614 Encyclopedia of the Solar System
that their sample of KBOs and Centaurs divided into two
distinct color groups—gray objects with 1.0<B-R<1.4
and red objects with 1.5<B-R<2.0. They found almost
no objects with 1.4<B-R<1.5. They did not have a physi-
cal explanation for their surprising result. Everyone agreed
KBOs and Centaurs did not exhibit only red surface colors.
As the groups pressed hard at telescopes to measure
more surface colors and test their initial findings, sample
sizes grew from 10 to more than 100 objects. Once the
sample sizes became large enough, it became apparent that
different dynamical classes of KBOs had different color sig-
natures.
9.1.1 CENTAUR OBJECTS
Figure 9a shows a histogram of the number of objects vs.
B-R color for a sample of 22 Centaur objects. Fourteen
objects have B-R<1.3 and eight objects have B-R>1.7.
Notice there are no objects with 1.3<B-R<1.7. Is it possi-
ble that Centaurs actually exhibit a uniform distribution of
B-R colors and either insufficient sampling or chance is re-
sponsible for the apparent split into two B-R color groups?
Application of statistical tests like the “dip test” tell us that
the probability of making observations in Figure 9a for an
actual uniform distribution of B-R colors is about 1 in 100.
The split into two B-R color groups appears to be real. Un-
like the earlier controversy, two groups, one led by Nuno
Peixinho and the other by this author, find the same highly
unusual split. What makes the split so unusual is that there
doesn’t seem to be any other physical property that cor-
relates with the color of a Centaur object. For example,
if Centaurs that came closest to the Sun were all gray, we
might suspect that the warmth of the Sun was chemically or
physically altering the surfaces and graying them. But there
is no statistically significant correlation between color and
perihelion distanceor any other orbital element.
9.1.2 CLASSICAL KBOS
Figure 9b shows a histogram of the number of objects vs.
B-R color for a sample of 21 classical KBOs. All 21 classical
objects have B-R>1.5, i.e., there are no gray objects at all
in the sample. Classical KBOs exhibit the color signature
originally expected for all KBOs.
9.1.3 SCATTERED DISK OBJECTS
Figure 9c shows a histogram of the number of objects vs.
B-R color for a sample of 20 SDOs. Seventeen of the 20
objects exhibit B-R<1.5. There appears to be a deficit of
red objects among this group.
FIGURE 9 Correlations between colors and orbital properties
of KBOs and Centaurs. (a) A sample of 22 Centaurs neatly divide
into two color groups; 14 objects exhibit B-R<1.3 and eight
objects exhibit B-R>1.7. Surprisingly, there are no Centaurs
with 1.3<B-R<1.7. (b) All 21 objects of a sample of classical
KBOs with q>40 AU, e<0.1 and e< 10 ◦are all red (B-R
>1.5). (c) A sample of 20 SDOs are mostly gray (B-R<1.5).
The mechanisms responsible for these correlations between
color and orbital properties are not well understood yet.