Fig. 30. Comparison of the optical near infrared reflection spectrum of D-type asteroid (368) Haidea (points) with the Tagish Lake meteorite (line), showing a nearly perfect match. Figure from [60]

proxy for the spectra and so for surface composition. Problems with this approach are numerous. Colors cannot, in general, be used to determine compositions Fig. 30. Colors are influenced by composition, but also by wavelength-dependent scattering effects in particulate regoliths, and by viewing geometry. On the other hand, colors can be used to classify objects into groups. The Holy Grail of colorimetric work on the KBOs has been for some years to find correlations between the colors and other properties such as size and orbital character [57,75,102,106,145-147]. Correlations like this might provide illuminating clues about the KBOs and their histories.

The use of color to learn about KBOs has been, to say the least, an up-hill battle. The first property to be measured was color diversity; the KBOs exhibit a range of surface optical colors that is large compared with the uncertainties of measurement. In fact, color diversity has emerged as the only physical property to be confirmed by every subsequent study. Later, color diversity at optical wavelengths was found to extend into the near infrared [27,31,33,74,106]. Moreover, the optical and infrared colors are correlated, which indicates that a single coloring agent is responsible for the wavelength dependence of the reflectivity across the wavelength range from B-band (0.45 |m) to J-band (1.2|m) and perhaps beyond to K-band (2.2|m).

The physical significance of color diversity is unclear. One possibility is that the different colors reflect intrinsically different compositions. This might be the case, but it is difficult to understand why the compositions of the measured KBOs would be so varied. After all, the measured objects are located in a comparatively narrow band between about 30 and 50 AU, where the radiation equilibrium temperatures vary from —40 to — 50 K. This very small temperature range could scarcely effect the compositions of the KBOs enough to cause major color differences.

For this reason, a second model was proposed to explain the color dispersion. In this "resurfacing model," the hemispherically averaged color of a KBO is time-dependent and determined by a competition between collisional resurfacing and cosmic ray processing. For example, suppose that cosmic ray processing causes an exposed surface to become redder on timescale rcl. This process competes with impact-driven resurfacing, in which impacts excavate "fresh" material from beneath the irradiated layer. If the excavated matter has a different (neutral?) color, the instantaneous, hemispheric average color will vary stochastically between extremes set by fully radiation-processed matter and fresh, excavated material. Substantial color fluctuations are possible when the timescale for resurfacing, rcon is — rcl.

Attractive though it at first seems, several predictions of the resurfacing model have not been confirmed by observations. The model predicts that rotational color variations on KBOs should be nearly as large as the color differences that exist between KBOs of a given size. This is not observed. The model also predicts that the range of colors observed should vary with KBO size, because the timescale for collisional resurfacing varies with object size while rcl does not. Again, this violates the observations. The model has been extended by the addition of color variations owing to possible outgassing effects [31] but the problems remain. Collisional resurfacing is unlikely to be responsible for the color dispersion of the KBOs, although it could conceivably be a contributing factor.

Tegler and Romanishin reported that the colors of KBOs were not just dispersed over a wide range but were bimodally distributed [145]. They continued to find bimodal color distributions with larger samples [146,147] but failed to receive observational support for this finding from independent observers [31,36,57,75]. The colors of the KBOs available at the time of writing (March 2006) are distinctly unimodal (see Figs. 31, 32 and 33). Recently, Peix-inho et al. [121] reported that, while the KBO colors are indeed unimodally distributed, the Centaurs appear bimodal (see the next section). This is more than an academic distinction: a bimodal color distribution would have placed strong constraints on the nature of the KBOs, had it been real.

Few of the long-sought correlations between colors and other physical and dynamical properties have turned out to be observationally robust. The correlation that seems most likely to be real is between color and perihelion distance [146] or, equivalently, between color and inclination [152] amongst the classical KBOs. The perihelion vs. inclination ambiguity arises because these quantities are loosely related amongst the Classical objects. Doressoundi-ram [35] finds that the color vs. perihelion distance correlation is slightly

Fig. 31. Color-color diagram for classical KBOs. From [32]
Fig. 32. Color-color diagram for resonant KBOs. From [32]
Fig. 33. Color-color diagram for scattered KBOs. From [32]

stronger than the color vs. inclination correlation. Trujillo and Brown [152] find that classical objects with small inclinations are redder, on average, than those with high inclinations. The latter observation has been factored into dynamical models by R. Gomes [53]. He asserts that the high inclination ("hot") Classical KBOs were scattered outward while the low inclination ("cold") Classical KBOs were formed exterior to Neptune, where they now reside [53]. Whether or not this is true, the central mystery that is unaddressed by dynamical models is why the cold and hot populations would have different colors. As measured by the B-I color index, the color vs. inclination (or color vs. perihelion) correlation appears secure at the 3a or 4a confidence level. However, the correlation is absent when V-R or V-I color indices are used [138]. One possibility is that the color correlation is forced by the B data (for example, there could be a B-band absorber whose distribution is correlated with inclination or perihelion distance but which would have no effect on color indices at wavelengths longer than B). As new observations are collected, it will be interesting to see whether or not the reported correlation will survive. No convincing explanation for the correlation, if real, has been suggested.

About a dozen KBOs possess both color and albedo determinations [25]. These are plotted in Fig. 34 together with corresponding data for the nuclei of comets, the Jovian Trojans, and Centaurs [80]. There it is seen that the wide dispersion of colors of the Centaurs and KBOs is matched by a wide dispersion in the albedos, with the large objects 2003 EL61 and Pluto defining one extreme. By comparison, the nuclei of the comets and the Jovian Trojans

Was this article helpful?

0 0

Post a comment