In July 2015, NASA’s New Horizons mission performed a flyby of Pluto, revealing details about the geology, surface composition and atmospheres of this world and its moons that are unobtainable from Earth. With a resolution as small as 80 metres per pixel, New Horizons’ images identified a large number of surface features, including a large basin filled with glacial ices that appear to be undergoing convection. Maps of surface composition show latitudinal banding, with non-volatile material dominating the equatorial region and volatile ices at mid- and polar latitudes. This pattern is driven by the seasonal cycle of solar insolation. New Horizons’ atmospheric investigation found the temperature of Pluto’s upper atmosphere to be much cooler than previously modelled. Images of forward-scattered sunlight revealed numerous haze layers extending up to 200 km from the surface. These discoveries have transformed our understanding of icy worlds in the outer Solar System, demonstrating that even at great distances from the Sun, worlds can have active geologic processes. This Review addresses our current understanding of the Pluto system and places it in context with previous investigations.
Before the July 2015 flyby of New Horizons through the Pluto system, astronomers had little information about the surface appearance of Pluto and its largest moon Charon. They knew that Pluto was redder (and getting redder with time) and had a higher albedo than Charon1,2,3, and also had a bright polar cap4 and a dark equatorial band5. There was considerable uncertainty about Pluto’s exact size: estimates of its radius ranged from 1,151 km to 1,178 km (refs 6,7). Charon’s diameter was estimated to be approximately half that of Pluto’s, and there was considerable uncertainty regarding the size of Pluto’s small moons. Figure 1 shows some key milestones in the quest to obtain ever-better spatial imagery of these far-away worlds.
Light curves to maps
The small angular sizes of Pluto and Charon — 0.1 arcseconds and 0.05 arcseconds at perihelion, respectively (in 1989) — makes it difficult to map their surfaces using ground-based telescopes (Box 1). The first surface ‘maps’ were constructed by indirect light-curve inversion techniques thanks to observations of mutual transit and occultation events of Pluto and Charon8,9,10,11, and this was followed by direct mapping using the Hubble Space Telescope (HST)3,12,13. These maps had only a few resolution elements across Pluto’s disk. Combined with hemispheric-composition data from ground-based spectra (see 'Compositional complexity in the Pluto system'), they formed the basis for choosing the New Horizons Pluto flyby encounter hemisphere (the face observed at closest approach) and multi-instrument observation sequence14. From seven days before closest approach until two days after, the New Horizons instrument suite made more than 400 different observations of Pluto and its moons.
Box 1 | From points of light to geological worlds
Data from New Horizons provided significant leaps in spatial resolution over earlier HST and ground-based images (Fig. 1). Clyde Tombaugh at Lowell Observatory in Flagstaff, Arizona, discovered Pluto in early 1930 after over nearly a year of searching, using a 32.5 cm telescope and recording images on photographic plates56. He used a blink comparator, an instrument that rapidly shifted back and forth between views of each of the plates, to reveal the movement of any object that had changed position or appearance between photographs. Decades later, with the launch of the HST, observations of Pluto with a resolution of around 200 km pixel–1 were obtained by the HST’s Faint Object Camera in 199413. Although the image was too coarse to provide information on Pluto’s surface geology, it confirmed significant albedo differences across the surface, as previously inferred from the mutual event observations in the 1980s. Subsequent HST imagery of Pluto taken in 2002–2003 showed its brightness had changed, with its north pole getting brighter and its southern regions getting redder and darker3. Images from New Horizons transformed our understanding of Pluto from an astronomer’s world to a geologic one, with spatial resolutions in selected areas as small as 80 m pixel–1. The closest-approach hemisphere is rich in geological diversity, with the most prominent feature being a heart-shaped bright region, the western half of which is a basin filled with nitrogen, methane and carbon monoxide ices, superimposed on a dark equatorial band. The New Horizons images also provided a means of determining Pluto’s size. The mean radius of Pluto is 1,188 ± 1.6 km (ref. 23) and there is no evidence of distortion.
Charon, Pluto’s largest moon, was discovered in 1978 by Christy and Harrington using the 1.55 m telescope at the US Naval Observatory Flagstaff Station57,58. The images showed a slight elongation in Pluto’s appearance, which occurred periodically with an ~6.4 day cadence — the same as Pluto’s rotational period. This elongation was eventually understood to be a moon (Charon) in orbit around Pluto. With less than 1 arcsec separation, resolving Pluto and Charon was challenging, but has since become more routine using 8 m ground-based facilities with adaptive optics, and high-angular-resolution submillimetre facilities such as the Atacama Large Millimeter/submillimeter Array and the HST. The New Horizons images transformed our understanding of Charon, just as they did for Pluto. The most prominent features on Charon’s closest-approach hemisphere are a dark red polar region and a tectonic belt that splits Charon’s visible surface geology into two distinct hemispheres: a relatively smooth region in the south, and a more rugged terrain in the north26.
The four smaller moons of Pluto were discovered using HST: Nix and Hydra in 200559, Kerberos in 201160 and Styx in 201261. After their discovery, it was possible using adaptive optics from the 10 m Keck Observatory to image Nix and Hydra, the larger of Pluto’s small moons. This is remarkable given their distance from Earth (~33 au) and small size (50 km × 35 km × 33 km for Nix and 65 km × 45 km × 25 km for Hydra62). New Horizons images of all four smaller moons reveal elongated shapes and brighter-than-expected reflectances. Infrared spectra taken by New Horizons confirmed the interpretation that the bright surfaces of these objects are due primarily to water ice. The surfaces of Nix and Hydra exhibit features that are consistent with impact craters, with 11 craters identified on Nix62.
Pluto and Charon emerge as geological worlds
Several months before the July 2015 flyby, New Horizons began to obtain imagery better than the 1994 HST Faint Object Camera images, and each following day revealed new details. With weeks to go before closest approach, it was already clear that Pluto had a more complex geology than previously predicted.
Because Pluto and Charon rotate synchronously about a common centre of gravity every 6.4 Earth days, New Horizons imagery in the last week before closest approach provided a map of all the illuminated surfaces of both bodies (Fig. 2). With some exceptions, terrain below 38° S latitude was not viewable by New Horizons because it was unilluminated as a result of the northern summer season. From the New Horizons data, we know that Pluto’s range of surface reflectance is extreme, varying in the range of 0.08–1.0, whereas Charon exhibits less variability across its surface15.
Much of Pluto’s encounter hemisphere is moderately to heavily cratered, and is thus likely to have its origins early in the Solar System’s history16. The size distribution of impact craters on both Pluto and Charon, when translated into impactor size, is roughly consistent with the observed size distribution of equivalent-sized Kuiper Belt objects, but reveals a surprising lack of impactors smaller than 1–2 km in diameter. The ancient terrain is cut by a sparse network of extensional faults, likely due to global expansion resulting from partial freezing of an internal ocean16, or phase changes in the ice shell17. However, many areas of Pluto are much younger. Large regions, particularly the north polar region, are partially covered in deposits of material — several kilometres thick — that have been locally eroded to reveal the older terrain beneath18. The composition, origin and removal method of these deposits remains a mystery. Surface textures in some areas, which resemble ‘penitentes’ formed by sublimation of terrestrial snowfields, but on a scale 1,000 times larger, suggest that sublimation, perhaps of methane, has been an important erosional process16.
The dominant geological feature on Pluto’s encounter hemisphere is a large 750 km × 1,400 km depression, 3–4 km deep, called Sputnik Planitia (SP)16,19. (Note that all names of surface features on Pluto and its moons are currently informal.) The floor of SP, which is extremely flat, is composed of bright N2, CH4 and CO ices (details below). From the lack of resolvable impact craters, this surface is geologically young (<10 Myr ago)16. This ice surface is clearly mobile: there are obvious flow features near the margins, and much of SP is occupied by polygonal features that appear to be active convection cells16,20,21. Computational models that assume the rheological properties of solid N2 and only present-day radiogenic heating predict resurfacing times of ~0.5 Myr, which is in agreement with the lack of impact craters. It is likely that SP occupies an ancient impact basin16,19, although it is also possible that the weight of the ice itself created the depression22. The location of SP, close to the anti-Charon point on Pluto (Fig. 2), is consistent with it being a positive mass anomaly that would be stable at this location due to tidal torques from Charon. This mass concentration might be explained if the thinner crust expected under a large impact basin allowed a denser liquid-water ocean underlying Pluto’s ice shell to reach closer to the surface at this location19,23 (Fig. 3), thus providing indirect evidence for such an ocean. New Horizons could not detect gravitational anomalies because the mass signature was small at the flyby distance.
The highland terrains to the north and east of SP feature extensive erosional features such as dendritic valley networks, which probably result from former glacial activity. There are also currently active glaciers, likely also dominated by N2, which flow down into SP from the east18,20. The western margin of SP features very different landforms: chaotic regions of jumbled mountain-sized water ice blocks reaching 5 km in height16. Because water ice is less dense than the nitrogen-dominated ices of SP that surround many of these mountains, they may be literally floating like icebergs in SP, and may have been transported to their current locations by the movement of the mobile nitrogen ice. To the southwest of SP are two large mountains with central depressions (informally called Wright Mons and Piccard Mons), which may be cryovolcanic constructs16. These are also lightly cratered, and thus geologically recent.
Pluto therefore displays a remarkable variety of geological processes, some of which continue to be active. Such complex and youthful activity was unexpected because Pluto experiences none of the tidal heating thought to power geological activity on many ice satellites. The most likely explanation is that Pluto’s feeble supply of radiogenic heat is sufficient to power geological activity owing to the presence of volatile ices in its interior. Nevertheless, Pluto’s surface is in general much older than that of Triton, Neptune’s largest moon24, which is similar in size, density and surface composition to Pluto. Pluto’s more rugged topography also suggests its ice shell is thicker and colder than that of Triton. This difference may provide evidence that Triton has an additional, probably tidal, heat source25.
Charon also surprised us with the discovery of its red polar region, which had not been predicted by ground-based observations, and its circumequatorial tectonic belt, which separates distinctive northern (more rugged, cratered) and southern (smoother) plains26. Elevation extremes appear to be ~10 km and are made possible by the strength of Charon’s dominant surface component (water ice) combined with its modest surface gravity. Charon’s crater counts indicate its surface overall is quite ancient (~4 Gyr ago)27.
Charon exhibits intriguing geological features not seen on Pluto, such as sunken mountains28 and landslides. The sunken mountains could be the result of flexure in Charon’s lithosphere or the flow of viscous cryovolcanic materials.
The many extensional features (for example, chasms, faults and rifts) of Charon suggest it was once slightly smaller and may have undergone a transition that increased its radius. The freezing of an internal ocean could have driven this expansion and would be consistent with its nearly spherical shape23. However, models of the freezing and expansion29 do not explain the north–south dichotomy in topography.
Compositional complexity in the Pluto system
Before the New Horizons flyby, Earth-based investigations of the composition of Pluto and Charon had revealed only limited information about the spatial resolution of compositional units. The surface of Pluto was known to have volatile ices such as nitrogen (N2), methane (CH4), carbon monoxide (CO) and ethane (C2H6) (ref. 30). Intriguingly, the CO was located primarily on the anti-Charon face of Pluto, and this was one factor in choosing the anti-Charon face of Pluto for the encounter hemisphere at closest approach14. At the surface temperature and pressure of Pluto, N2, CH4 and CO are volatile and can be transported via sublimation and condensation. The main driver of the atmospheric pressure on Pluto is the temperature of the most volatile ice (N2), which changes seasonally with insolation. At Pluto’s surface temperatures (~40 K), pure CH4 ice does not coexist with pure N2 ice; instead they form solid solutions31. It is also likely that Pluto’s CO is in solid solution with its other volatile ices.
Composition maps of Pluto’s surface have been produced from New Horizons data using three different methods: (1) equivalent widths of selected features32; (2) modelling of the bidirectional reflectance33; and (3) principal component analysis34. All three methods yield the same broad results: SP is a vast reservoir of volatile ices (N2, CH4 and CO), and outside SP there is a general latitudinal dependence of tholin composition (red material resulting from the irradiation of hydrocarbons)35 near the equator, N2 ice at mid-latitudes, and CH4 ice north of ~60° (Fig. 4).
A number of models have demonstrated that volatile ices are stable in SP over long timescales. Modelling the insolation on surface elements of Pluto found that the high albedo of ices in SP causes cold-trapping of the volatile ices over timescales of millions of years36. Additionally, global circulation models involving the evolution of volatile ices (over timescales of thousands of years) show that these ices accumulate in the basin used to model SP37.
The latitudinal dependence of Pluto’s composition (excluding SP, which is explained above) can be understood largely by the distribution of insolation over Pluto’s surface (Fig. 5). Near the equator (Box 2) there is always some insolation over each Pluto day, resulting in long-term loss of volatile ices and thus leaving behind the dark equatorial band of non-volatile tholins. Pluto’s north polar region has been experiencing summer with continuous insolation from 1987 to the current day. This intense insolation has caused sublimation of the most volatile of Pluto’s ices (N2) at the north pole, and atmospheric transport has redistributed the N2 further south34. This can be seen in Fig. 4, where CH4 is the most abundant species in the north.
While water ice was expected on Pluto’s surface, it had not been identified in Earth-based spectroscopic observations38. Observations at higher spatial resolution by New Horizons allowed the identification and mapping of water ice on Pluto’s surface for the first time. The signature of water ice is evident at mid- and equatorial-latitudes where volatile ices are depleted. The surface of Pluto’s north pole is low in water ice and is instead dominated by methane ice, which sublimates and condenses through seasonal cycles of insolation to cover the non-volatile water ice. There are a number of mountain ranges at the western edge of SP. The surface composition of these mountains is more than 30% water ice. From the shape, location and composition of these mountains, they appear to be blocks of water ice crust that have broken off and travelled from the edge of SP. Because water ice is less dense than N2 ice, it is possible that these mountains are floating on the glacial ices of SP.
In contrast, water ice is ubiquitous on Charon. Spectral features of water ice include the 1.65 µm band, which indicates that the water ice is in a crystalline form. Additionally, multiple sites exhibit an absorption feature at 2.1 µm, which is associated with ammonia hydrate. Local concentrations of ammonia hydrate are associated with some craters. Given that the destruction of ammonia ice by ultraviolet radiation takes ~10 Myr, these features are relatively young.
Unexpectedly, colour observations by New Horizons revealed that Charon’s north pole is dark and red. Analysis of Charon’s south pole (illuminated by sunlight reflected off Pluto) showed it was dark compared with the equatorial and mid-latitude regions, just like Charon’s north pole. It is proposed that the red colouration is a result of tholins that accumulate each season39. CH4 escapes Pluto’s atmosphere at a rate of 5 × 1025 molecules s−1 (ref. 40) and ~2.5% of this will intercept Charon. These CH4 molecules will bounce around Charon’s surface, where some will be cold-trapped at the winter poles. Irradiation by Lyα and cosmic rays over the long polar winter will convert the captured CH4 to non-volatile, red tholin-like material, producing about 0.16 mm per million Earth years.
Pluto’s cold and hazy atmosphere
In 1976, following the detection of methane ice on Pluto’s surface, it was suggested that Pluto might have an atmosphere41. The first definitive observation of Pluto’s atmosphere was from a stellar occultation42 in 1988. As Pluto moved in front of the Galactic Plane as seen from the Earth in the 2000s, more occultation observations were made, with observations nearly every year over the period 2002–2016. Analysis of these observations demonstrated that Pluto’s atmospheric pressure increased by a factor of two from 1988 to 2002 as Pluto receded from perihelion43,44. This may seem paradoxical because one might expect Pluto’s atmosphere to be most dense at perihelion, where sunlight is the most intense, but it is explained by a high thermal inertia in Pluto’s subsurface45 that retains heat from perihelion. Later occultations showed that Pluto’s atmospheric pressure has been continuing to increase with a tripling of atmospheric pressure from 1988 to 201546. Ground-based spectroscopic observations provided the first direct detection of atmospheric gases with the detection of methane47. Later work detected CO and hydrogen cyanide48 in Pluto’s atmosphere.
New Horizons provided unique information about Pluto’s atmosphere through ultraviolet solar and radio science occultations and high-phase-angle imaging. One surprising discovery was the blue haze layers encircling Pluto40, which were observed after closest approach thanks to the strong forward-scattering properties of the haze particles. The haze extends to an altitude of >200 km with a scale height of 50 km (ref. 40). The scattering and extinction properties of the haze suggest that the haze particles are fractal aggregates40,49.
The composition of Pluto’s atmosphere was measured by observing sunlight during a solar occultation. Different molecules absorbed the light at specific ultraviolet wavelengths in the range of 52–187 nm. This experiment provided the surprising result that the temperature in the upper atmosphere is much colder (69 K) than previously expected (~100 K; ref. 40). The cooling mechanism is currently an open question. The observation of solar occultation also provided definitive detection of acetylene (C2H2), ethylene (C2H4) and ethane (C2H6), in addition to the expected N2 (dominant species) and CH4 (surface mixing ratio ~0.3%)50. The solar occultation data show a decrease in concentration of C2 hydrocarbons below 200–400 km altitude, which is attributed to condensation51.
Measurements of Pluto’s surface pressure and temperature were made using an uplink radio science experiment, using four Deep Space Network stations to transmit a combination of right- and left-circularly polarized radiation through Pluto’s atmosphere to New Horizons52. The radio science experiment was sensitive to the atmosphere in approximately the lowest 100 km altitude. Above ~20 km altitude, both the ingress and egress atmospheric profiles are consistent with each other and with stellar occultation observations53. Below ~20 km altitude, the ingress and egress temperature profiles differ significantly: the atmosphere is much warmer over the egress location (a dark area on the surface) than the ingress location (near the edge of SP)53. This is also reflected in the temperature gradients in the lowest 10 km: 6.4 K km–1 at ingress and 3.4 K km–1 at egress53. The mean surface pressure is 11.5 ± 0.7 µbar at a reference radius of 1,189.9 ± 0.2 km (ref. 53).
Pluto’s atmosphere should be considered in the context of other outer planet atmospheres, such as Triton and Titan. The main constituent in each of these three atmospheres is N2, with methane as a minor constituent. The photochemistry in each of these atmospheres is expected to be similar, with ultraviolet radiation of methane producing hydrocarbons and ultraviolet radiation of N2 producing ions, leading to haze40. The scattering properties of Pluto’s haze particles are consistent with fractal aggregates, as on Titan49. All three atmospheres have haze layers, with the least haze in Triton’s atmosphere and the most opaque haze in Titan’s. Triton and Pluto have tenuous atmospheres, with atmospheric pressures of the order of 10 µbar (Triton’s atmospheric pressure was determined to be 14 µbar by Voyager 2 in 198954), while Titan’s is substantially thicker, with a pressure of ~1.5 bar. For comparison, the atmospheric pressure of 14 µbar occurs at ~400 km altitude on Titan. The atmospheric pressure of both Triton and Pluto has been observed to vary significantly with time.
New Horizons searched for an atmosphere around Charon, but no evidence of an atmosphere was found55.
Box 2 | Seasons and insolation
Pluto experiences complex seasonal variation across a number of different timescales. Owing to Pluto’s eccentric orbit, insolation at aphelion (49 au) is 2.6 times lower than that at perihelion (29 au) over the course of its 248 year orbit. Due to Pluto’s large obliquity (~120°), its polar regions experience extended periods of arctic night. Pluto’s obliquity is not constant: it precesses with a peak-to-peak amplitude of 23° over ~3 Myr. When considering these effects on Pluto and Charon, latitudes between 13° N and 13° S always receive sunlight during a diurnal cycle (6.4 days), latitudes poleward of 38° experience arctic seasons every Pluto year, and latitudes poleward of 77° never have the sun directly overhead63. These longitudinal regions show a correlation with albedo and composition (Fig. 5).
On even longer timescales (~3.7 Myr), Pluto’s orbital longitude of perihelion regresses. The combination of all these affects results in epochs of extreme seasons36. An example of this was 0.9 Myr ago, when Pluto’s subsolar point at perihelion was at high northern latitude, resulting in a short but intense artic summer36.
Models of atmospheric pressure, given the obliquity precession and orbital variations, predict possible high pressures (as much as hundreds of millibars, depending on assumptions of polar albedo), which could explain geomorphic features that suggest possible previous liquids on the surface of Pluto, like the dendritic terrain and lake-like features near the edge of SP64.
The historic July 2015 flyby provided new insights into the Pluto system. We are inspired to ask new questions, thanks to the discovery of glaciers composed of N2, CH4 and CO flowing on Pluto, mountains of water ice, a global blue atmosphere with haze and evidence suggestive of a subsurface ocean. More laboratory studies are needed to better understand the Pluto system and other cold distant Kuiper Belt objects. Experiments to better understand the rheology of Pluto’s volatile ices (such as flow rates, stain and stress parameters) are needed. Furthermore, measurements of the visible and near-infrared optical constants of ice mixtures such as N2, CH4 and CO are needed to help improve reflectance models. Additionally, ultraviolet optical constants are needed at low temperatures for the C2 hydrocarbons seen in Pluto’s atmosphere.
Pluto and its moons have now returned to the realm of astronomy. Pluto is only a few pixels across when imaged through the soon-to-be launched James Webb Space Telescope (JWST), yet JWST’s unprecedented spectral sensitivity and factor of approximately six improvement in spectral resolution compared with New Horizons will allow for the detection of new species and regular measurement of Pluto’s surface ices. NASA’s airborne Stratospheric Observatory for Infrared Astronomy can be utilized in combination with JWST and ground-based telescopes to monitor Pluto’s evolving atmosphere via occultation events. Continued monitoring of Pluto’s hemispherical surface composition with ground-based near-infrared telescopes and the Atacama Large Millimeter/submillimeter Array observatory, which can also probe its atmosphere down to altitudes of several hundred kilometres, are all important parts in this next chapter of understanding the Pluto system.
Publisher’s note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
This work was supported by the NASA New Horizons project. We thank the engineers and staff of the New Horizons team whose dedication enabled the initial reconnaissance of the Pluto system.