Planet Nine

Orbit

Planet Nine is hypothesized to follow a highly elliptical orbit around the Sun lasting 10,000–20,000 years.^[24] The planet's semi-major axis is estimated to be 7014104718509490000♠700 AU,^{[upper-alpha 2]} roughly 20 times the distance from Neptune to the Sun, and its inclination to be about 6999523598775598300♠30°±10°.^{[2][3][26][upper-alpha 3]} The high eccentricity of Planet Nine's orbit could bring it as close as 7013299195741400000♠200 AU at its perihelion and take it as far away as 7014179517444840000♠1,200 AU at its aphelion.^[27][28]

The aphelion, or farthest point from the Sun, would be in the general direction of the constellation of Taurus,^[29] whereas the perihelion, the nearest point to the Sun, would be in the general direction of the southerly areas of Serpens (Caput), Ophiuchus, and Libra.^[30][31]

Brown thinks that if Planet Nine is confirmed to exist, a probe could fly by it in as little as 20 years, with a powered slingshot around the Sun.^[32]

Size and composition

Planet Nine is hypothesized to be two to four times the diameter of Earth;^[2][6] similar to the ice giants Uranus and Neptune.^[33]

The planet is estimated to have 10 times the mass^[21][26] and two to four times the diameter of Earth.^[34][6][35] An object with the same diameter as Neptune has not been excluded by previous surveys in visible light.^[2] However, the infrared survey by the Wide-field Infrared Survey Explorer (WISE) may have the capabilities to detect such a planet.^[36][37][8]

Brown thinks that if Planet Nine exists, its mass is large enough to clear its feeding zone in 4.6 billion years (with possible exceptions for some combinations of semi-major axis and mass) and that its gravity dominates the outer edge of the Solar System, which is sufficient to make it a planet by current definitions.^[38] Jean-Luc Margot has also stated that Planet Nine satisfies his criteria and would qualify as a planet—if and when it is detected.^[39][40]

Brown speculates that the predicted planet is most probably an ejected ice giant, similar in composition to Uranus and Neptune: a mixture of rock and ice with a small envelope of gas.^[2][6] In fact, if it once orbited the region of the gas/ice giants, the planet probably acquired an atmosphere of hydrogen and helium.^[41] Based on its size, the considerable gravitational pull of Planet Nine could theoretically promote life in subsurface oceans of its moons, were it to have any. Subsurface oceans have been discovered on Europa of Jupiter, Enceladus of Saturn, and subsurface water is postulated for Triton of Neptune.^[42]

Observations: Orbital clustering amongst high perihelion objects

The clustering of the orbits of extreme trans-Neptunian objects was first described by Chad Trujillo and Scott S. Sheppard, who noted similarities between the orbits of Sedna and 2012 VP₁₁₃. Upon further analysis they observed that the arguments of perihelion (which indicate the orientation of elliptical orbits within their orbital planes) of 12 eTNOs with perihelia greater than 7012448793612100000♠30 AU and semi-major axes greater than 7013224396806050000♠150 AU were clustered near zero degrees. Trujillo and Sheppard proposed that this alignment was caused by a massive unknown planet beyond Neptune via the Kozai mechanism.^[4] (see Trujillo and Sheppard (2014) for more details.)

Caltech's Konstantin Batygin and Michael E. Brown, looking to refute the mechanism proposed by Trujillo and Sheppard, also examined the orbits of the extreme trans-Neptunian objects.^[1] After eliminating the objects in Trujillo and Sheppard's original analysis that were unstable due to close approaches to Neptune or were affected by Neptune mean-motion resonances, they determined that the arguments of perihelion for the remaining six objects (namely Sedna, 2012 VP113, 2004 VN₁₁₂, 2010 GB174, 2000 CR₁₀₅, and 2010 VZ₉₈) were clustered around 7000555014702134197♠318°±8°. This was out of alignment with how the Kozai mechanism would align these orbits, at c. 0° or 180°.^{[1][upper-alpha 4]}

Orbital correlations among six distant trans-Neptunian objects led to the hypothesis. (See: Final frame orbits)

Batygin and Brown also found that the orbits of the six objects with semi-major axes greater than 250 AU and perihelia beyond 30 AU (namely Sedna, 2012 VP113, 2004 VN₁₁₂, 2010 GB174, 2007 TG422, and 2013 RF98) were aligned in space with their perihelia in roughly the same direction, resulting in a clustering of their longitudes of perihelion. The orbits of the six objects were also tilted with respect to that of the ecliptic and approximately co-planar, producing a clustering of their longitudes of ascending nodes. They determined that there was only a 0.007% likelihood that this combination of alignments was due to chance.^{[1][20][46][47]} These six objects had been discovered by six different surveys on six different telescopes. That made it less likely that the clumping might be due to an observation bias such as pointing a telescope at a particular part of the sky. The observed clustering should be smeared out by the object's varied precession rates in a few hundred million years.^{[upper-alpha 5]} This indicates that it couldn't be due to an event in the distant past, like a passing star, and is most likely being maintained by an object orbiting the Sun.^[1]

In a later article Trujillo and Sheppard noted a correlation between the longitude of perihelion and the argument of perihelion of the eTNOs with semi-major axes greater than 150 AU. Those with a longitude of perihelion of 0–120° have arguments of perihelion between 280–360°, and those with longitude of perihelion of 180–340° have argument of perihelion 0–40°. The statistical significance of this correlation was 99.99%. They suggested that the correlation is due to the orbits of these objects avoiding close approaches to a massive planet by passing above or below its orbit.^[48]

Extreme trans-Neptunian objects

Among the extreme trans-Neptunian objects are two high-perihelion objects: Sedna and 2012 VP113. Sedna and 2012 VP113 are distant detached objects with perihelia greater than 70 AU. Their high perihelia keep them at a sufficient distance to avoid significant gravitational perturbations from Neptune. Previous explanations for the high perihelion of Sedna include a close encounter with an unknown planet on a distant orbit and a distant encounter with a random star or a member of the Sun's birth cluster that passed near the Solar System.^[49][50][51]

Since early 2016, ten more extreme trans-Neptunian objects have been discovered with orbits that have a perihelion greater than 30 AU and a semi-major axis greater than 250 AU bringing the total to sixteen. Most eTNOs have perihelia significantly beyond Neptune, which orbits 7012448793612100000♠30 AU from the Sun.^[21][52] Generally, TNOs with perihelia smaller than 7012538552334520000♠36 AU experience strong encounters with Neptune.^[1] Most of the eTNOs are relatively small, but currently relatively bright because they are near their closest distance to the Sun in their elliptical orbits. These are also included in the orbital diagrams and tables below.

The extreme trans-Neptunian object orbits

6 original and 8 TNO object orbits with current positions near their perihelion in purple, with hypothetical Planet Nine orbit in green

Closer up view of 13 TNO current positions

Extreme trans-Neptunian objects with perihelion greater than 7012448793612100000♠30 AU and a semi-major axis greater than 7013373994676750000♠250 AU^[53]

Object	Orbit							Orbital plane			Body
	Barycentric[upper-alpha 6] Orbital period (years)	Barycentric Semimajor axis (AU)	Perihelion (AU)	Barycentric Aphelion (AU)	Current distance from Sun (AU)	Eccent.	Argum. peri ω (°)	inclin. i (°)	Longitude of		Hv	Current mag.	Diameter (km)
									Ascending node ☊ or Ω (°)	Perihelion ϖ=ω+Ω (°)
Sedna	11,400	507	76.04	936	85.5	0.85	311.5	11.9	144.5	96.0	1.5	20.9	1,000
2004 VN112	5,900	327	47.32	607	47.7	0.85	327.1	25.6	66.0	33.1	6.5	23.3	200
2007 TG422	11,300	503	35.57	970	37.3	0.93	285.7	18.6	112.9	38.6	6.2	22.0	200
2010 GB174	6,600	351	48.76	654	71.2	0.87	347.8	21.5	130.6	118.4	6.5	25.1	200
2012 VP113	4,300	266	80.27	441	83.5	0.69	292.8	24.1	90.8	23.6	4.0	23.3	600
2013 RF98	6,900	364	36.10	690	36.8	0.90	311.8	29.6	67.6	19.4	8.7	24.4	70
2013 FT28	5,050	295	43.60	546	57.0	0.86	40.2	17.3	217.8	258.0 (*)	6.7	24.4	200
2014 FE72	66,000	1,600	36.31	3,200	61.5	0.98	134.4	20.6	336.8	111.2	6.1	24.0	200
2014 SR349	5,160	299	47.57	549	56.3	0.84	341.4	18.0	34.8	16.2	6.6	24.2	200
2013 SY99	19,700	730	49.91	1,410	60.3	0.93	32.4	4.2	29.5	61.9	6.7	24.5	250
2015 GT50	5,510	310	38.45	580	41.7	0.89	129.2	8.8	46.1	175.3 (*)	8.5	24.9	80
2015 KG163	17,730	680	40.51	1,320	40.8	0.95	32.0	14.0	219.1	251.1 (*)	8.1	24.3	100
2015 RX245	8,920	430	45.48	815	61.4	0.89	65.4	12.2	8.6	74.0	6.2	24.2	250
uo5m93[56]	4,760	283	39.48	526	41.7	0.86	43.3	6.8	165.9	209.3 (*)	8.9	25.0	?
2015 BP519 "Caju"[57]	9,500	449	35.25	863	52.7	0.92	348.1	54.1	135.2	123.3	4.3	21.5	550[58]
2015 TG387	40,000	1,200	64.94	2,300	77.7	0.94	118.2	11.7	300.8	59.0	5.3	24.3	300
Ideal elements under hypothesis	—	>250	>30	—	—	>0.5	—	10~30	—	2~120	—	—	—
Hypothesized Planet Nine	~15,000	~700	~200	~1,200	~1,000?	~0.6	~150	~30	7001910000000000000♠91±15	7002241000000000000♠241±15		>22	~40,000

• (*) longitude of perihelion, ϖ, outside expected range;
•    are the objects included in the original study by Trujillo and Sheppard (2014).^[4]
•    has been added in the 2016 study by Brown and Batygin.^[1][2]
• All other objects have been announced later.

The most extreme case is that of 2015 BP₅₁₉, nicknamed Caju, which has both the highest inclination^[59] and the farthest nodal distance; these properties make it a probable outlier within this population.^[60]

Simulations: Observed clustering reproduced, inference on Planet Nine orbit and mass

The clustering of the orbits of extreme trans-Neptunian objects and raising of their perihelia is reproduced in simulations that include Planet Nine. In simulations conducted by Batygin and Brown swarms of large semi-major axis scattered disk objects^{[upper-alpha 7]} that began with random orientations were sculpted into roughly collinear groups of spatially confined orbits by a massive distant planet in a highly eccentric orbit. These surviving objects, about 10% of the population with initial semi-major axes of 250–550 AU,^[61] were in orbits that were oriented with their long axes anti-aligned with respect to the massive planet and were roughly co-planar with it. The objects were also found to be in resonance with the massive planet. The resonances included high-order resonances, for example 27:17, and were interconnected, yielding an orbital evolution that was fundamentally chaotic, causing their semi-major axes to vary unpredictably on million-year timescales.^[1] The perihelia of these objects were also raised temporarily, producing Sedna-like orbits, before being returned to orbits more typical of typical trans-Neptunian objects after several hundred million years.^[62] Some of the eTNOs also evolved into orbits perpendicular to the plane of the Solar System, which Batygin and Brown later discovered had also been observed.^[1] Similar results also occur if simulations began with the objects initially interacting with multiple planets, like in various versions of the Nice model. Lawler et al. and Nesvorny et al. both found that objects scattered outward by the giant planets would be captured in a cloud centered on Planet Nine's semi-major axis with most objects at greater semi-major axis. A significant fraction of these objects had inclinations of greater than 60°, and for an initial 20 Earth mass planetesimal disk roughly 0.3–0.4 Earth masses remained in the Planet Nine cloud at the end of a 4 billion year simulation.^[63][44]

Batygin and Brown found that the distribution of the orbits of known extreme trans-Neptunian objects is best reproduced in simulations using a 10 M_⊕^{[upper-alpha 8]} planet in the following orbit:

• semi-major axis a ≈ 7014104718509490000♠700 AU (orbital period 700^1.5=18,520 years)^{[upper-alpha 9]}
• eccentricity e ≈ 0.6, (perihelion ≈ 7013418874037960000♠280 AU, aphelion ≈ 7014167549615184000♠1,120 AU)
• inclination i ≈ 30° to the ecliptic
• longitude of the ascending node Ω ≈ 7000164060949687467♠94°.^{[upper-alpha 10]}
• argument of perihelion ω ≈ 139° and longitude of perihelion ϖ = 7000410152374218668♠235°±12°^[64]

This orbit results in strong anti-alignment beyond 250 AU, weak alignment between 150 AU and 250 AU, and little effect inside 150 AU. Anti-alignment occurs with variable success using a semi-major axis between 7013598391482800000♠400 AU and 7014224396806050000♠1500 AU and an eccentricity between 0.5 and 0.8. Anti-alignment weakens as Planet Nine's inclination is increased.^[25] Simulations conducted by Becker et al. found a similar range for the stability of eTNOs, semi-major axes ranging from 500 to 1200 AU and eccentricities ranging from 0.3 to 0.6 with lower eccentricities being favored at smaller semi-major axes. They noted that while stability was favored with smaller eccentricities, anti-alignment was more likely at higher eccentricities near the borders of stability.^[65] Lawler et al. found that the population captured was smaller in simulations with a planet in a circular orbit which also produced few high inclination objects.^[63] Investigations by Cáceres et al. showed that a planet with a lower perihelion led to a narrower confinement of orbits of the eTNOs, with a perihelion of 90 AU or higher being consistent with the distribution of the classical Kuiper belt objects.^[66]

Dynamics: Effect on other objects in the Solar System

Planet Nine modifies the orbits of extreme trans-Neptunian objects via a combination of effects. On very long timescales exchanges of angular momentum with Planet Nine causes the perihelia of anti-aligned objects to rise until their precession reverses direction, maintaining their anti-alignment, and later fall, returning them to their original orbits. On shorter timescales mean-motion resonances with Planet Nine provides phase protection, which stabilizes their orbits by slightly altering the objects' semi-major axes, keeping their orbits synchronized with Planet Nine's and preventing close approaches. The inclination of Planet Nine's orbit weakens this protection, resulting in a chaotic variation of semi-major axes as objects hop between resonances. The orbital poles of the objects circle that of the Solar System's Laplace plane, which at large semi-major axes is warped toward the plane of Planet Nine's orbit, causing their poles to be clustered toward one side.^[61]

Apsidal anti-alignment

Secular evolution of eTNOs induced by Planet Nine for objects with semi-major axis of 250 AU.^[67][61] Blue:anti-aligned, Red:aligned, Green:metastable, Orange:circulating. Crossing orbits above black line.^{[upper-alpha 11]}

The anti-alignment and the raising of the perihelia of extreme trans-Neptunian objects with semi-major axes greater than 250 AU is produced by the secular effects of Planet Nine. Secular effects act on timescales much longer than orbital periods so the perturbations two objects exert on each other are the average between all possible configurations. Effectively the interactions become like those between two wires of varying thickness, thicker where the objects spend more time, that are exerting torques on each other, causing exchanges of angular momentum but not energy. Thus secular effects can alter the eccentricities, inclinations and orientations of orbits but not the semi-major axes.^[68][69]

Exchanges of angular momentum with Planet Nine cause the perihelia of the anti-aligned objects to rise and fall while their longitudes of perihelion librate, or oscillate within a limited range of values. When the angle between an anti-aligned object's perihelion and Planet Nine's (delta longitude of perihelion on diagram) climbs beyond 180° Planet Nine exerts a positive average torque on the object's orbit. This torque increases the object's angular momentum,^{[upper-alpha 12]} causing the eccentricity of its orbit to decline (see blue curves on diagram) and its perihelion to rise away from Neptune's orbit. The object's precession then slows and eventually reverses as its eccentricity declines. After delta longitude of perihelion drops below 180° the object begins to feel a negative average torque and its eccentricity grows and perihelion falls. When the object's eccentricity is once again large it precesses forward, returning the object to its original orbit after several hundred million years.^[61][68][70]

The behavior of the orbits of other objects varies with their initial orbits. Stable orbits exist for aligned objects with small eccentricities. Although objects in these orbits have high perihelia and have yet to be observed, they may have been captured at the same time as Planet Nine due to perturbations from a passing star.^[71][72] Aligned objects with lower perihelia are only temporarily stable, their orbits precess until parts of the orbits are tangent to that of Planet Nine, leading to frequent close encounters.^[69][61] After crossing this region the perihelia of their orbits decline causing them to encounter the other planets leading to their ejection.^{[72][upper-alpha 13]}

The curves the orbits follow vary with semi-major axis of the object and if the object is in resonance. At smaller semi-major axes the aligned and anti-aligned regions shrink and eventually disappear below 150 AU, leaving typical Kuiper belt objects unaffected by Planet Nine. At larger semi-major axis the region with stable aligned orbits shifts to lower eccentricities and becomes narrower.^[67] The opposite happens for anti-aligned objects, shifting these regions to higher eccentricities and lower perihelia, with perihelia of 40 AU becoming stable for semi-major axes greater than 1000 AU.^[73][74] The anti-alignment of resonant objects, for example if Sedna is in a 3:2 resonance with Planet Nine as proposed by Malhotra, Volk and Wang,^[75] is maintained by similar secular effects inside the mean-motion resonances.^[61][67] The secular dynamics is more complex if Planet Nine and the eTNOs are in inclined orbits, alignment and anti-alignment in this case is more the result of sticky chaos rather than confinement, with orbits evolving widely but spending more time in regions of relative stability associated with secular resonances.^[73]

Evolution and long-term stability of anti-aligned orbits

An example of phase-protection in a mean-motion resonance: The orbital resonances of Orcus and Pluto in a rotating frame with a period equal to Neptune's orbital period. (Neptune is held stationary.)

The long term stability of anti-aligned extreme trans-Neptunian objects with orbits that intersect that of Planet Nine is due to their being captured in mean-motion resonances. Objects in mean-motion resonances with a massive planet are phase protected, preventing them from making close approaches to the planet. When the orbit of a resonant object drifts out of phase,^{[upper-alpha 14]} causing it to make closer approaches to a massive planet, the gravity of the planet modifies its orbit, altering its semi-major axis in the direction that reverses the drift. This process repeats as the drift continues in the other direction causing the orbit to appear to rock back and forth, or librate, about a stable center when viewed in a rotating frame of reference.^[76][61] In the example at right, when the orbit of a plutino drifts backward it loses angular momentum when it makes closer approaches ahead of Neptune,^{[upper-alpha 15]} causing its semi-major axis and period to shrink, reversing the drift.^[77]

In a simplified model where all objects orbit in the same plane and the giant planets are represented by rings,^{[upper-alpha 16]} objects captured in strong resonances with Planet Nine could remain in them for the lifetime of the Solar System. At large semi-major axes, beyond a 3:1 resonance with Planet Nine, most of these objects would be in anti-aligned orbits. At smaller semi-major axes the longitudes of perihelia of an increasing number of objects could circulate, passing through all values ranging from 0° to 360°, without being ejected,^{[upper-alpha 17]} reducing the fraction of objects that are anti-aligned.^[61][69] 2015 GT50 may be in one of these circulating orbits.^[78]

If this model is modified with Planet Nine and the eTNOs in inclined orbits the objects alternate between extended periods in stable resonances and periods of chaotic diffusion of their semi-major axes. The distance of the closest approaches varies with the inclinations and orientations of the orbits, in some cases weakening the phase protection and allowing close encounters. The close encounters can then alter the eTNO's orbit, producing stochastic jumps in its semi-major axis as it hops between resonances, including higher order resonances. This results in a chaotic diffusion of an object's semi-major axis until it is captured in a new stable resonance and the secular effects of Planet Nine shift its orbit to a more stable region.^[61][69] The chaotic diffusion reduces the range of longitudes of perihelion that anti-aligned objects can reach while remaining in stable orbits.^[72]

Neptune's gravity can also drive a chaotic diffusion of semi-major axes when all objects are in the same plane.^[1] Distant encounters with Neptune can alter the orbits of the eTNOs, causing their semi-major axes to vary significantly on million year timescales.^[48] These perturbations can cause the semi-major axes of the anti-aligned objects to diffuse chaotically while occasionally sticking in resonances with Planet Nine. At semi-major axes larger than Planet Nine's, where the objects spend more time, anti-alignment may be due to the secular effects outside mean-motion resonances.^[74]

The phase protection of Planet Nine's resonances stabilizes the orbits of objects that interact with Neptune, via its resonances, for example 2013 FT28, or by close encounters for objects with low perihelia like 2007 TG422 and 2013 RF98.^[48] Instead of being ejected following a series of encounters these objects can hop between resonances with Planet Nine and evolve into orbits no longer interacting with Neptune.^[65][79] A shift in the position of Planet Nine in simulations from the location favored by an analysis of Cassini data to a position near aphelion has been shown to increase the stability of some of the observed objects, possibly due to this shifting the phases of their orbits to a stable range.^[80][81]

Clustering of orbital poles (nodal alignment)

Tilting of Laplace Plane by Planet Nine

The clustering of the orbital poles, which produces an apparent clustering of the longitude of the ascending nodes and arguments of perihelion of the extreme TNOs, is the result of a warping of the Laplace plane of the Solar System toward that of Planet Nine's orbit. The Laplace plane defines the center around which the pole of an object's orbit precesses with time. At larger semi-major axes the angular momentum of Planet Nine causes the Laplace plane to be warped toward that of its orbit.^{[upper-alpha 18]} As a result, when the poles of the eTNO orbit precess around the Laplace plane's pole they tend to remain on one side of the ecliptic pole. For objects with small inclination relative to Planet Nine, which were found to be more stable in simulations, this off-center precession produces a libration of the longitudes of ascending nodes with respect to the ecliptic making them appear clustered.^[61] In simulations the precession is broken into short arcs by encounters with Planet Nine and the positions of the poles are clustered in an off-center elliptical region.^[82] In combination with the anti-alignment of the longitudes of perihelion this can also produce clustering of the arguments of perihelion.^[61]

Objects in perpendicular orbits with large semi-major axis

The orbits of the five objects with high-inclination orbits (nearly perpendicular to the ecliptic) are shown here as cyan ellipses with the hypothetical Planet Nine in orange. Those of four are towards the left in this view, and that of one (2012 DR30) is towards the right, with an aphelion over 7014299195741400000♠2,000 AU.

Planet Nine can deliver extreme trans-Neptunian objects into orbits roughly perpendicular to the plane of the Solar System.^[83][84] Several objects with high inclinations, greater than 50°, and large semi-major axes, above 250 AU, have been observed.^[85] Their high inclination orbits can be generated by a high order secular resonance with Planet Nine involving a linear combination of the orbit's arguments and longitudes of perihelion: Δϖ - 2ω. Low inclination eTNOs can enter this resonance after first reaching low eccentricity orbits. The resonance causes their eccentricities and inclinations to increase, delivering them into perpendicular orbits with low perihelia where they are more readily observed. The orbits then evolve into retrograde orbits with lower eccentricities after which they pass through a second phase of high eccentricity perpendicular orbits before returning to low eccentricity, low inclination orbits. Unlike the Kozai mechanism this resonance causes objects to reach their maximum eccentricities when in nearly perpendicular orbits. In simulations conducted by Batygin and Brown this evolution was relatively common, with 38% of stable objects undergoing it at least once.^[61] Saillenfest et al. also observed this behavior in their study of the secular dynamics of eTNOs and noted that it caused the perihelia to fall below 30 AU for objects with semi-major axes greater than 300 AU, and with Planet Nine in an inclined orbit it could occur for objects with semi-major axes as small as 150 AU.^[73] In simulations the arguments of perihelion of the objects with roughly perpendicular orbits and reaching low perihelia are clustered near or opposite Planet Nine's and their longitudes of ascending node are clustered around 90° in either direction from Planet Nine's.^[86][1] This is in rough agreement with observations with the differences attributed to distant encounters with the known giant planets.^[1] Seven high inclination objects with semi-major axes greater than 250 AU and perihelia beyond Jupiter's orbit are currently known:

High-inclination trans-Neptunian objects with a semi-major axis greater than 7013373994676750000♠250 AU^[1][61][87]

Object	Orbit						Body
	Perihelion (AU) Figure 9[1]	Semimaj. (AU) Figure 9[1]	Current distance from Sun (AU)	inc (°)[85]	Eccen.	Arg. peri ω (°)	Mag.	Diam. (km)
(336756) 2010 NV1	9.4	323	14	141	0.97	133	22	20–45
(418993) 2009 MS9	11.1	348	12	68	0.97	129	21	30–60
2010 BK118	6.3	484	11	144	0.99	179	21	20–50
2013 BL76	8.5	1,213	11	99	0.99	166	21.6	15–40
2012 DR30	14	1,404	17	78	0.99	195	19.6	185[88]
2014 LM28	16.8	268	17	85	0.94	38	22	46
2015 BP519	35.3	449	53	54	0.92	348	21.5	550

Temporary or coincidental nature of clustering

The results of the Outer Solar System Survey (OSSOS) suggest that the observed clustering is the result of a combination of observing bias and small number statistics. OSSOS, a well-characterized survey of the outer Solar System with known biases, observed eight trans-Neptunian objects with semi-major axis > 150 AU with orbits oriented on a wide range of directions. After accounting for the known observational biases of the survey, no evidence for the arguments of perihelion (ω) clustering identified by Trujillo and Sheppard was seen^{[upper-alpha 20]} and the orientation of the orbits of the objects with the largest semi-major axis was statistically consistent with random.^{[110][109][111]} A previously released article by Mike Brown analyzed the discovery locations of eccentric trans-Neptunian objects. While identifying some biases he found that even with these biases the clustering of longitudes of perihelion of the known objects would be observed only 1.2% of the time if their actual distribution was uniform.^[112]

Simulations of 15 known extreme trans-Neptunian objects evolving under the influence of Planet Nine also revealed a number of differences with observations. Cory Shankman et al. simulated clones (objects with similar orbits) of 15 objects with semi-major axis > 150 AU and perihelion > 30 AU under the influence of a 10 Earth-massed Planet Nine in Batygin and Brown's proposed orbit.^{[upper-alpha 21]} While longitude of perihelion alignment of the objects with semi-major axis > 250 AU was observed in their simulations, the alignment of the arguments of perihelion was not. The simulations also revealed an increase in the inclinations of many objects, thereby predicting a larger reservoir of high-inclination TNOs that had not yet been observed.^[70] A previously published article concluded that observations were insufficient to determine if this reservoir exists, however.^[63] The perihelia of the objects rose and fell smoothly, in contrast with the observed absence of extreme TNOs with perihelia between 50 AU and 70 AU. Their perihelia also reached values where the objects would not be observed, requiring a population of Sednas significantly larger than current estimates. They then, after declining, fell to values low enough for the objects to enter planet-crossing orbits leading to their ejection from the Solar System, requiring an initial population inconsistent with current models of the early Solar System to explain current observations.^{[upper-alpha 22]} Based on these challenges Shankman et al. concluded that the existence of Planet Nine is unlikely and that the currently observed alignment of the existing TNOs is a temporary phenomenon that will disappear as more objects are detected.^[113][70]

Inclination instability due to mass of undetected objects

Ann-Marie Madigan and Michael McCourt postulate that an inclination instability in a distant massive belt is responsible for the alignment of the arguments of perihelion of the eTNOs. The inclination instability occurs in a disk of particles in eccentric orbits around a massive object. The self-gravity of this disk causes its spontaneous organization, increasing the inclinations of the objects and aligning the arguments of perihelion, forming it into a cone above or below the original plane.^[114] This process requires an extended time and significant mass of the disk, on the order of a billion years for a 1–10 Earth-mass disk.^{[115][upper-alpha 23]} While an inclination instability can align the arguments of perihelion and raise perihelia, producing detached objects, it does not align the longitudes of perihelion.^[112] Mike Brown considers Planet Nine a more probable explanation, noting that current surveys do not support the existence of a scattered-disk region of sufficient mass to support this idea of "inclination instability".^[116][117] In Nice model simulations that included the self-gravity of the planetesimal disk an inclination instability did not occur due to a rapid precession of the objects' orbits and their being ejected on too short of a timescale.^[118]

Object in lower-eccentricity orbit

Renu Malhotra, Kathryn Volk, and Xianyu Wang have proposed that the four detached objects with the longest orbital periods, those with perihelia beyond 7012598391482800000♠40 AU and semi-major axes greater than 7013373994676750000♠250 AU, are in n:1 or n:2 mean-motion resonances with a hypothetical planet. Two other objects with semi-major axes greater than 7013224396806050000♠150 AU are also potentially in resonance with this planet. Their proposed planet could be on a lower eccentricity, low inclination orbit, with eccentricity e < 0.18 and inclination i ≈ 11°. The eccentricity is limited in this case by the requirement that close approaches of 2010 GB174 to the planet are avoided. If the eTNOs are in periodic orbits of the third kind,^{[upper-alpha 24]} with their stability enhanced by the libration of their arguments of perihelion, the planet could be in a higher inclination orbit, with i ≈ 48°. Unlike Batygin and Brown, Malhotra, Volk and Wang do not specify that most of the distant detached objects would have orbits anti-aligned with the massive planet.^[75][120]

Alignment due to the Kozai mechanism

Previous models with additional planets

Attempts to detect planets beyond Neptune by indirect means such as orbital perturbation date back beyond the discovery of Pluto. A few observations were directly related to the Planet Nine hypothesis:

• George Forbes was the first to postulate the existence of trans-Neptunian planets in 1880, and his work is considered similar to more recent Planet Nine theories. In Forbes's model the planet had a semi-major axis of ∼300 AU (roughly three hundred times the distance from Earth to the Sun); locations were based on clustering of the aphelion distances of periodic comets.^[130]
• The discovery of Sedna with its peculiar orbit in 2004 led to the conclusion that something beyond the known eight planets had perturbed Sedna away from the Kuiper belt. That object could have been an unknown planet on a distant orbit, a random star that passed near the Solar System, or a member of the Sun's birth cluster.^[49][50][51] The announcement in March 2014 of the discovery of a second sednoid, 2012 VP113, which shared some orbital characteristics with Sedna and other extreme trans-Neptunian objects, further raised the possibility of an unseen super-Earth in a large orbit.^[131][132]
• In 2008 Tadashi Mukai and Patryk Sofia Lykawka suggested that a distant Mars- or Earth-sized minor planet currently in a highly eccentric orbit between 100 and 7013299195741400000♠200 AU and orbital period of 1000 years with an inclination of 20° to 40° was responsible for the structure of the Kuiper belt. They proposed that the perturbations of this planet excited the eccentricities and inclinations of the trans-Neptunian objects, truncated the planetesimal disk at 48 AU, and detached the orbits of objects like Sedna from Neptune. During Neptune's migration this planet is posited to have been captured in an outer resonance of Neptune and to have evolved into a higher perihelion orbit due to the Kozai mechanism leaving the remaining trans-Neptunian objects on stable orbits.^{[133][134][135][136]}
• In 2012, after analysing the orbits of a group of trans-Neptunian objects with highly elongated orbits, Rodney Gomes of the National Observatory of Brazil proposed that their orbits were due to the existence of an as yet undetected planet. This Neptune-massed planet would be on a distant orbit that would be too far away to influence the motions of the inner planets, yet close enough to cause the perihelia of scattered disc objects with semi-major axes greater than 300 AU to oscillate, delivering them into planet-crossing orbits similar to those of (308933) 2006 SQ372 and (87269) 2000 OO67 or detached orbits like that of Sedna. Alternatively the unusual orbits of these objects could be the result of a Mars-massed planet on an eccentric orbit that occasionally approached within 33 AU.^[137][138] Gomes argued that a new planet was the more probable of the possible explanations but others felt that he could not show real evidence that suggested a new planet.^[139] Later in 2015, Rodney Gomes, Jean Soares, and Ramon Brasser proposed that a distant planet was responsible for an excess of centaurs with large semi-major axes.^[140][141]

Visibility and location

Due to its extreme distance from the Sun, Planet Nine would reflect little sunlight, potentially evading telescope sightings.^[6] It is expected to have an apparent magnitude fainter than 22, making it at least 600 times fainter than Pluto.^{[2][upper-alpha 26]} If Planet Nine exists and is close to its perihelion, astronomers could identify it based on existing images. For its aphelion, the largest telescopes would be required. However, if the planet is currently located in between, many observatories could spot Planet Nine.^[26] Statistically, the planet is more likely to be closer to its aphelion at a distance greater than 500 AU.^[2] This is because objects move more slowly when near their aphelion, in accordance with Kepler's second law.

Searches of existing data

The search in databases of stellar objects performed by Batygin and Brown has already excluded much of the sky the predicted planet could be in, save the direction of its aphelion, or in the difficult to spot backgrounds where the orbit crosses the plane of the Milky Way, where most stars lie.^[30] This search included the archival data from the Catalina Sky Survey to magnitude c. 19, Pan-STARRS to magnitude 21.5, and infrared data from WISE.^[2][2][30]

David Gerdes who helped develop the camera used in the Dark Energy Survey claims that it is quite possible that one of the images taken for his galaxy map may actually contain a picture of Planet Nine, and if so, purpose-built software, which was used to identify objects such as 2014 UZ224, can help to find it.^[145]

Michael Medford and Danny Goldstein, graduate students at the University of California, Berkeley, are also examining archived data using a technique that combines multiple images, taken at different times. Using a supercomputer they will offset the images to account for the calculated motion of Planet Nine, allowing many faint images of a faint moving object to be combined to produce a brighter image.^[95]

A search combining multiple images collected by WISE and NEOWISE data has also been conducted without detecting Planet Nine. This search covered regions of the sky away from the galactic plane at the "W1" wavelength (the 3.4 μm wavelength used by WISE) and is estimated to be able to detect a 10 Earth mass object out to 800–900 AU.^[146][8]

Ongoing searches

Because the planet is predicted to be visible in the Northern Hemisphere, the primary search is expected to be carried out using the Subaru Telescope, which has both an aperture large enough to see faint objects and a wide field of view to shorten the search.^[132] Two teams of astronomers—Batygin and Brown, as well as Trujillo and Sheppard—are undertaking this search together, and both teams cooperatively expect the search to take up to five years.^[21][147] Brown and Batygin initially narrowed the search for Planet Nine down to roughly 2,000 square degrees of sky near Orion, a swath of space, that in Batygin's opinion, could be covered in about 20 nights by the Subaru Telescope.^[148] Subsequent refinements by Batygin and Brown have reduced the search space to 600–800 square degrees of sky.^[149]

A zone around the constellation Cetus, where Cassini data suggest Planet Nine may be located, is being searched as of 2016 by the Dark Energy Survey—a project in the Southern Hemisphere designed to probe the acceleration of the Universe.^[150] DES observes about 105 nights per season, lasting from August to February.

Radiation

Although a distant planet such as Planet Nine would reflect little light, it would still be radiating the heat from its formation as it cools due to its large mass. At its estimated temperature of 47 K (−226.2 °C), the peak of its emissions would be at infrared wavelengths.^[34] This radiation signature could be detected by Earth-based submillimeter telescopes, such as ALMA,^[151] and a search could be conducted by cosmic microwave background experiments operating at mm wavelengths.^{[152][153][154][upper-alpha 27]} Additionally, Jim Green of NASA is optimistic that it could be observed by the James Webb Space Telescope, the successor to the Hubble Space Telescope, that is expected to be launched in 2021.^[156]

Citizen science

Cassini measurements of perturbations of Saturn

An analysis of Cassini data on Saturn's orbital residuals was inconsistent with Planet Nine being located with a true anomaly of −130° to −110° or −65° to 85°. The analysis, using Batygin and Brown's orbital parameters for Planet Nine, suggests that the lack of perturbations to Saturn's orbit is best explained if Planet Nine is located at a true anomaly of 7000205599785884932♠117.8°+11°
−10°. At this location, Planet Nine would be approximately 7013942466585410000♠630 AU from the Sun,^[173] with right ascension close to 2^h and declination close to −20°, in Cetus.^[174] In contrast, if the putative planet is near aphelion it could be moving projected towards the area of the sky with boundaries: right ascension 3.0^h to 5.5^h and declination −1° to 6°.^[175]

An improved mathematical analysis of Cassini data by astrophysicists Matthew Holman and Matthew Payne tightened the constraints on possible locations of Planet Nine. Holman and Payne developed a more efficient model that allowed them to explore a broader range of parameters than the previous analysis. The parameters identified using this technique to analyze the Cassini data was then intersected with Batygin and Brown's dynamical constraints on Planet Nine's orbit. Holman and Payne concluded that Planet Nine is most likely to be located within 20° of RA = 40°, Dec = −15°, in an area of the sky near the constellation Cetus.^[176][150]

The Jet Propulsion Laboratory has stated that according to their mission managers and orbit determination experts, the Cassini spacecraft is not experiencing unexplained deviations in its orbit around Saturn. William Folkner, a planetary scientist at JPL stated, "An undiscovered planet outside the orbit of Neptune, 10 times the mass of Earth, would affect the orbit of Saturn, not Cassini ... This could produce a signature in the measurements of Cassini while in orbit about Saturn if the planet was close enough to the Sun. But we do not see any unexplained signature above the level of the measurement noise in Cassini data taken from 2004 to 2016."^[177] Observations of Saturn's orbit neither prove nor disprove that Planet Nine exists. Rather, they suggest that Planet Nine could not be in certain sections of its proposed orbit because its gravity would cause a noticeable effect on Saturn's position, inconsistent with actual observations.

Analysis of Pluto's orbit

An analysis of Pluto's orbit by Matthew J. Holman and Matthew J. Payne found perturbations much larger than predicted by Batygin and Brown's proposed orbit for Planet Nine. Holman and Payne suggested three possible explanations: systematic errors in the measurements of Pluto's orbit; an unmodeled mass in the Solar System, such as a small planet in the range of 60–7013149597870700000♠100 AU (potentially explaining the Kuiper cliff); or a planet more massive or closer to the Sun instead of the planet predicted by Batygin and Brown.^[178][113]

Optimal orbit if objects are in strong resonances

An analysis by Sarah Millholland and Gregory Laughlin indicates that the commensurabilities (period ratios consistent with pairs of objects in resonance with each other) of the extreme TNOs are most likely to occur if Planet Nine has a semi-major axis of 654 AU. They used 11 then-known extreme TNOs with their semi-major axis over 200, and perihelion over 30 AU , with five bodies close to four simple ratios (5:1, 4:1, 3:1, 3:2) with a 654 AU distance: 2002 GB32, 2000 CR105 (5:1), 2001 FP185 (5:1), 2012 VP113 (4:1), 2014 SR349, 2013 FT28, 2004 VN112 (3:1), 2013 RF98, 2010 GB174, 2007 TG422, and (90377) Sedna (3:2). Beginning with this semi-major axis they determine that Planet Nine best maintains the anti-alignment of their orbits and a strong clustering of arguments of perihelion if it is near aphelion and has an eccentricity e ≈ 0.5, inclination i ≈ 30°, argument of perihelion ω ≈ 150°, and longitude of ascending node Ω ≈ 50° (the last differs from Brown and Batygin's value of 90°).^{[upper-alpha 28]} The favored location of Planet Nine is a right ascension of 30° to 50° and a declination of −20° to 20°. They also note that in their simulations the clustering of arguments of perihelion is almost always smaller than has been observed.^[130]

A previous analysis by Carlos and Raul de la Fuente Marcos of commensurabilities among the known eTNOs using Monte Carlo techniques revealed a pattern similar to that of the Kuiper belt, where accidental commensurabilities occur due to objects in resonances with Neptune. They find that this pattern would be best explained if the eTNOs were in resonance with an additional planetary-sized object beyond Pluto and note that a number of these objects may be in 5:3 and 3:1 resonances if that object had semi-major axis of ≈700 AU.^[180]

A later analysis by Elizabeth Bailey, Michael Brown and Konstantin Batygin found that if Planet Nine is in an eccentric and inclined orbit the capture of many of the eTNOs in higher order resonances and their chaotic transfer between resonances prevent the identification of Planet Nine's semi-major axis using current observations. They also determined that the odds of the first six objects observed being in N/1 or N/2 period ratios with Planet Nine are less than 5% if it is in an eccentric orbit.^[181][182]

Ascending nodes of large semi-major axis objects

In an article by Carlos and Raul de la Fuente Marcos evidence is shown for a possible bimodal distribution of the distances to the ascending nodes of the extreme TNOs. This correlation is unlikely to be the result of observational bias since it also appears in the nodal distribution of large semi-major axis centaurs and comets. If it is due to the extreme TNOs experiencing close approaches to Planet Nine, it is consistent with a planet with a semi-major axis of 300–400 AU.^[183][184]

Orbits of nearly parabolic comets

An analysis of the orbits of comets with nearly parabolic orbits identifies five new comets with hyperbolic orbits that approach the nominal orbit of Planet Nine described in Batygin and Brown's initial article. If these orbits are hyperbolic due to close encounters with Planet Nine the analysis estimates that Planet Nine is currently near aphelion with a right ascension of 83°–90° and a declination of 8°–10°.^[185] Scott Sheppard, who is skeptical of this analysis, notes that many different forces influence the orbits of comets.^[113]

Possible disrupted binary

Similarities between the orbits of 2013 RF98 and (474640) 2004 VN112 have led to the suggestion that they were a binary object disrupted near aphelion during an encounter with a distant object. The visible spectra of (474640) 2004 VN₁₁₂ and 2013 RF₉₈ are also similar but very different from that of 90377 Sedna. The value of their spectral slopes suggests that the surfaces of (474640) 2004 VN₁₁₂ and 2013 RF₉₈ can have pure methane ices (like in the case of Pluto) and highly processed carbons, including some amorphous silicates.^[186][187] The disruption of a binary would require a relatively close encounter with Planet Nine,^[188] however, which becomes less likely at large distances from the Sun.