Kuiper Belt: Dynamics 593
of the discovered objects are clustered in the 36–48 AU
range, but several others form a “tail” structure extending
beyond 50 AU. Their semimajor axes range up to several
100 AU. Their perihelion distances are generally between
30 and 38 AU (dotted curves in the bottom right panel of
Fig. 2), so that on average the orbital eccentricities increase
with semimajor axes. These objects are dynamically unsta-
ble because they suffer sufficiently close encounters with
Neptune. At each encounter, they receive an impulse-like
acceleration, which changes the semimajor axis of their or-
bits. The perihelion distance remains roughly constant dur-
ing an encounter, so that the eccentricity changes together
with the semimajor axis. Thus, under the scattering gravi-
tational action of Neptune, these objects move in a sort of
random walk in the region confined by the dotted curves.
For this reason, this population of objects is now called the
scattered disk. The name “disk” is justified, because the
orbital inclinations, although large, are significantly smaller
than 90◦, giving this population a disk-like structure.
Up to now, we have used “Kuiper Belt” to denote gener-
ically the population of objects witha>30 AU. However,
the existence of the scattered disk suggests that we should
reserve the name “Kuiper Belt” for the population of objects
that do not suffer encounters with Neptune and therefore
have orbits that either do not significantly change with time,
or do so very slowly. Adopting this definition, the objects of
the Kuiper Belt are plotted with blue and green dots in
Fig. 2, while the scattered disk objects are plotted in red.
As one sees, scattered disk objects can also havea<50 AU,
provided that they have a small perihelion distance and are
not in one of the most prominent mean-motion resonances
with Neptune (indicated by the vertical lines labeled 3:4, 2:3
and 1:2 in Fig. 2). In Fig. 2, the scattered disk seems to be
outnumbered by the Kuiper Belt bodies. However, the scat-
tered disk objects are more difficult to discover, given that
most of them have very elongated orbits and spend most of
the time very far from the Sun. Accounting for this difficulty,
astronomers have estimated that the scattered disk and the
Kuiper Belt should constitute roughly equal populations.
All solar system bodies should have accreted on quasi-
circular orbits. This is a necessary condition for small plan-
etesimals being able to stick together and form larger
objects. Indeed, if the eccentricities are large, the relative
encounter velocities are such that, upon collisions, plan-
etesimals do not grow, but fragment into smaller pieces.
This consideration suggests that the scattered disk objects
formed much closer to Neptune, on quasi-circular orbits,
and have been transported outward by the scattering ac-
tion of that planet. The fact that the scattering action is still
continuing implies that the origin of the scattered disk does
not necessarily require that the primordial solar system was
different from the current one.
However, recent observations have revealed that, in ad-
dition to the Kuiper Belt and the scattered disk, there is
a third category of objects, represented with magenta dots
in Fig. 2. Their orbital distribution mimics that of the scat-
tered disk objects, but their perihelion distance is somewhat
larger, so that they avoid the scattering action of Neptune.
Their orbits do not significantly change over the age of the
solar system. Among the objects with these orbital proper-
ties are 1995 TL 8 (a∼52 AU,q∼40 AU), 2000 CR 105
(a∼225 AU,q∼44 AU), 90377 Sedna (a∼500 AU,
q∼76 AU), and the recently discovered 136199 Eris (a∼
67 .5 AU,q=38 AU, the largest Trans-Neptunian Ob-
ject (TNO) known so far) and 2004 XR 190 (a∼ 57 .4 AU,
q∼51 AU—exceptional for its inclination of about 45◦).
For the previously listed reasons, these bodies also should
have formed closer to Neptune on much more circular or-
bits, and presumably they have been transported outward
through close encounters with the planet. However, given
that they do not undergo close encounters now, their ex-
istence suggests that the solar system was different in the
past (either the planetary orbits were different or the envi-
ronment was different—rogue planets, passing stars, etc.),
so that the scattered disk extended further out in perihelion
distance during the primordial times. We will come back to
this in Section 7.
If we look at Fig. 2 more in detail (left panels), the Kuiper
Belt can also be subdivided in a natural way in subpop-
ulations. Several objects (green dots) are located in mean-
motion resonances with Neptune. As explained in Section 2,
the mean-motion resonances provide a protection mecha-
nism, so that resonant objects can avoid close encounters
with Neptune even if their perihelion distance is smaller
than 30 AU, as it is in the case of Pluto. For this reason, reso-
nant Kuiper Belt objects can be on much more elliptic orbits
than the nonresonant ones, the eccentricities of the former
ranging up to 0.35. The objects in the 2:3 mean-motion reso-
nance with Neptune are usually called the Plutinos (because
they share the same resonance as Pluto), while those in the
1:2 resonance are sometimes called twotinos. In Fig. 2, the
resonant population seems to constitute a substantial frac-
tion of the Kuiper Belt population. However, resonant ob-
jects are easier to discover because at perihelion they come
closer to the Sun than the nonresonant ones. When account-
ing for this fact, astronomers estimate that, all together, the
objects in mean-motion resonances constitute about 10%
of the total Kuiper Belt population.
The nonresonant Kuiper Belt objects (blue dots) are usu-
ally referred to as classical. This adjective is attributed be-
cause their orbital distribution is the most similar to what
the astronomers were expecting, before the discoveries of
trans-Neptunian objects began: that of a disk of objects on
stable, low-eccentricity, nonresonant orbits. However, even
the classical population has unexpected properties. Their
eccentricities are moderate—a necessary condition to avoid
encounters with Neptune, given that they are not protected
by any resonant mechanism. Nevertheless, the eccentrici-
ties are definitely larger than those of the protoplanetary
disk in which the objects had to form. Some mechanism