data Fig. 2. To compare Charon’s southern pole with the northern one as seen with the same instrument, we also ran a Hapke model with the same parameters for a LORRI full-resolution sunlit approach image obtained 2015 July 14 at 02:44 ut, from a distance of 470,000 km, with image scale of 2.3 km per pixel. The MET label for this image was 299147776. It and the corresponding model are shown in Extended Data Fig. 2d and e and the latitude-dependent ratio is included as a dashed curve in Fig. 2d and Extended Data Fig. 2f, normalized to unity over the mean from 0° to 10° N. MVIC colour observations were not sufficiently sensitive to detect Pluto- shine on Charon’s night side. Thermal model. A standard one-dimensional finite element model9 was used to account for the vertical heat flow within each element of Charon’s surface in response to diurnally and seasonally varying insolation10 . Diurnal and annual timescales differ greatly, so we represented Charon’s surface with a large number of layers (400) and broke its year into a large number of time steps (2 ×​ 106 , about an hour per time step). We assumed a uniform bolometric bond albedo AB =​ 0.3 and emissivity ε =​ 0.9, ignoring the lower albedos of Charon’s poles. Diurnal ­thermal inertia Γ values were varied from 2.5 to 40 J m−2 K−1  s−1/2 , spanning the range reported for Charon and Kuiper belt objects from Spitzer and Herschel ­observations11,13 to icy Saturnian satellites observed by Cassini12 . The density and heat capacity of H2O ice are fixed, so low thermal inertias require inefficient ­conduction between ice grains. Conduction can rise with compaction at depth, leading to higher seasonal thermal inertias. Simulations with higher conductivities up to that of monolithic H2O ice33 at depths as shallow as few tens of centimetres below the surface can raise polar winter temperatures by a few kelvin without greatly affecting diurnal temperature variations. Sunlight was assumed to be the only energy source, providing 870 mW m−2 at Pluto’s mean heliocentric distance. We did not include radiogenic heating. If Charon’s rock fraction of around 60% (ref. 34) has chondritic radionuclide abundances, heat from their decay would ­contribute an additional 1.5 mW m−2 , preventing the temperature from falling below about 13 K at present, and higher in the distant past. Figure 3a shows the model ­surface temperature history over the course of three centuries for Charon’s northern ­hemisphere, assuming an intermediate diurnal thermal inertia value of Γ =​ 10 J m−2 K−1  s−1/2 , no radiogenic heat, and no high-conductivity ­subsurface layer. High and low Γ bounding cases are shown in Extended Data Fig. 3. Diurnal variations are much greater for small Γ, as is the range of seasonal surface ­temperature extremes. For all of these scenarios, Charon’s high latitudes experience multi-decade episodes that are sufficiently cold to cold-trap CH4 as ice. As the spin pole of the system precesses and the longitude of the perihelion of its heliocentric orbit regresses10,14 , the insolation patterns evolve, modifying the ­surface temperature history over the course of a Pluto year. This variability is accommodated in Fig. 3b by averaging the longest continuous period below 25 K during a Pluto year over both hemispheres and over three million years of orbital history. Loss mechanisms and cold-trapping. CH4 molecules can be lost from Charon’s surface environment by several mechanisms. First we consider thermal escape to space. For a Maxwell–Boltzmann distribution at 60 K, corresponding to the highest temperatures reached on Charon’s summer hemisphere (see Fig. 3a), about 1% of CH4 molecules exceed Charon’s 590 m s−1 escape velocity. The majority, which do not escape hop on ballistic trajectories until they achieve escape velocity or encounter the cold pole and stick. A typical latitude boundary of the cold pole from Fig. 3b is 45°. A cold pole above that latitude occupies about 1/7th of Charon’s surface. If each ballistic hop is assumed to arrive in a completely new random location, it would take an average of seven random hops to encounter the cold pole at a time when it is that size. With 1% lost to space per hop, 93% would make it to the pole, leading to an accumulation rate of 2 ×​ 1012 molecules m−2  s−1 . For a CH4 ice density35 of 516 kg m−3 , the deposition rate is 9 ×​ 10−17 m s−1 which adds up to about 0.3 μ​m at the pole over one Charon winter. Because the accumulation time is strongly latitude-dependent, the resulting ice distribution would be, too. Another potential loss mechanism is ionization. If a molecule is ionized, it becomes coupled to the magnetic field in the solar wind and is thus swept from the system. The cross-section of a CH4 gas molecule to Lyα​radiation is estimated22 as 1.8 ×​ 10−21 m2 . The solar Lyα​flux at Charon’s mean heliocentric distance of 39 au is 1.9 ×​ 1012  cm−2  s−1 , so the probability of photoionization is about 0.1 per Earth year. At typical thermal speeds of 200–300 m s−1 , CH4 molecules can traverse Charon’s diameter in less than 2 h, so the probability of photoionization during seven ballistic hops is low. Likewise, little CH4 would be photoionized between its escape from Pluto and arrival at Charon considering that Pluto’s exobase ­temperature is around 70 K, so molecules escaping at the tail of the Maxwellian velocity distribution would reach Charon in just a few tens of hours. Surface pressure. An estimate of the surface pressure at Charon can be obtained by assuming steady state, with arriving CH4 molecules undergoing an average of seven ballistic hops before being lost to space or cold-trapped on a 45° radius cold pole, so each element of Charon’s surface experiences seven times the globally averaged CH4 arrival flux of 2.7 ×​ 107 molecules cm−2  s−1 at speeds that are consistent with the Boltzmann velocity distribution, resulting in a momentum flux or pressure of roughly 1 ×​ 10−11 Pa, which is three orders of magnitude smaller than the 3σ upper limit from New Horizons36 . When the cold pole is smaller, the pressure increases because more hops occur, and the reverse is true when the cold pole is larger, so this pressure is very approximate. However, it enables us to estimate the threshold temperature of 25 K for cold-trapping CH4 via its vapour pressure18 . Assuming the radius of a CH4 molecule is 2 ×​ 10−10 m, this pressure corresponds to a mean free path that is >​100 Charon radii, so collisions between CH4 gas molecules can be ignored. Photolysis of CH4 ice. The interplanetary medium scatters Lyα​photons from the Sun. The flux from this source observed during the New Horizons flyby4 was about 50% larger than predicted37 , or about 140 R averaged over the anti-Sun ­hemisphere, which illuminates the winter pole, equating to a flux of 3.5 ×​ 1011 Lyα​ photons m−2  s−1 . This appears to be the largest source of night-side ultraviolet ­illumination at wavelengths between 10 nm and 133 nm that efficiently break bonds in the CH4 ­molecule22,38 . Integrating over these wavelengths for the 1,000 stars with largest apparent ultraviolet fluxes using IUE spectra and Kurucz39 models and dividing by two to account for the visible 2π​steradians of sky contributes an additional 1.1 ×​ 1010 ­photons m−2  s−1 . The Lambert absorption coefficient of CH4 ice at the Lyα​ ­wavelength (1216 Å)38 is 19 μ​m−1 so a 53-nm-thick CH4 ice coating has ­optical depth unity. The Lyα​cross-section of individual CH4 ice molecules has been reported22 as 1.4 ×​ 10−21  m2 , enabling an estimation of the probability of photolysis of an unshielded CH4 ice molecule on Charon’s winter pole as 1.5% per Earth year. For 2.7 ×​ 1011 molecules m−2  s−1 arriving at Charon, ­concentration by cold-­trapping on ­approximately 1/7th of Charon’s surface area results in the © 2016 Macmillan Publishers Limited, part of Springer Nature. All rights reserved.
• 6.
  LETTERRESEARCH ­accumulation of around3 nm of CH4 ice per Earth year, or a layer that is ­optically thick to Lyα​in 18 yr. Only the skin reachable by ultraviolet photons can be pho- tolysed. For steady state at this ­deposition rate, around 26% of the CH4 would be ­processed before becoming buried beneath enough CH4 ice to shield it from further ultraviolet ­photolysis. Over a ­century-long accumulation cycle at this rate, about 21% is ­processed. Impacts and gardening. Only one of the bright-ray craters in Charon’s polar region appears to be resolved. Its darker, inner region could mark the rim, or it could be ejecta given that many of Charon’s other craters feature dark inner ejecta. If the dark/bright boundary is the rim, the diameter of the crater itself is around 5–6 km; otherwise, scaling from fully resolved craters, the diameter could be as small as approximately 1.5 km. In addition to uncertainty about crater size, the impactor flux in the outer Solar System for objects below around 100 km is not well known. Using the ‘knee’ model40,41 (consistent with the size distribution of larger craters on Pluto and Charon7 ), one crater 5–6 km or larger in diameter should typically form somewhere on Charon approximately every million years. The probability of occurrence in dark polar material is lower, in proportion to its smaller surface area. If the crater is smaller than 5–6 km, the occurrence frequency would be higher, but the knee model used predicts many more craters smaller than 10 km than are observed on Pluto or Charon7 , suggesting that the timescale for the formation of the observed craters could be considerably longer than a million years. No detailed modelling has so far been conducted for impact gardening of outer Solar System objects. Two studies of lunar impact gardening found similar results for 107  yr timescales27,42 . A 50% probability that material has been excavated down to a depth of 1 cm over that interval was estimated27 , while ref. 42 found that all of the study points in a grid were excavated at least once down to a depth of around 1 cm. The two studies diverged for longer timescales, with ref. 42 deriving a ­disturbed depth of 10 cm in 108  yr and the model of ref. 27 taking 109  yr to achieve the same disturbed depth. The rate of gardening in the lunar example is calculated for impactors of all sizes, from the largest objects to hit the Moon down to micrometeorites and dust- sized particles. Extrapolation to the outer Solar System requires knowledge of the impactor flux, which, as described above, is poorly constrained. The lunar flux used by ref. 27 is larger than the best estimate from refs 40 and 41 for Charon for ­impactors larger than around 0.1 mm by up to several orders of magnitude, but smaller than the extrapolated knee model for smaller impactors by similar amounts. The dust flux (particles below 1 cm in diameter) estimated from the New Horizons Student Dust Counter43 is four to six orders of magnitude below the knee model. Extrapolating from that model to much smaller sizes or from the New Horizons dust measurement to much larger sizes is highly uncertain, illustrating the broad range of the potential impactor flux. Other instances. A natural question to ask is whether our hypothesized ­mechanism for the production of Charon’s dark, red poles should produce ­similar deposits elsewhere. Unlike the case for Charon, escape velocities on Pluto’s smaller satellites are negligible compared with the thermal speeds. They intercept Pluto’s escaping atmosphere only in proportion to their geometric cross-sections, with no ­enhancement effect from CH4 molecules ballistically hopping around their ­surfaces. We consider Nix, as it exhibits a prominent red spot29 , although similar calculations could be done for any of the small satellites. Nix’s effective radius is 20 km and its orbital distance is 48,760 km, so it intercepts 1.7 ×​ 10−7 of Pluto’s escaping CH4, or 8 ×​ 1018 molecules s−1 . Nix’s escape velocity is roughly 10–20 m s−1 , so only CH4 molecules that happen to hit an extremely cold region on their first impact will stick. Assuming that Nix has a long-duration cold pole like Charon does (it might not, if its pole precesses rapidly), the polar orientation with respect to radially outflowing gas from Pluto would diminish the accumulation rate by another factor of four, leading to an accumulation rate of around 1 ×​ 10−7  μ​m yr, about 20,000×​slower than the accumulation rate at Charon’s cold pole. Such slow accumulation might not be competitive with impact erosion of Nix’s surface. Code availability. New Horizons MVIC and LORRI images were processed using the US Geological Survey’s Integrated Software for Images and Spectrometers (ISIS), available at https://isis.astrogeology.usgs.gov. Thermal models were based on J.R. Spencer’s thermprojrs.pro, available at https://www.boulder.swri.edu/~​ spencer/thermprojrs. Hapke reflectance models were based on M.W. Buie’s bidr2. pro, available at http://www.boulder.swri.edu/~​buie/idl/pro/bidr2.html. 30. Hapke, B. Theory of Reflectance and Emittance Spectroscopy (Cambridge Univ. Press, 1993). 31. Buie, M. W., Grundy, W. M., Young, E. F., Young, L. A. & Stern, S. A. Pluto and Charon with the Hubble Space Telescope I: monitoring global change and improved surface properties from light curves. Astron. J. 139, 1117–1127 (2010). 32. Buratti, B. J. et al. Global albedos of Pluto and Charon from LORRI New Horizons observations. Preprint at http://arxiv.org/abs/1604.06129 (2016). 33. Ross, R. G. & Kargel, J. S. in Solar System Ices (eds Schmitt, B. et al.) 33–62 (Kluwer Academic, 1998). 34. McKinnon, W. B. et al. The Pluto–Charon system revealed: geophysics, activity, and origins. In Proc. 47th Lunar and Planetary Science Conference Contribution No. 1995 (Lunar and Planetary Institute, 2016). 35. Bol’shutkin, D. N., Gasan, V. M. & Prokhvatilov, A. I. Temperature dependence of the crystalline lattice parameter of methane in the range of 11–70 °K. J. Struct. Chem. 12, 670–672 (1972). 36. Stern, S. A. et al. New Horizons constraints on Charon’s present day atmosphere. Preprint available at http://arxiv.org/abs/1608.06955 (2016). 37. Gladstone, G. R., Pryor, W. R. & Stern, S. A. Lyα​@Pluto. Icarus 246, 279–284 (2015). 38. Baratta, G. A., Leto, G. & Palumbo, M. E. A comparison of ion irradiation and ultraviolet photolysis of CH4 and CH3OH. Astron. Astrophys. 384, 343–349 (2002). 39. Kurucz, R. L. SYNTHE Spectrum Synthesis Programs and Line Data (Smithsonian Astrophysical Observatory, 1993). 40. Greenstreet, S., Gladman, B. & McKinnon, W. B. Impact and cratering rates onto Pluto. Icarus 258, 267–288 (2015). 41. Greenstreet, S. Gladman, B. & McKinnon, W. B. Corrigendum to “Impact and cratering rates onto Pluto” [Icarus 258 (2015) 267–288]. Icarus 274, 366–367 (2016). 42. Arnold, J. R. A Monte Carlo model for the gardening of the lunar regolith. Moon 13, 159–172 (1975). 43. Poppe, A. R. Interplanetary dust influx to the Pluto–Charon system. Icarus 246, 352–359 (2015). © 2016 Macmillan Publishers Limited, part of Springer Nature. All rights reserved.
• 7.
  LETTER RESEARCH Extended DataFigure 1 | Three MVIC colour images obtained on approach showing Charon’s northern pole as Charon rotates. a, Observation obtained 2015 July 11 3:35 ut, with MET label 0298891582. b, Observation obtained 2015 July 13 3:38 ut, with MET label 0299064592. c, The same observation as Fig. 1a of the main text, obtained 2015 July 14 at 10:42 ut, with MET label 0299176432. Unlike in Fig. 1a, these images are not re-projected or divided by photometric models. North, as defined by the angular momentum vector, is up. © 2016 Macmillan Publishers Limited, part of Springer Nature. All rights reserved.
• 8.
  LETTERRESEARCH Extended Data Figure2 | Additional panchromatic observations of Charon’s poles. a, Second stack of 120 images of Charon’s southern hemisphere illuminated by Pluto-shine obtained approximately 2 h after the stack in Fig. 2a. b, Corresponding photometric model. c, Observation/ model ratio. d, Sunlit northern hemisphere. e, Corresponding photometric model. f, Similar to Fig. 2d, with the first Pluto-shine stack indicated by blue points and the second stack indicated by red points (offset left and right for clarity). The horizontal bars indicate the widths of the latitude bins and the vertical bars indicate the standard deviation of the mean within each latitude bin. © 2016 Macmillan Publishers Limited, part of Springer Nature. All rights reserved.
• 9.
  LETTER RESEARCH Extended DataFigure 3 | Thermal models for cases of low and high thermal inertia. a, Thermal history for lower limit thermal inertia Γ =​ 2.5 J m−2 K−1  s−1/2 . b, Thermal history for upper limit thermal inertia Γ =​ 40 J m−2 K−1  s−1/2 . © 2016 Macmillan Publishers Limited, part of Springer Nature. All rights reserved.