February 12, 2026
in where they detect KBOs, these survey biases could directly translate into biases in the distribution of longitude of perihelion. There is another well-known longitude bias in KBO surveys: nearly all surveys avoid the galactic plane, where the density of background stars makes the discovery of faint KBOs challenging. This bias is, however, symmetric and would beunable to cause clustering of a in one direction. Nonetheless, it is important to evaluate the possibility that the combination of choices of survey observing locations, weather, and stellar density could conceivably conspire to cause apparent clustering where none is present.
Brown (2017) addresses the issue of observational bias in a by using the entire cat- alog of KBO discoveries to construct individual bias functions for each of the distant KBOs. The probability distribution function of a for each distant object indeed shows considerable bias, particularly against the galactic plane, but no overall preference for a single direction is found. Brown and Batygin (2019) updated the probability that the observed clustering is due to random chance by using these calculated bias functions to perform Monte Carlo simulations where they randomly select a for each of the distant objects by assuming a uniform distribution modified by the observational bias of the individual object. The clustering seen in these Monte Carlo samples (as measured by a modified version of the Rayleigh Z-test) exceeds that of the real data only 4% of the time. Thus, including observational biases, a is clustered at the 96% confidence level. It is interesting to note that although observational biases exist, they make little differ- ence to the assessment of significance of the clustering. This result is not surprising given the lack of biases that would plausibly cause clustering in a single direction.
Clustering of the Orbital Planes
A second intriguing feature exhibited by long-term stable, distant KBOs is that they cluster around a common orbital plane, which is appreciably inclined with respect to the ecliptic. This clustering is easily discernible in the polar projection of the angular momentum vectors shown in Figure 6, where the gray and purple points predominantly occupy the upper half of the graph. It is this effect, combined with the perihelion con- finement discussed above, that gives rise to the approximate alignment of orbits in physical space (Batygin and Brown, 2016a). An important side-effect of this align- ment is the grouping of the argument of perihelion, v = a Ω, first pointed out by Trujillo and Sheppard (2014). It is further worth noting that although clustering of v
remains robust beyond a & 250 AU, detections of new KBOs have largely eliminated this feature in the 150 AU . a . 250 AU orbital domain.
Unlike apsidal confinement – which is largely captured by a single parameter –
clustering of the orbital planes requires simultaneous alignment of the longitudes of ascending node as well as the inclinations of long-period KBOs, in order to ensue. For this reason, rather than quoting a single degree of grouping as we did for a, we instead consider the mean inclination and node of the distant objects, as well as the rms dispersion of KBO inclinations about this mean plane. Quantitatively, these numbers are i 7 deg, Ω 82 deg and σi 15 deg respectively. These quantities are illustrated graphically in the inset of Figure 6 by a dashed circle that is centered on a point marked with an symbol. To further exemplify the (e, a) and (i, Ω) data in a uniform fashion, in the bottom panel of Figure 8 we show the longitudes of ascending node as a function of the semi-major axis, and label each object by its inclination.
Much in the same way that the apsidal confinement discussed in the previous sub- section is susceptible to differential precession, clustering of the orbital planes is sub- ject to dispersal due to the differential regression of the ascending node, induced by thegravity of the giant planets. Applying Lagrange’s equations as before, we obtain the rate of nodal regression from equation (3) as follows (Murray and Dermott, 1999):
dΩ –1 b2¯ 3 , G M cos(i) X8 mj a2
, = – .
Noting that the differential regression of the node operates on a comparable timescale to perihelion precession (given by equation 4), we assert that in absence of a sustained restoring torque that maintains the near-alignment of long-period KBO orbits in phys- ical space, the quadrupolar gravitational field associated with the known giant planets would randomize the orientations of the distant orbits on a timescale that is short rela- tive to the age of the solar system8.
Once again, we need to consider the possibility that the clustering of the orbital planes is due to chance, along with observational biases (Shankman et al., 2017). Like a, the measured distributions of both Ω and i are easily biased by where surveys have detected objects. For example, most KBO surveys target at or near the ecliptic, re- quiring that objects in the survey be near either their ascending or descending node. A strong bias likewise exists for survey inclinations to be approximately equal to the ecliptic latitude at which the survey is undertaken. As before, no obvious set of biases should lead to the sort of clustering seen in the data. Nonetheless a full analysis is clearly needed.
Brown and Batygin (2019) extend the method previously developed in Brown (2017) to calculate simultaneous bias functions for Ω and i. As before, they use the entire col- lection of KBO discoveries to calculate a probability distribution function of discovered orbital planes for each distant object under the assumption that Ω is uniform and i is
distributed identically to the a ≤ 250 AU scattered KBOs. They find that the 14 KBOs with a > 250 AU are clustered in their orbital poles at the 96.5% confidence level.
Both the apsidal orientation and orbital plane clustering are moderately significant, but the combined probability of both occurring simultaneously is only 0.2% (Brown and Batygin, 2019). The distant KBOs are thus distinctly clustered at the 99.8% confi- dence level.
Highly Inclined TNOs
A third puzzling population of trans-Neptunian objects is comprised of minor bod- ies that reside on orbits that are strongly inclined with respect to the plane of the solar system. Indeed, the current census of known TNOs contains 49 objects with incli- nations above i > 40 deg, with 10 of them occupying orbits with i > 90 deg. With the exception of the recently discovered KBO 2015 BP519 (i = 54 deg, q = 36 AU, a = 450 AU; Becker et al. 2018) all presently known high-inclination objects are Cen- taurs, meaning that they have q < 30 AU, and therefore veer into inter-planetary space 8Accounting for Neptune scattering (which equations 4 and 5 explicitly neglect) would further reduce the characteristic timescale for randomization. However, by restricting our consideration of the observations to the subset of stable and metastable objects, we largely circumvent this complication.
Figure 9: Trans-Neptunian objects with high inclinations. This diagram shows the i > 50 deg subset of the distant solar system’s small body population, as viewed from the ecliptic plane. Long-period Centaurs with q < 30 AU and a ≥ 250 AU are shown in orange, while the KBO 2015 BP519 is denoted in pink. The orbits of more proximate high-inclination TNOs are rendered in cyan. Generically speaking, trans-Neptunian objects with i > 50 deg cannot be natually explained by the standard model of the solar system’s formation and dynamical evolution.
at perihelion. For consistency with the preceding discussion, we sub-divide the high- inclination population of TNOs into a long-period component with a > 250 AU, and a sub-population occupying the more proximate part of the Kuiper belt (Figure 9).
Objects with orbital inclinations in excess of a few tens of degrees are neither a natural result of the solar system formation process (which unfolds in a geometrically thin, dissipative disk of gas and dust; Armitage 2010), nor an expected outcome of the solar system’s post-nebular evolution (Morbidelli et al., 2008). Indeed, detailed numerical simulations of the solar system’s early dynamical relaxation carried out by Levison et al. (2008) generate an inclination dispersion of TNOs that is largely confined to i . 30 deg. While the more finely-tuned simulation suite of Nesvorny´ (2015b) boost
the upper envelope of the inclination distribution to i 40 deg, perpendicular and
strongly retrograde objects such as Drac (Gladman et al., 2009), Niku (Chen et al., 2016) as well other long-period Centaurs (Gomes et al., 2015) are never produced in these calculations. This picture is further consistent with the recent simulation suite of Becker et al. (2018), who demonstrate that even the i = 54 deg orbit of 2015 BP519 has a negligible chance of being produced self-consistently through Neptune scattering.
In order to understand why interactions with the canonical giant planets do notroutinely yield highly inclined orbits, it is useful to consider a simplified toy-model of the dynamical evolution of scattered disk objects, wherein Neptune’s orbit is taken to be circular, and acts as the sole source of perturbations. Within the framework of this circular restricted three body problem, the conservation of the test particle’s specific energy is replaced by a near-conservation of the Tisserand parameter (Murray and Dermott, 1999):
T = a8 + 2 , a .1 – e2 cos(i), (6)
where a8 is Neptune’s semi-major axis. Notably, this quantity is nothing other than the Jacobi constant (which is an exact integral of motion), in the limit of small planet/star mass ratio.
Employing the conservation of T , let us consider the fate of a particle that is ini- tialized on a nearly planar, circular orbit, in the immediate vicinity of Neptune (e.g. at a = 33 AU). Facilitated by chaotic diffusion arising from overlap of Neptunian reso- nances (Wisdom 1980; Murray 1986; see also Gomes et al. 2008 and the references therein), the test particle will experience a random walk in energy, thereby slowly in- creasing its semi-major axis, while maintaining its perihelion distance close to its point of origin, at q a8. The characteristic inclination attained by this particle throughout its evolution is then
i = arccos a T – a8 , a8 ! ~ 10 deg, (7)
2 a q (2a – q)
where we have set cos(i) 1 in evaluating equation (6), due to the assumed near- coplanarity of the initial condition. This simple calculation shows that it is the sym- metry entailed by the conservation of T that prevents significant excitation of orbital inclination in the scattered disk by Neptune.
Of course, the real dynamics of the solar system are more complicated than those encompassed by the circular restricted three body problem. Nevertheless, the quali- tative picture implied by the above calculation illuminates the relevant limitations on the degree of orbital excitation that can ensue in the scattered disk population of the Kuiper belt. To restate this result simply, an external gravitational influence is required to generate the high-inclination orbits observed in the distant solar system.
One potential solution to this conundrum is to imagine that rather than being native to the Kuiper belt, these highly inclined objects are sourced from the Oort cloud. That is, perturbed by the combined action of the Galactic tide and passing stars (e.g., Heisler and Tremaine 1986), very long period objects acquire near-parabolic trajectories, and upon crossing the orbit of Neptune get scattered inwards, becoming exotic members of the scattered disk (Levison et al., 2001). Although such a scenario can in principle be invoked for high-i Centaurs (see, however, Gomes et al. 2015), considerable fine- tuning would likely be required to account for q > 30 AU objects like 2015 BP519 in this manner.
As always, we could consider observational bias here, but in this case there is nearly no need. Surveys for KBOs which observe near the ecliptic are biased against high in- clination objects by a factor of 1/ sin(i) (Brown, 2001). Thus, the detection of any
KBOs and Centaurs with large inclinations suggests that the true population of high-i TNOs is much more prominent. In other words, independent of observational bias, nearly perpendicular as well as strongly retrograde TNOs exist, and require an expla- nation beyond that which can be formulated within the framework of an eight-planet solar system.
The Planet Nine Hypothesis: Analytical Theory
The current observational census of long-period Kuiper belt objects indicates that the dynamical origin of the anomalous structure of the distant Kuiper belt requires sustained perturbations beyond those that can be generated by the known giant planets of the solar system. In other words, a separate source of gravitational influence in the trans-Neptunian region is required to explain the anomalous patterns exhibited by the data. A series of recent studies (Batygin and Brown, 2016a,b; Batygin and Morbidelli, 2017; Millholland and Laughlin, 2017; Becker et al., 2017; Hadden et al., 2018; Khain et al., 2018a; Ca´ceres & Gomes, 2018; Li et al., 2018) have demonstrated that the existence of an additional, multi-Earth mass planetary member of the solar system – Planet Nine – can successfully resolve this theoretical tension. More specifically, (1) the apsidal confinement of distant KBOs, (2) perihelion detachment of long-period
orbits (3) clustering of the a & 250 AU orbital planes, and (4) excitement of extreme TNO inclinations can all be simultaneously explained by Planet Nine (P9)-induced
dynamical evolution.
In order to define the Planet Nine hypothesis as a specific theoretical prediction, we begin by presenting a purely analytical description of the dynamical mechanisms through which P9 is envisioned to sculpt the distant Kuiper belt. By and large, the following discussion will draw upon ideas of secular perturbation theory of celestial mechanics. Within the framework of this formalism, the Keplerian motion of all objects in the system is averaged over, leaving long-term exchange of angular momentum (but not energy) among the constituent bodies as the only physically active process. In other words, instead of computing the evolution of test-particles under the influence of point masses as done in the direct treatment of the gravitational N-body problem, the aim of secular theory lies in calculating the long-term behavior of test-orbits, subject to perturbations from massive wires that trace out the planetary trajectories9. Inherently, this mathematical procedure is equivalent to the so-called Gaussian averaging method, wherein the orbiting bodies are replaced by massive wires with line-densities that are inversely proportional to the instantaneous Keplerian velocities (Touma et al., 2009).
Secular Forcing
In order to classify the various flavors of secular interactions that can arise due to a distant planetary perturber, it is of considerable use to examine the approximate form of the P9-KBO coupling function i.e., Planet Nine’s orbit-averaged gravitational
No comments:
Post a Comment