A thoroughly sporadic column from astronomer Mike Brown on space and science, planets and dwarf planets, the sun, the moon, the stars, and the joys and frustrations of search, discovery, and life. With a family in tow. Or towing. Or perhaps in mutual orbit.

A new window on astronomers

One of the goals of this column is to make it more clear what astronomers actually do all day (or night) long. As I have been discussing, one of the things that I do frequently is writing proposals to convince other astronomers to let me use their telescope. The once-a-year proposals for use of the Hubble Space Telescope were due on Friday at 5pm (I pressed the “send” button at 4:49pm). Here is the proposal that I submitted. While it is written for other astronomers, so often flies into astronomical shorthand, I think it is at least moderately readable by anyone generally interested in what is going on in the outer part of the solar system. Plus, there is no chance I could possibly attempt to write anything else this week. In that spirit, here is:

A compositional-dynamical survey of the Kuiper belt: a new window on the formation of the outer solar system.

The eight planets overwhelmingly dominate the solar system by mass, but their small numbers, coupled with their stochastic pasts, make it impossible to construct a unique formation history from the dynamical or compositional characteristics of them alone. In contrast, the huge numbers of small bodies scattered throughout and even beyond the planets, while insignificant by mass, provide an almost unlimited number of probes of the statistical conditions, history, and interactions in the solar system. Studies of these small bodies have been exploited for many years in the inner part of the solar system, where combined dynamical and compositional observations of asteroids have been used to trace chemical gradients, study early radioactivity, and detect and analyze collisional histories in the region of the terrestrial planets (Bottke et al. 2005 and references therein). While a similar study of the Kuiper belt would offer similar promise for understanding the formation of the region of the giant planets, the typical objects in the Kuiper belt are 10,000 times fainter than those in the asteroid belt, so this promise has been hampered by the difficulty of obtaining concrete observations of the surface compositions of these objects.

Instead, attempts to understand the formation and evolution of the Kuiper belt have largely been dynamical simulations where a hypothesized starting condition is evolved under the gravitational influence of the early giant planets and an attempt is made to reproduce the current observed populations(Levison and Morbidelli 2003, Tsiganis et al. 2005, Charnoz and Morbidelli 2007, Lykawka and Mukai 2008). With little compositional information known for the real Kuiper belt, the test particles in the simulation are free to have any formation location and history as long as they end at the correct point. Allowing compositional information to guide and constrain these studies would add an entire new dimension to our understanding of the formation and evolution of the outer solar system.

New visible-infrared capabilities of WFC3 allow such compositional information of a large number of Kuiper belt objects to be obtained for the first time. Here we propose to exploit these capabilities to perform the first ever large-scale dynamical-compositional study of Kuiper belt objects (KBOs) and their progeny to study the chemical, dynamical, and collisional history of the region of the giant planets.

Kuiper belt compositions: the current view.
Combining compositional and dynamical information on small bodies has proved a powerful technique in the inner solar system for understanding the formation of the terrestrial planetary region, but it has only been used to a very limited extent in the outer solar system. Color diversity.The earliest attempts to jointly consider outer solar system compositions and dynamics simultaneously were attempted using only the colors of KBOs.While colors are a poor proxy for composition, they have proved a fascinating early tracer of the dynamical homogeneity -- and lack thereof -- of the Kuiper belt. The earliest photometric observations (Jewitt and Luu 1998) suggested that KBOs came in a wide variety of colors and that there was no relationship between the color and any orbital or physical parameter of the object. To date, this great heterogeneity remains unexplained, though it clearly points to a wide diversity of formation or evolutionary histories throughout the Kuiper belt.

The cold classical KBOs.
Subsequent observations of colors of larger numbers of KBOs eventually showed that one dynamical subset of the Kuiper belt, the ``cold classical KBOs'' on dynamically cold low inclination and eccentricity orbits, consists exclusively of objects that are red (Tegler and Romanishin 2000, Trujillo and Brown 2002). While the color red is impossible to interpret compositionally without more spectral information, the existence of this red grouping has been used to argue that the cold classicals are a unique population whose dynamical coherence has been maintained through the dramatic evolution of the outer solar system (Morbidelli and Brown 2004). The need to retain this group of objects is one of the key constraints on -- and sometimes the death of -- models of the evolution of the outer solar system and is the earliest example of the power of combining (even limited) compositional information with small body dynamics.

While color groupings have proved interesting for helping to understanding the evolution and rearrangement of the outer solar system, the actual cause for the different colors remains unknown.Infrared spectroscopy would allow a direct probe of the surface ices common in the outer solar system, but for many years few infrared spectra were available, as few KBOs were bright enough for even the lowest resolution spectroscopy with the largest telescopes. This difficulty was partially alleviated by our wide field search for the largest KBOs (Brown 2008), which finally provided a moderate number of bright observable objects, and by long term programs at VLT and Keck that slowly obtained spectra of the very brightest of these (i.e. Barucci et al. 2006, Barkume et al. 2008). The most systematic survey to date is our Keck survey (Barkume et al. 2008), which obtained 1.5 to 2.5 micron spectra of 45 objects in the outer solar system. Three results from this small sample provide examples and details of what could be expected from a much larger survey.

Fragments from a giant primordial collision.
One small set of KBOs stood out in the Keck survey for their unique spectra (Fig 1a). This collection of objects has surfaces which look like laboratory spectra of pure uncontaminated water ice. Moreover, all of these pure water ice objects have nearly identical orbits(Fig. 2a), and the largest of them, the nearly Pluto-sized 2003 EL61, had previously been speculated to have suffered a giant impact at some point in its past which gave it its rapid spin and system of at least two moons (Brown et al. 2006). The compositional and dynamical association of the water ice objects with 2003 EL61 itself made it clear that the small set of pure water ice objects were fragments of the giant impact that had shaped 2003 EL61 (Brown et al. 2007). This impact is the largest anywhere in the solar system for which we have multiple extant fragments identified, providing a unique laboratory into the types of massive collisions which shaped the solar system.

It is expected that the 2003 EL61 impact occurred during the time of solar system clearing when the Kuiper belt was significantly more dense than its current state. A model by Levison et al. (2008) suggests that the impact actually occurred between two objects which were themselves in the processes of being scattered out of the solar system. As would be expected from a collision of objects that were on unstable orbits, some of the 2003 EL61 family is itself in an unstable region of space. In Ragozzine and Brown (2007), we exploited these instabilities to develop a dynamical chronometer to use the current spread in orbital elements of 2003 EL61 fragments to determine the time of the 2003 EL61 impact. To date, with the small number of family members known, we can only place a lower limit of 1 Gyr on the age. But with more objects discovered we will be able to more precisely date this impact, and thus date the time of solar system clearing.While almost all models to date assume that major clearing occurred 4.5 Gyr ago, the new and to date quite successful Nice model (Tsiganis et al. 2005 and papers following) posits that solar system clearing was delayed by ~1 Gyr and did not largely occur until the time of the Late Heavy Bombardment. The study of the dynamics of this compositionally unique set of objects could answer one of the most important questions about the timing of major events in the outer solar system.

The methane giants.
Schaller and Brown (2007) suggested that a small number of the largest and coldest objects should have enough surface gravity to maintain their volatiles against loss to space over the age of the solar system. In their model, the final loss to space is controlled by the slow leakage of Jeans escape from a vapor-pressure controlled atmosphere. The loss is an intimate function of the object size and of the precise orbit. The results of the model predictions to date have been nearly perfect: almost everything that the model suggests should have volatiles on the surface (predominantly methane; Fig 1a) does, and nothing that the model suggests shouldn't have volatiles has been found to have volatiles. This success opens the possibility of being able to find outliers with unusual dynamical or compositional histories by finding objects whose predictions don't fit within the framework of the model. Indeed, the one object which doesn't fit the model prediction is 2003 EL61, the giant parent of the collisional family. We presume that the impact took away most of the volatiles on the outermost fragments, but, more importantly, even if we had know nothing else about 2003 EL61, its failure to have a predictable surface composition would have quickly drawn attention to it.

Overall spectral diversity.
Once the 2003 EL61 family is removed from the spectral sample, no apparent compositional-dynamical correlation or pattern is seen in the remaining 40 objects (Fig 2a). While the compositions of asteroids are strongly stratified as a function of heliocentric distance, the KBOs have no such stratification. Just as objects with different optical colors are jumbled throughout the Kuiper belt, so are objects with different infrared spectra. Unlike the asteroid belt, however, where compositional differences are glaring and distinct, in the Kuiper belt, the spectra of almost every KBO fits along a smooth continuum with the only differences being the amount of absorption due to water ice and the optical color (Fig 1b). While initially unexpected, the lack of other significant surface ices is now understood as a natural consequence of thermal escape of the more volatile ices (Schaller and Brown 2007).

Oddly, however, little correlation appears between the optical colors and the amount of water ice absorption (Fig 1b), conflicting with the commonly held conceptual view that KBO surfaces are a simple mix between red colors due to irradiated organics and blue colors due to fresh water ice exposed by collision (Jewitt and Luu 1998) or that KBOs can be compositionally classified by optical colors alone (Barruci et al. 2006).

While the cause of this continuous diversity is unknown, the broad possibilities are limited: the surfaces can reflect either primordial differences in the objects, subsequent evolution of the objects, or both. Primordial differences would likely reflect formation location, while evolution could reflect both thermal and collisional history.

Whatever the cause of the surface composition variability, understanding the reason would allow significant new insights into the evolution of the outer solar system. If the variations are primarily primordial, we could use KBO composition to reconstruct the initial locations of the objects that are now jumbled in the Kuiper belt, while if the variations are evolutionary we will be able to use compositions to reconstruct collisional or thermal histories of different regions of the Kuiper belt. In either case, with the current small number of objects known it is impossible to determine the cause of the variability, but the promise for this potential tool is strong.

The proposal continues on for another few pages, describing precisely how we want to use the Hubble Space Telescope to answer some of these questions that we had set out here.


  1. Mike, again there wasn't answer.
    Where are bigger planetoids behind Pluto, Pluto included now!?
    All are close to their extremes of their orbits!!!!
    Eris present-current distance 97AU!!!,Aphelia97AU,..perihelia38AU
    2003EL61 52AU!!!,....52AU,...... 35AU
    2005FY9 52AU!!!!,...52AU,.......39AU
    2003UB313-Sedna 90AU!!! 975,6AU,.... 76,2AU
    close toperih. 49AU,.........30AU
    Orcus 48,3AU 30,5AU
    Quaoar 45AU 42AU
    I made my conclusions. Probability to have all planetoids close to perihelia, or aphelia is less than 1:10000!!!! Pavel Smutny

  2. I continue in previous comment,...

    Orcus is circa 48AU from Sun what is quite close to aphelia!!!!!!!

    Qaoar is circa 43AU from us-nearly circular orbit.,...
    Mike, please let us know your theory why are dwarf planets behind Pluto, Pluto included so close to their extreme points on their orbits. My theory was p├║resented in previous discussion comments,...Pavel Smutny

  3. Amendment to orbital speeds of planetoids due to opposite motion of Sun with inner planets toward X is circa 0.5km/s when planetoids behind Pluto have orb. speeds circa till 4km/s. This makes effect of more elipsoid paths of those planetoids than they are in real, prolongation about till 50 percents. Next conclusion is that X so must be much closer to us than those Planetoids. X is somewhere behind Saturn what is aprox. 5 years to perihelia of X,…2012

  4. Hahaha... I came from seeing the Sciencemag article from 2yrs ago.

    I suppose the previous commentor... eventually got some of the information he wanted, albeit after a long 8yr wait, haha!

    But... you can see where this is going!

    Your discovery/ confirmation was 2yrs ago!

    - So?

    Where our we regarding Planet X, now? Eh?

    If you saw it 2yrs ago... we have to know more by now, right?