(lecture12)=
# Meteorites, Asteroids, Minor Planets & Comets

```{highlights}
**Learning objectives:** By the end of this lecture, you will be able to classify meteorites and use isotopic chronometers to date the early solar system, describe the dynamics and populations of the asteroid belt and trans-Neptunian region, explain the origin and structure of comets, and use small bodies as formation fossils that record the solar system's dynamical and chemical history.
```

```{seealso}
**Slides:** [Download Lecture 12 (PDF)](../_static/slides/lecture12.pdf)
```

The previous lectures have followed the planets one by one: rocky worlds in {ref}`Lecture 9 <lecture09>` and {ref}`Lecture 10 <lecture10>`, gas and ice giants in {ref}`Lecture 11 <lecture11>`.
This lecture is about everything that did *not* become a planet.
Small bodies are the leftover pieces of the planet-building process: asteroids and meteorites in the inner solar system, comets and Kuiper Belt objects (KBOs) in the outer solar system, dwarf planets like Ceres and Pluto in between, and the Oort cloud at the gravitational edge of the Sun's domain.
Most of them are physically tiny.
Together they carry far more information about how the solar system formed than the eight planets do, because they have been processed much less.

We follow the same descriptive-first structure as in the planet lectures.
Part 1 introduces meteorites, the only solar-system samples we can study in a laboratory and the source of every absolute age in this course.
Part 2 surveys the dynamical populations: asteroid belt, near-Earth asteroids, Kuiper Belt, scattered disk, dwarf planets, and the inferred Oort cloud.
Part 3 covers the messengers and visitors: in-situ space missions, sample-return missions, planetary defence, and the new and remarkable category of interstellar interlopers.
Pluto is treated here as the largest known KBO, not as a planet. {numref}`fig:l12-small-bodies-overview` collects the small bodies visited by spacecraft to set the scale for the rest of the lecture.

```{figure} figures/small_bodies_overview.avif
:name: fig:l12-small-bodies-overview
:width: 700px
:align: center

Size comparison of representative small bodies visited by spacecraft, scaled against terrestrial landmarks for reference. Comet and asteroid nuclei span a range from sub-kilometre fragments to several hundred kilometres for the largest main-belt asteroids and dwarf planets. The major dynamical reservoirs from which these bodies are drawn (main belt asteroids and Trojans, asteroids that share Jupiter's orbit while librating around its leading and trailing Lagrange points, inside Jupiter's orbit; the Kuiper Belt and scattered disk just beyond Neptune; and the Oort cloud at $10^4$-$10^5$ AU) are surveyed in the rest of this lecture. Image credit: NASA/JPL-Caltech, public domain.
```

## Part 1: Meteorites, samples of the early solar system

### Why meteorites matter

A **meteorite** is a rocky or metallic fragment from interplanetary space that has survived atmospheric passage to reach the ground.
Its **parent body** is the object, usually an asteroid, that produced the fragment.
Meteorites are catalogued either as **falls**, recovered shortly after a witnessed fireball, or **finds**, collected after weathering on Earth {cite:p}`Krot2014`.
The majority are recovered from Antarctica and deserts, where dark stones are easy to find against pale sand or ice and chemical weathering is slow.

Meteorites provide the oldest rocks available to terrestrial laboratories.
Earth's geological record begins around $4.0$ Gyr ago and lunar rocks postdate Moon formation by hundreds of Myr, but primitive meteorites preserve grains that condensed directly from the protoplanetary disk.
Some primitive meteorites contain **presolar grains**, mineral grains older than the Sun that condensed in the ejecta of dying stars {cite:p}`Zinner2014`.

The oldest known solar system solids yield an absolute age of $4567.30 \pm 0.16$ Myr {cite:p}`Connelly2012`.
This date defines time zero ($t = 0$) against which to date planetesimal differentiation, core formation, and the giant impact that formed the Moon.

### A taxonomy of stones from space

**Chondrites** are primitive meteorites directly inherited from the protoplanetary disk.
Their defining components are **chondrules**: millimetre-scale silicate spheres crystallised from molten droplets in the disk.

```{figure} figures/allende.avif
:name: fig:l12-allende
:width: 480px
:align: center

A polished slab of the Allende carbonaceous chondrite (CV3), which fell in Mexico on 8 February 1969. Allende is rich in millimetre-scale chondrules (round, lighter inclusions) and bright Calcium-Aluminium-rich Inclusions (CAIs), the oldest known solids in the solar system. The matrix is dark, fine-grained, water-bearing material. Image credit: James St. John, CC BY 2.0 (Wikimedia Commons).
```

Other meteorites are **differentiated meteorites** that melted and segregated into crusts and mantles (**achondrites**), metallic cores (**iron meteorites**), and mixed silicate-metal boundaries (**stony irons**).

```{figure} figures/iron_widmanstatten.avif
:name: fig:l12-widmanstatten
:width: 480px
:align: center

Polished and acid-etched surface of an iron meteorite displaying the **Widmanstätten pattern**: an interlocking lattice of kamacite and taenite, two Fe-Ni phases. The pattern can only form if the solid metal cools at a rate slower than $\sim 100$ K per million years inside the metallic core of a parent body of asteroid size or larger. Image credit: H. Raab, CC BY-SA 3.0 (Wikimedia Commons).
```

The **Widmanstätten pattern** ({numref}`fig:l12-widmanstatten`) forms as Fe-Ni phases unmix during slow cooling ($< 100$ K per million years) in metallic cores, indicating parent bodies tens to hundreds of kilometres in radius {cite:p}`Goldstein2009`.

**Pallasites** ({numref}`fig:l12-pallasite`) are stony irons sampling the core-mantle boundary of a differentiated parent body.

```{figure} figures/pallasite.avif
:name: fig:l12-pallasite
:width: 480px
:align: center

Polished slice of the Esquel pallasite. Yellow-green olivine crystals are surrounded by a silvery Fe-Ni metallic matrix. Pallasites likely sample the core-mantle boundary of a differentiated planetesimal. Image credit: UCLA Meteorite Gallery, CC BY-SA 4.0 (Wikimedia Commons).
```

### The chondrite groups

Within the chondrites, three broad groupings dominate.
**Ordinary chondrites** are the most common falls (about 80% of observed falls), originating from dry, silicate-rich inner main belt asteroids and subdivided into H, L, and LL groups.
**Carbonaceous chondrites** are volatile-rich and chemically primitive, subdivided into groups including CI and CM.
**Enstatite chondrites** are the most reduced, dominated by metallic iron, suggesting formation in hot, low-oxygen disk regions close to the Sun.

CI chondrites occupy a special place in cosmochemistry because their non-volatile composition matches the solar photosphere {cite:p}`Lodders2003`.
Samples returned from Ryugu by Hayabusa2 in 2020 confirm that CI chondrites represent real, accessible asteroids {cite:p}`Yokoyama2023`.

```{figure} figures/chondrite_thin.avif
:name: fig:l12-chondrite-thin
:width: 480px
:align: center

Cross-polarised thin-section image of a chondrule in the NWA 5930 chondrite. The bright, interlocking crystals are olivine and pyroxene, frozen from a molten droplet that cooled in seconds to minutes inside the protoplanetary disk. Each chondrule is roughly a millimetre across. Image credit: H. Raab, CC BY-SA 3.0 (Wikimedia Commons).
```

### Chondrules and CAIs: the oldest solids

The two most diagnostic components of a chondrite are millimetre-scale objects that formed in the disk before incorporation into parent bodies.

**Calcium-Aluminium-rich Inclusions** (CAIs) are the oldest known solar system materials, consisting of refractory oxides and silicates of Ca, Al, Mg, and Ti that condensed from hot gas above $\sim 1400$ K.
They range from sub-millimetre aggregates to centimetre-scale lumps, appearing white or grey against the darker chondrite matrix ({numref}`fig:l12-allende`).
Three independent radiogenic chronometers ($^{238}\mathrm{U}$-$^{206}\mathrm{Pb}$, $^{235}\mathrm{U}$-$^{207}\mathrm{Pb}$, and $^{232}\mathrm{Th}$-$^{208}\mathrm{Pb}$) yield the same CAI age of $4567.30$ Myr to within $\pm 0.16$ Myr {cite:p}`Connelly2012`.
That age defines the zero point of the solar system timescale, against which all other events are dated.

**Chondrules** are roughly spherical beads of olivine, pyroxene, glass, and metal that crystallised from molten droplets ({numref}`fig:l12-chondrite-thin`), making up 60 to 80% of most chondrites by volume.
Their textures record peak temperatures of $\sim 1500$-$1900$ K and cooling rates from hundreds to thousands of K per hour over timescales of seconds to hours {cite:p}`HewinsConnolly2005`.

How chondrules formed remains an unsolved problem in cosmochemistry.
Candidate heating mechanisms include nebular shocks from gravitational instabilities, magnetic reconnection in current sheets, the X-wind model, and impact jetting between molten planetesimals {cite:p}`Connolly2016`.
Pb-Pb ages of individual chondrules run from the CAI age to $2.6$ Myr later, but primary chondrule formation was concentrated in the first million years after CAIs; the existing chondrules were then reworked over the lifetime of the disk {cite:p}`Connelly2012,Bollard2017`.
Chondrules therefore formed while the disk cooled and the first planetesimals grew.

{numref}`fig:chondrule-cai-timeline` places the two components on the time axis and sketches a chondrule's heating event.

<!-- Generated by scripts/figures/L12_small_bodies/fig_chondrule_cai_timeline.py -->
```{figure} figures/chondrule_cai_timeline.avif
:name: fig:chondrule-cai-timeline
:width: 100%
:align: center

(a) The oldest solids on the solar system time axis: CAIs at 4567.30 Myr (three Pb chronometers agree to within 0.16 Myr) define $t = 0$, and individual chondrules date from $t = 0$ to $2.6$ Myr. (b) A chondrule's heating event: peak temperatures of 1500 to 1900 K followed by cooling at hundreds to thousands of kelvin per hour, drawn as linear cooling from a 1700 K peak at 100 and at 1000 K per hour; CAIs condensed from gas above about 1400 K. Course-original figure.
```

### Isotopic dating of the early solar system

**Radioactive decay** provides the chronometer of planetary science.
A parent isotope decays to a daughter isotope at a rate proportional to its abundance:

$$
\dv{N}{t} = -\lambda N, \qquad N(t) = N_0 \exp(-\lambda t),
$$

where $\lambda$ is the decay constant and $t_{1/2} = \ln 2 / \lambda$ is the **half-life**.
Measuring parent and daughter abundances gives the time elapsed since cooling below the **closure temperature**, locking daughter atoms in minerals.

**Long-lived chronometers** have half-lives exceeding the age of the solar system and remain active today.
Primary systems include $^{238}\mathrm{U} \to {}^{206}\mathrm{Pb}$ ($t_{1/2} = 4.47$ Gyr), $^{235}\mathrm{U} \to {}^{207}\mathrm{Pb}$ ($t_{1/2} = 0.704$ Gyr), $^{87}\mathrm{Rb} \to {}^{87}\mathrm{Sr}$, and $^{147}\mathrm{Sm} \to {}^{143}\mathrm{Nd}$.
The Pb-Pb system is the most precise because its two parallel decay chains provide an internal cross-check.

**Short-lived chronometers** had half-lives of a few Myr or less and are extinct today.
Recorded only as daughter-isotope anomalies, they resolve processes on disk and planetesimal timescales:

| Parent | Daughter | $t_{1/2}$ (Myr) | What it dates |
|---|---|---|---|
| $^{26}\mathrm{Al}$ | $^{26}\mathrm{Mg}$ | 0.717 | CAI condensation, chondrule heating, planetesimal melting |
| $^{53}\mathrm{Mn}$ | $^{53}\mathrm{Cr}$ | 3.7 | Aqueous alteration, early planetesimal differentiation |
| $^{182}\mathrm{Hf}$ | $^{182}\mathrm{W}$ | 8.9 | Core formation in differentiated bodies |
| $^{129}\mathrm{I}$ | $^{129}\mathrm{Xe}$ | 16 | Volatile retention in chondrites |
| $^{244}\mathrm{Pu}$ | fission Xe | 80 | Old-rock retention of fissiogenic Xe |

Cross-calibrating long-lived absolute ages with short-lived relative ages yields a high-resolution timeline for the first 10 Myr.
Anchored to the Pb-Pb age of CAIs ($t = 0$), this dates planetesimal differentiation from $t \approx 0.5$ to $4$ Myr and protoplanet growth from $t \approx 5$ Myr onward {cite:p}`Kleine2009,Kruijer2017`.
Most chondrules date to $t \approx 1$ to $3$ Myr by this cross-calibration, though the earliest individual Pb-Pb ages extend to $t = 0$.
The rapid decay of two of these clocks, $^{26}\mathrm{Al}$ and $^{60}\mathrm{Fe}$, over the first 10 Myr of solar-system history is shown in {numref}`fig:l12-short-lived-decay`.

```{figure} figures/short_lived_decay.avif
:name: fig:l12-short-lived-decay
:width: 500px
:align: center

Abundance decay curves of the key short-lived radionuclides $^{26}\mathrm{Al}$ ($t_{1/2} = 0.72$ Myr; {cite:p}`CastilloRogez2009`) and $^{60}\mathrm{Fe}$ ($t_{1/2} = 2.62$ Myr; {cite:p}`Rugel2009`) over the first 10 Myr after CAI formation. Rapid decay of these extinct isotopes powered early planetesimal melting and differentiation. Course-original figure.
```

## Blackboard derivation: The Pb-Pb isochron age of CAIs

````{admonition} Blackboard derivation: The Pb-Pb isochron age of CAIs
:class: tip

We now derive the Pb-Pb isochron age of the oldest solar system solids.
This is the single most important number in this lecture and the cleanest illustration of how multiple radioactive systems can be combined to bypass model assumptions.

**Setup.**
Uranium has two long-lived isotopes that both decay to lead: $^{238}\mathrm{U}$ to $^{206}\mathrm{Pb}$ with decay constant $\lambda_{238} = \ln 2 / (4.4683 \text{ Gyr})$, and $^{235}\mathrm{U}$ to $^{207}\mathrm{Pb}$ with $\lambda_{235} = \ln 2 / (0.7038 \text{ Gyr})$.
Lead also has a primordial isotope, $^{204}\mathrm{Pb}$, that has no significant radiogenic source and is therefore a stable reference.
We will measure ratios relative to $^{204}\mathrm{Pb}$ throughout.

**Step 1: Radiogenic daughter from the parent measured today.**
For each U-Pb system, the present amount of radiogenic daughter is the parent that has decayed:

$$
{}^{206}\mathrm{Pb}^* = {}^{238}\mathrm{U} \, [\exp(\lambda_{238} t) - 1],
$$

$$
{}^{207}\mathrm{Pb}^* = {}^{235}\mathrm{U} \, [\exp(\lambda_{235} t) - 1].
$$

Here $t$ is the elapsed time since the system closed (the "age"), and the asterisk distinguishes radiogenic Pb produced *in situ* from any inherited initial Pb.

**Step 2: Real rocks carry initial Pb.**
A real rock contains both initial and radiogenic Pb.
Dividing each isotope by the stable reference $^{204}\mathrm{Pb}$, we can write the present-day measured ratios as

$$
\left( \frac{{}^{206}\mathrm{Pb}}{{}^{204}\mathrm{Pb}} \right)_{\!\text{now}}
\!=
\left( \frac{{}^{206}\mathrm{Pb}}{{}^{204}\mathrm{Pb}} \right)_{\!0}
\!+
\frac{{}^{238}\mathrm{U}}{{}^{204}\mathrm{Pb}} \, [\exp(\lambda_{238} t) - 1],
$$

$$
\left( \frac{{}^{207}\mathrm{Pb}}{{}^{204}\mathrm{Pb}} \right)_{\!\text{now}}
\!=
\left( \frac{{}^{207}\mathrm{Pb}}{{}^{204}\mathrm{Pb}} \right)_{\!0}
\!+
\frac{{}^{235}\mathrm{U}}{{}^{204}\mathrm{Pb}} \, [\exp(\lambda_{235} t) - 1],
$$

where the subscript $0$ denotes the *initial* (closure-time) ratio and "now" denotes the modern measurement.
These two equations contain three unknowns each: the elapsed time $t$, the initial Pb ratio, and the U/Pb ratio of the rock.
Knowing only one isotope ratio you cannot solve for $t$.

{numref}`fig:board-sketch-l12` shows where the argument is heading: several minerals from one rock fall on a straight line in the Pb-Pb plane, the line's slope gives the age, and its intercept region gives the initial lead.

<!-- Generated by scripts/figures/board_sketches/render_board_sketches.py -->
```{figure} figures/board_sketch_l12.avif
:name: fig:board-sketch-l12
:width: 45%
:align: center

Setup sketch of the board derivation of the Pb-Pb isochron age: measured $^{207}\mathrm{Pb}/^{204}\mathrm{Pb}$ against $^{206}\mathrm{Pb}/^{204}\mathrm{Pb}$ for several phases of one rock (orange points) fall on the isochron; its slope depends on the age $t$ alone, and the initial Pb composition sits at the low-ratio end. Course-original figure.
```

**Step 3: Subtract the initial ratios and divide.**
For each isotope, move the initial ratio to the left-hand side and divide by the U/Pb ratio, so that the time function stands alone on the right; for $^{206}\mathrm{Pb}$ this gives

$$
\frac{1}{{}^{238}\mathrm{U}/{}^{204}\mathrm{Pb}}
\left[ \left( \frac{{}^{206}\mathrm{Pb}}{{}^{204}\mathrm{Pb}} \right)_{\!\text{now}}
- \left( \frac{{}^{206}\mathrm{Pb}}{{}^{204}\mathrm{Pb}} \right)_{\!0} \right]
= \exp(\lambda_{238} t) - 1,
$$

and similarly for $^{207}\mathrm{Pb}$.
Now consider a *suite* of cogenetic samples, that is, several sub-samples of the same rock or several minerals that formed at the same time and from the same Pb reservoir.
They all share the same initial Pb ratios and the same age $t$.
What differs between samples is the U/Pb ratio, because U and Pb partition differently into different mineral phases (U is incompatible in olivine, for example, while Pb is incompatible in apatite).

Take the ratio of the two equations above for the same sample, dividing the $^{207}$Pb/$^{204}$Pb expression by the $^{206}$Pb/$^{204}$Pb expression.
The $^{204}$Pb terms cancel and the U-isotope ratio simplifies to the present-day $^{235}\mathrm{U}/{}^{238}\mathrm{U}$ value, leaving

$$
\frac{({}^{207}\mathrm{Pb}/{}^{204}\mathrm{Pb})_{\text{now}} - ({}^{207}\mathrm{Pb}/{}^{204}\mathrm{Pb})_0}
{({}^{206}\mathrm{Pb}/{}^{204}\mathrm{Pb})_{\text{now}} - ({}^{206}\mathrm{Pb}/{}^{204}\mathrm{Pb})_0}
=
\frac{{}^{235}\mathrm{U}}{{}^{238}\mathrm{U}} \cdot \frac{\exp(\lambda_{235} t) - 1}{\exp(\lambda_{238} t) - 1}.
$$

The right-hand side depends *only on time* (and on the present-day $^{235}\mathrm{U}/{}^{238}\mathrm{U}$ ratio of the rock, which is well-measured: about $1/137.8$ in CAIs once a small natural variability {cite:p}`Brennecka2010` is corrected for).
The left-hand side is the slope of a line in a plot of $({}^{207}\mathrm{Pb}/{}^{204}\mathrm{Pb})$ versus $({}^{206}\mathrm{Pb}/{}^{204}\mathrm{Pb})$: this is the **Pb-Pb isochron**.

**Why this is self-calibrating.**
For a suite of cogenetic samples that share an age and an initial Pb composition, all the measured points lie on a straight line.
The slope depends only on $t$, and the intercept gives the initial Pb ratio.
You do *not* need to assume an initial Pb composition before fitting the data: it falls out of the fit.
In contrast, a single U-Pb system (just $^{238}\mathrm{U}$-$^{206}\mathrm{Pb}$) requires an external estimate of the initial Pb to extract an age.
The double system removes that assumption; this is why the Pb-Pb method is the gold standard for absolute dating in cosmochemistry.

**Step 4: The CAI age.**
{cite:t}`Connelly2012` performed Pb-Pb dating on individual CAIs from the Efremovka CV3 chondrite (CAIs 22E, 31E, 32E), complemented by the Allende CAI SJ101, after acid leaching to remove non-radiogenic contamination.
The measured isochrons gave a CAI age of

$$
t_{\text{CAI}} = 4567.30 \pm 0.16 \text{ Myr}.
$$

The same work reported chondrule ages from Allende and NWA 5697 extending from about $4567.3$ Myr down to $4564.7$ Myr, leading {cite:t}`Connelly2012` to argue that chondrule formation began essentially contemporaneously with CAIs. The later Pb-Pb work of {cite:t}`Bollard2017` on 22 chondrules supports this early start: primary chondrule formation fell within the first million years after CAIs, and the existing chondrules were then reworked over the lifetime of the disk; most chondrules therefore carry short-lived-chronometer ages of $\sim 1$--$3$ Myr after CAIs.
The CAI age has been confirmed by independent samples and laboratories to within the same precision {cite:p}`Amelin2010`.
This is the absolute zero of the solar system clock ({numref}`fig:l12-pb-pb` shows the Pb-Pb isochrons for the dated CAIs and chondrules).

```{figure} figures/pb_pb_dating.avif
:name: fig:l12-pb-pb
:width: 445px
:align: center

Pb-Pb isochrons for (A) the Efremovka CAI 22E, (B) an Allende chondrule (C30), and (C) an NWA 5697 chondrule (C2). Each panel plots $^{207}\mathrm{Pb}/^{206}\mathrm{Pb}$ versus $^{204}\mathrm{Pb}/^{206}\mathrm{Pb}$ for cogenetic mineral fractions; on this inverse isochron the intercept at zero $^{204}\mathrm{Pb}/^{206}\mathrm{Pb}$ gives the radiogenic $^{207}\mathrm{Pb}/^{206}\mathrm{Pb}$ ratio, which yields the absolute age. The single-isochron age for CAI 22E, $4567.35 \pm 0.28$ Myr, is consistent with the weighted-mean CAI age $4567.30 \pm 0.16$ Myr derived from multiple CAIs in the main text. The Allende chondrule is contemporaneous with CAIs to within uncertainties, while the NWA chondrule is about 2.6 Myr younger. Figure adapted from {cite:t}`Connelly2012`.
```

````

### Petrographic and shock metamorphism

The **petrographic type** of a chondrite, on a scale from 1 to 7, records how much its parent body altered it after accretion.
By convention, type 3 is the least altered.
Types 2 and 1 record progressively heavier **aqueous alteration**, where low-temperature reactions with water form clays and erase original chondrules.
Types 4 through 7 record increasing **thermal metamorphism**, where higher temperatures coarsen mineral grains, equilibrate compositions, and destroy chondrule outlines.

The **shock stage**, denoted S1 (unshocked) through S6 (heavily shocked, partially melted), records impact history {cite:p}`Stoffler1991`.
Shock features include planar fractures in olivine and pyroxene crystals, dark melt veins, and high-pressure minerals such as ringwoodite and majorite that form at pressures of tens of GPa.

{numref}`fig:petrologic-types` lays out both scales.

<!-- Generated by scripts/figures/L12_small_bodies/fig_petrologic_types.py -->
```{figure} figures/petrologic_types.avif
:name: fig:petrologic-types
:width: 100%
:align: center

The two scales that record what a parent body did to a chondrite. Petrographic type 3 is the least altered; types 2 and 1 record increasing aqueous alteration, in which water forms clays and erases chondrules, and types 4 to 7 increasing thermal metamorphism, in which grains coarsen, compositions equilibrate and chondrule outlines vanish. The shock stage runs from S1 (unshocked) to S6 (heavily shocked, partially melted), with planar fractures, melt veins and high-pressure minerals such as ringwoodite and majorite formed at tens of GPa. Course-original figure.
```

### Differentiated bodies, lunar and Martian meteorites

Achondrites, irons, and stony irons all come from parent bodies that experienced bulk melting.
The primary heat source was the decay of $^{26}\mathrm{Al}$, with a half-life of 0.7 Myr and an initial abundance of about $5 \times 10^{-5}$ relative to stable $^{27}\mathrm{Al}$.
This delivered enough heat to melt planetesimals larger than $\sim 20$ km formed at the start of the CAI epoch, whereas bodies accreting after $\sim 1.5$ Myr never melted and remained chondrites {cite:p}`Hevey2006`.

The most important achondrite group are the **HED meteorites** (howardites, eucrites, and diogenites), which make up $\sim 6\%$ of falls and share a unique parent, the asteroid (4) Vesta.
The Dawn mission confirmed this link ({numref}`fig:l12-vesta`), showing that the giant Rheasilvia impact basin excavated deep layers and likely ejected the HED meteorites into space {cite:p}`Russell2012`.

```{figure} figures/vesta.avif
:name: fig:l12-vesta
:width: 480px
:align: center

Mosaic of (4) Vesta from NASA's Dawn mission, taken in 2011--2012. Vesta is the second-largest asteroid in the main belt and the parent body of the HED meteorite suite. The southern hemisphere is dominated by the giant Rheasilvia impact basin (visible at lower right), which is responsible for delivering Vesta-derived material to Earth as meteorites. Image credit: NASA/JPL-Caltech/UCLA/MPS/DLR/IDA, public domain.
```

In addition to asteroidal achondrites, two groups come from larger planetary bodies:

- **Lunar meteorites** match Apollo and Luna sample compositions and sample regions of the Moon never visited by spacecraft.
- **Martian meteorites** (the SNC group) contain trapped atmospheric noble gases whose composition matches measurements in Mars's atmosphere {cite:p}`Bogard1983`.

Both groups are launched when large impacts eject material above the parent body escape velocity, eventually reaching Earth.

### Oxygen isotopes as a parent-body fingerprint

The **oxygen three-isotope plot** ($\delta^{17}\mathrm{O}$ versus $\delta^{18}\mathrm{O}$) is the primary tool for identifying meteorite parent bodies {cite:p}`Clayton1973`.
Oxygen has three stable isotopes: $^{16}\mathrm{O}$, $^{17}\mathrm{O}$, and $^{18}\mathrm{O}$.
On Earth, natural processes operate by **mass-dependent fractionation**, where heavier isotopes react slightly slower and fractionation in $^{17}\mathrm{O}/^{16}\mathrm{O}$ is half that in $^{18}\mathrm{O}/^{16}\mathrm{O}$.
All terrestrial samples therefore lie on a single line of slope $1/2$.

Meteorites do not follow this terrestrial trend.
In CAIs, oxygen isotopes define a slope-1 line displaced from the terrestrial line, indicating mixing with a nearly pure $^{16}\mathrm{O}$ reservoir {cite:p}`Yurimoto2004`.
Each meteorite group occupies a distinct cluster: ordinary chondrites plot above the terrestrial line, carbonaceous chondrites plot below it, enstatite chondrites and lunar samples lie on the terrestrial line, and the HED and SNC groups plot on lines consistent with Vesta and Mars {cite:p}`Greenwood2017`.

This isotopic clustering identifies parent bodies of unknown meteorites and classifies returned samples.
Hayabusa2 samples from Ryugu plot with CI chondrites, whereas OSIRIS-REx samples from Bennu fall near ungrouped carbonaceous chondrites while sharing CI-like bulk composition {cite:p}`Yokoyama2023,Lauretta2024`.

{numref}`fig:oxygen-three-isotope` sketches the diagram.

<!-- Generated by scripts/figures/L12_small_bodies/fig_oxygen_three_isotope.py -->
```{figure} figures/oxygen_three_isotope.avif
:name: fig:oxygen-three-isotope
:width: 100%
:align: center

Schematic oxygen three-isotope plot, $\delta^{17}$O against $\delta^{18}$O. Mass-dependent fractionation puts every terrestrial sample, the Moon and the enstatite chondrites on the terrestrial fractionation line of slope 1/2; CAIs define a slope-1 line that records mixing with a nearly pure $^{16}$O reservoir. Ordinary chondrites plot above the terrestrial line, carbonaceous chondrites below it, CI chondrites together with the Ryugu samples, and the HED and SNC groups on their own parallel lines for Vesta and Mars. Cluster positions are schematic; the slopes and the ordering above and below the terrestrial line follow the text. Course-original figure.
```

### The NC-CC isotopic dichotomy: three competing interpretations

**Nucleosynthetic anomalies** split meteorites into non-carbonaceous (**NC**) and carbonaceous (**CC**) reservoirs that did not mix for at least the first $\sim 3$–$4$ Myr {cite:p}`Warren2011,Kruijer2017`.

Three competing physical interpretations explain this dichotomy ({numref}`fig:l12-spitzer-ages`).

```{figure} figures/spitzer_nc_cc_ages.avif
:name: fig:l12-spitzer-ages
:width: 500px
:align: center

Hf-W metal-silicate constraints show NC core formation at $\Delta t_{\text{CAI}} = 1.0$ to $2.6$ Myr ($\varepsilon^{182}\mathrm{W} = -3.37$ to $-3.20$) and CC core formation at $3.0$ to $3.6$ Myr ($\varepsilon^{182}\mathrm{W} = -3.16$ to $-3.10$). The separation of roughly $2$ Myr between the two reservoirs is the key constraint that all three interpretations must satisfy (from Figure 5 of {cite:t}`SpitzerPt2021`).
```

```{figure} figures/lichtenberg2021_fig1.avif
:name: fig:l12-lichtenberg-fig1
:width: 600px
:align: center

Planetesimal formation rate $\dd \Sigma_{\text{plts}} / \dd t$ across orbital distance $r$ and time $t$ after CAI formation, where outward snow-line migration produces early dry NC bodies (Reservoir I) and inward migration forms later ice-rich CC bodies (Reservoir II). This bifurcation arises from disk thermal evolution without requiring an early formed Jupiter barrier (from Figure 1 of {cite:t}`Lichtenberg2021`).
```

Three competing hypotheses explain the physical cause of this bifurcation:

- A **spatial** cause {cite:p}`Kruijer2017`: a $\sim 20\,\Mearth$ early Jupiter core opens a disk gap, blocking inward drift of CC pebbles ({ref}`Lecture 2 <lecture02>`).
- A **dynamical** cause {cite:p}`Lichtenberg2021`: snow-line migration produces two distinct planetesimal reservoirs ({numref}`fig:l12-lichtenberg-fig1`), with Jupiter halting pebble drift once it reaches the pebble isolation mass.
- A **temporal** cause {cite:p}`Schiller2018,Spitzer2020`: the composition of the inner-disk solids changed rapidly with time during the infall from the parent molecular cloud, so planetesimals that formed at different times locked in different compositions.

## Part 2: Small body populations and dynamics

We move now from individual rocks to entire populations.
Where in the solar system do small bodies live, what shapes their orbits, and what does the present distribution tell us about the dynamical history of the planets?

### The asteroid belt: structure and orbital dynamics

The **asteroid belt** is the largest reservoir of small bodies inside Neptune's orbit, situated between $a = 2.1$ AU and $a = 3.3$ AU with $e < 0.3$ ({numref}`fig:l12-asteroid-orbits`).
Its total mass is approximately $4 \times 10^{-4}\,\Mearth$, of which Ceres contributes nearly $40\%$ {cite:p}`Bottke2020`.

```{figure} figures/demeo_carry_belt_inclination.avif
:name: fig:l12-asteroid-orbits
:width: 700px
:align: center

The asteroid belt in context: orbital inclination as a function of heliocentric distance for catalogued asteroids. Each point is a single body and the colour intensity scales with the local number density. Yellow points show the high-density regions of the main belt between $\sim 2.1$ and $\sim 3.3$ AU, with the Hungarias and Hildas labelled and the Jupiter Trojans visible at $5.2$ AU. The vertical Kirkwood resonances at $2.5$, $2.8$, and $3.3$ AU appear as gaps in the density. From Figure 1 of {cite:t}`DeMeo2014`.
```

The **Kirkwood gaps** are depletions in semimajor axis where mean-motion resonances with Jupiter pump eccentricities to planet-crossing values over $10^4$-$10^6$ years.
These dynamical sinks continuously supply near-Earth asteroids {cite:p}`Wisdom1985` ({numref}`fig:l12-kirkwood`).

```{figure} figures/granvik_kirkwood_resonances.avif
:name: fig:l12-kirkwood
:width: 700px
:align: center

Steady-state orbit distributions of the near-Earth objects that leave the main belt through the three principal Jupiter mean-motion resonances: the 3:1J complex at $a \approx 2.50$ AU (left), the 5:2J complex at $a \approx 2.82$ AU (middle), and the 2:1J complex at $a \approx 3.27$ AU (right). Each column holds two panels for one escape route: inclination against semimajor axis on top, eccentricity against semimajor axis below. The colour scale is the fraction of the steady-state population in each bin, from black (empty) through purple and orange to yellow (highest). The yellow concentration in each column is close to the semimajor axis of its own resonance, and the population spreads to eccentricities above $0.5$ and to inclinations of several tens of degrees, which is what makes these orbits planet-crossing. Assembled from the steady-state column of Figure 5 of {cite:t}`Granvik2018`.
```

### Asteroid families and spectral taxonomy

**Asteroid families** are clusters of asteroids sharing similar **proper elements** (orbital elements averaged over secular variations), representing collisional debris from a disrupted parent body.
Their size distribution and orbital spread constrain parent body sizes, impact energies, and disruption ages {cite:p}`Nesvorny2015` ({numref}`fig:l12-asteroid-families`).

```{figure} figures/asteroid_families_nesvorny.avif
:name: fig:l12-asteroid-families
:width: 700px
:align: center

Asteroid families in the main belt revealed by hierarchical clustering, plotted in proper inclination $i_{\text{P}}$ versus proper semimajor axis $a_{\text{P}}$. The principal Hirayama families (Themis, Eos, Koronis, Eunomia, Vesta, Flora) appear as concentrations in $(a_{\text{P}}, i_{\text{P}})$ space (from Figure 1a of {cite:t}`Nesvorny2015`).
```

Asteroids are classified into **spectral types** based on visible and near-infrared reflectance.
Major classes include carbonaceous **C-types** (carbonaceous chondrite analogues), silicaceous **S-types** (ordinary chondrite analogues with 1 and 2 $\mu$m absorptions), metallic **M-types**, basaltic **V-types**, and primitive **D- and P-types**.

Rocky S-types dominate the inner belt inside the snow line, whereas volatile-rich C-types dominate the outer belt ({numref}`fig:l12-demeo-mass`) {cite:p}`DeMeo2014`.

```{figure} figures/demeo_carry_belt_mass.avif
:name: fig:l12-demeo-mass
:width: 700px
:align: center

Compositional mass distribution across the asteroid belt in $0.02$ AU bins from the Hungarias to the Trojans, showing total mass (grey) and the spectral classes; the horizontal line at $10^{18}$ kg is the limit of the 1980s surveys, and below it the modern data resolve the individual classes. Silicaceous S-types dominate the inner belt while carbonaceous C-, B-, P-, and D-types dominate the outer belt, with V-type material concentrated near $a = 2.4$ AU, reflecting the disk volatile gradient (from Figure 3 of {cite:t}`DeMeo2014`).
```

This taxonomic structure records giant-planet migration in the early solar system, which implanted carbonaceous bodies into the outer belt and reshuffled the asteroid belt {cite:p}`Walsh2011,Raymond2017` ({ref}`Lecture 2 <lecture02>`).

### Near-Earth asteroids and the impact hazard

A **near-Earth asteroid** (NEA) is one with perihelion $q < 1.3$ AU.
Because the typical NEA residence time is only $\sim 10$ Myr before ejection or collision, the steady-state population requires continuous resupply {cite:p}`Bottke2002`.

The main asteroid belt supplies new NEAs through Kirkwood gaps (particularly the 3:1 and 5:2 resonances), the $\nu_6$ **secular resonance** (a matching of orbital precession rates where an asteroid's apsidal precession matches Saturn's), and the Yarkovsky effect.

About 2,500 NEAs are classified as **Potentially Hazardous Asteroids** (PHAs): larger than $\sim 140$ m and approaching Earth's orbit within $0.05$ AU.

The **Yarkovsky effect** is a non-gravitational force caused by asymmetric thermal re-emission of absorbed sunlight on a rotating body.
For a 1 km basaltic main-belt asteroid ($\rho \sim 2.5$ g cm$^{-3}$, $a \sim 2$ AU), $\mathrm{d}a/\mathrm{d}t \sim 10^{-4}$ AU per Myr, so a drift of $0.01$ AU into a nearby Kirkwood resonance takes $\sim 10^8$ years.
The resulting recoil slowly changes the semimajor axis, drifting asteroids over $10^8$ years into resonances that pump eccentricities onto planet-crossing orbits {cite:p}`Bottke2006`.

Because recoil force scales with absorbed solar power while acceleration scales inversely with mass $m \propto \rho R^3$, the secular semimajor-axis drift rate scales as

$$
\dv{a}{t} \propto \frac{1}{D \, \rho \, \sqrt{a}},
$$

so a smaller, less dense, closer-in body drifts faster {cite:p}`Bottke2006`.
This $1/D$ supply scaling explains why smaller NEAs dominate the population, with drift directly measured on (6489) Golevka ({numref}`fig:l12-yarkovsky`).

```{figure} figures/yarkovsky_detection_vokrouhlicky.avif
:name: fig:l12-yarkovsky
:width: 500px
:align: center

Direct detection of the Yarkovsky effect on the near-Earth asteroid (6489) Golevka. The plot shows the orbital solution in range vs range-rate offset (relative to a fit using only gravitational perturbations), projected into the plane of radar observables. The grey ellipse labelled "pure gravity" represents the 90% confidence region for the orbital solution if Yarkovsky is excluded; the grey ellipse labelled "with Yarko" is the predicted solution including the nominal Yarkovsky force. The Arecibo measurements of May 2003 (black symbol with $\sim 5$ mm/s uncertainty) fall squarely on the Yarkovsky-included prediction. From Figure 2 of {cite:t}`Vokrouhlicky2015`.
```

The **YORP effect** (Yarkovsky-O'Keefe-Radzievskii-Paddack) is a thermal recoil torque from asymmetric radiation that modifies the rotation rate and spin axis of bodies smaller than $\sim 10$ km, driving them to rotational fission or slowing them to near zero.
This spin-up has been directly detected on the near-Earth asteroid (54509) YORP ({numref}`fig:l12-yorp`).

```{figure} figures/yorp_detection_vokrouhlicky.avif
:name: fig:l12-yorp
:width: 500px
:align: center

Direct detection of the YORP effect on the small near-Earth asteroid (54509) YORP. The vertical axis shows the additional sidereal rotation phase (in degrees) accumulated relative to a constant-rotation model; the horizontal axis is time in days since 27 July 2001. Black points are independent measurements from successive radar and optical apparitions; the grey curve is a quadratic fit corresponding to a rotational acceleration $\dd \omega / \dd t \simeq 350 \times 10^{-8}$ rad d$^{-2}$. The accelerating spin is the predicted YORP signature: an asymmetric thermal recoil torque acting on an irregular rotating body. Adapted from Figure 5 of {cite:t}`Vokrouhlicky2015`.
```

### Impact frequency and planetary defence

The impact rate follows a power law with exponent $-2.3$ ({numref}`fig:l12-impact-freq`); the $10$ km Chicxulub impactor that ended the Cretaceous at $66$ Ma is the largest well-dated example (cratering physics in {ref}`Lecture 7 <lecture07>`).

```{figure} figures/neo_sfd_schunova.avif
:name: fig:l12-impact-freq
:width: 600px
:align: center

Cumulative size-frequency distribution of near-Earth objects versus absolute magnitude $H_V$ ({cite:t}`Schunova2017`, Fig. 9). The power-law distribution spans five orders of magnitude, consistent with models from {cite:t}`Bottke2002` and {cite:t}`Granvik2018`.
```

Impact rates scale inversely with size: 10 to 80 m airbursts occur every decade to few millennia (e.g., Chelyabinsk, {numref}`fig:l12-chelyabinsk`, {cite:p}`Brown2013`, and Tunguska), 140 m impacts every $\sim 30{,}000$ years, 1 km disruptions every $\sim 500{,}000$ years, and 10 km extinctions every $\sim 100$ Myr.

```{figure} figures/chelyabinsk.avif
:name: fig:l12-chelyabinsk
:width: 480px
:align: center

Vapour trail of the 15 February 2013 Chelyabinsk meteor, formed when a 19 m asteroid disintegrated at $\sim 30$ km altitude releasing 0.5 Mt TNT equivalent. The shock wave shattered windows and injured about 1,500 people (image credit: Alex Alishevskikh, CC BY-SA 2.0).
```

In 2022, NASA's DART mission demonstrated **planetary defence** by deflecting Dimorphos, shortening its orbital period by $33.0 \pm 1.0$ minutes ({numref}`fig:l12-dimorphos`) {cite:p}`Thomas2023`, with ejecta recoil giving momentum enhancement $\beta = 3.61^{+0.19}_{-0.25}$ {cite:p}`Cheng2023`.

Equating asteroid gravity to differential solar tides defines the stable **Hill sphere** and yields the Hill radius

$$
r_H = a \left( \frac{m}{3 \, \Msun} \right)^{1/3},
$$

For Didymos ($a \approx 1.64$ AU, mass $m \approx 5.6 \times 10^{11}$ kg), the Hill radius is $r_H \approx 110$ km {cite:p}`Daly2023`.
Dimorphos orbits at $\sim 1.2$ km, well inside the stable limit of $0.5 \, r_H$.

```{figure} figures/dart_lightcurve_daly.avif
:name: fig:l12-dimorphos
:width: 700px
:align: center

Photometric lightcurves of Didymos-Dimorphos six days after DART impact, showing Didymos's 2.26 hr rotation (top) and Dimorphos's new $11.372$ hr orbit (bottom). Kinetic impact reduced the orbital period by $33.0 \pm 1.0$ minutes ($3\sigma$) via ejecta momentum enhancement ({cite:t}`Thomas2023`, Fig. 3).
```

ESA's Hera mission (launched 2024) will survey the crater to refine momentum-transfer efficiency, making planetary defence a quantitative engineering discipline {cite:p}`Michel2022`.

### Ceres and the dwarf planets of the inner solar system

Ceres is the largest body in the asteroid belt and the only inner solar system **dwarf planet**, a near-spherical body with a mean radius of $470$ km and a mass of $9.4 \times 10^{20}$ kg.

```{figure} figures/ceres_occator.avif
:name: fig:l12-occator
:width: 320px
:align: center

The Occator crater on Ceres, imaged by NASA's Dawn mission. The bright deposits in the crater floor (Cerealia Facula and Vinalia Faculae) are evaporitic salts, primarily sodium carbonate ($\mathrm{Na_2CO_3}$), interpreted as relict brine that erupted from a subsurface reservoir and froze. Image credit: NASA/JPL-Caltech/UCLA/MPS/DLR/IDA, public domain.
```

In Occator crater ({numref}`fig:l12-occator`), bright evaporitic salts of sodium carbonate and ammonium chloride formed when subsurface brine erupted and froze {cite:p}`DeSanctis2016`.
This indicates that Ceres has, or had until recently, a partially liquid water layer at depth and probably an internal heat source, making it similar to icy moons such as Enceladus and Europa ({ref}`Lecture 11 <lecture11>`).

Its surface ammonia and high water content suggest Ceres formed beyond the snow line and was captured into the asteroid belt during the Nice-Model rearrangement {cite:p}`McKinnon2008`.
This is one of the more direct pieces of evidence that the asteroid belt is a mixture of inner-disk and outer-disk material.

### The Kuiper Belt and scattered disk

Beyond Neptune lies the **trans-Neptunian region**, a reservoir of small bodies with distinct dynamical sub-populations ({numref}`fig:l12-kuiper-orbits`) {cite:p}`Bannister2018`.

The **cold classical Kuiper Belt** consists of unscattered primordial orbits between 42 and 47 AU {cite:p}`Parker2011`.
The **hot classical Kuiper Belt** has inclinations up to $30^{\circ}$ excited by giant-planet instability.
The **resonant population** is locked in mean-motion resonances, dominated by **plutinos** at $a \approx 39.4$ AU.
The **scattered disk** has perihelia near Neptune ($q \sim 30$ AU), while **detached objects** have decoupled perihelia ($q > 40$ AU).

```{figure} figures/bannister_ossos_orbits.avif
:name: fig:l12-kuiper-orbits
:width: 700px
:align: center

Orbital parameters of the 1142 characterised trans-Neptunian objects discovered by the Outer Solar System Origins Survey (OSSOS) and affiliated programmes (CFEPS, MiLat, AlexWP), classified into Jupiter-coupled (dark blue), centaurs (cyan), classical (light blue), resonant (red), scattering (orange), and detached (dark red) populations. Top: orbital inclination versus heliocentric distance. Middle: orbital inclination versus semimajor axis. Bottom: eccentricity versus semimajor axis. Pale blue vertical lines mark the locations of the principal Neptune mean-motion resonances; the cold classical population is the dense low-inclination cluster between 42 and 47 AU. From Figure 4 of {cite:t}`Bannister2018`.
```

```{figure} figures/petit_classical_kbo.avif
:name: fig:l12-kuiper-inclination
:width: 600px
:align: center

The three components of the CFEPS-L7 synthetic model of the main classical Kuiper belt, separated into the broad-inclination "hot" component (top left), the dynamically active "stirred" component (top right), the dense low-inclination "kernel" near $a \approx 44$ AU (bottom left), and the combined model (bottom right). Each panel shows orbital eccentricity $e$ vs semimajor axis $a$ (upper sub-panel) and inclination $i$ vs $a$ (lower sub-panel). The kernel and the broader hot component coexist in the same range of $a$ but occupy completely different parts of $(e, i)$ space, evidence that the cold classicals were never strongly perturbed by Neptune while the hot classicals were excited by the giant-planet instability described by the Nice Model. From Figure 4 of {cite:t}`Petit2011`.
```

The bimodal inclination distribution, resonant populations, and scattered disk provide empirical evidence for giant-planet migration under the Nice Model ({ref}`Lecture 2 <lecture02>`) ({numref}`fig:l12-kuiper-inclination`).

### Pluto: a dwarf planet visited

The July 2015 New Horizons flyby revealed Pluto as a geologically diverse dwarf planet {cite:p}`Stern2015` ({numref}`fig:l12-pluto-color`).

```{figure} figures/pluto_color_stern.avif
:name: fig:l12-pluto-color
:width: 500px
:align: center

True-colour Ralph instrument image of Pluto acquired by New Horizons during the 2015 flyby, showing the bright heart-shaped Tombaugh Regio and dark equatorial band Cthulhu Macula rich in **tholins**. From Figure 3 of {cite:t}`Stern2015`.
```

Pluto's density of $1860 \pm 13$ kg m$^{-3}$ (radius $1187 \pm 4$ km, mass $1.303 \times 10^{22}$ kg) implies roughly two-thirds rock and one-third water ice {cite:p}`Stern2015`.

The most prominent feature is **Sputnik Planitia**, a $\sim 1{,}000$ km wide nitrogen-ice basin where polygonal cells overturn every $\sim 10^{6}$ years via solid-state convection powered by radiogenic decay in the rocky interior {cite:p}`McKinnon2016` ({numref}`fig:l12-sputnik-convection`).
The absence of craters indicates a surface age under 10 Myr.

```{figure} figures/sputnik_convection_mckinnon.avif
:name: fig:l12-sputnik-convection
:width: 700px
:align: center

Numerical model of solid-state convection in a layer of $\mathrm{N_2}$ ice in Sputnik Planitia, driven at Rayleigh number $\mathrm{Ra}_b \approx 3 \times 10^5$ with overturn timescales of $\sim 10^6$ years. Panels show temperature, surface velocity, dynamic topography, and heat flow across one convective cell (from Figure 4 of {cite:t}`McKinnon2016`).
```

A positive gravity anomaly at Sputnik Planitia suggests a **subsurface ocean**, a liquid layer between the rocky core and ice shell {cite:p}`Nimmo2016`.

Pluto has a thin nitrogen atmosphere (around 10 $\mu$bar) undergoing slow escape, closer to Jeans escape than to hydrodynamic escape because the upper atmosphere is so cold {cite:p}`Gladstone2016`.

### Charon and the small moons

Charon (radius $606$ km) and Pluto form a binary: their barycentre lies outside Pluto, the only such case among planet-moon systems in the solar system.
Its polar cap, **Mordor Macula**, is interpreted as photochemically processed methane that escaped Pluto's atmosphere and froze onto Charon {cite:p}`Grundy2016` ({numref}`fig:l12-mordor`) ({numref}`fig:l12-pluto-charon`).

```{figure} figures/charon_mordor.avif
:name: fig:l12-mordor
:width: 480px
:align: center

Charon imaged by New Horizons. The dark reddish region at the north pole is **Mordor Macula**, an accumulation of photochemically processed organic compounds (tholins) sourced from Pluto's escaping atmosphere. The chasm visible across the equator is part of an extensional tectonic system. Image credit: NASA/JHUAPL/SwRI, public domain.
```

```{figure} figures/pluto_charon_features_stern.avif
:name: fig:l12-pluto-charon
:width: 700px
:align: center

Cylindrical projections of Pluto (top) and Charon (bottom) with informally named features overlaid, derived from New Horizons imaging during the 2015 flyby. The colour-coded regions on Pluto separate the named maculae, regiones, planitiae, terrae, and montes; the colour-coded regions on Charon highlight the polar Mordor Macula and the equatorial chasms (Argo, Serenity, Macross). The Pluto-Charon system is a true binary whose barycentre lies outside Pluto. From Figure 2 of {cite:t}`Stern2015`.
```

The four small moons (Styx, Nix, Kerberos, Hydra) are likely fragments of the same collision that produced Charon {cite:p}`Canup2011`.

### Arrokoth: a pristine cold classical KBO

The New Horizons flyby of (486958) Arrokoth at $44.6$ AU revealed a **contact binary**: two lobes joined at a narrow neck by a gentle merger under a few m s$^{-1}$ {cite:p}`Stern2019`.

This morphology indicates formation by the **streaming instability**, the gravitational collapse of a pebble swarm into a bound object.
Arrokoth shows that $\sim 35$ km planetesimals form by gentle gravitational collapse rather than hierarchical accretion {cite:p}`McKinnon2020` ({numref}`fig:l12-arrokoth`).

```{figure} figures/arrokoth_color_stern.avif
:name: fig:l12-arrokoth
:width: 500px
:align: center

Enhanced colour image of (486958) Arrokoth (then informally "Ultima Thule") at 1.5 km/pixel resolution, taken by New Horizons during the 1 January 2019 flyby. The two flattened lobes ("Ultima" and "Thule") are joined at a narrow neck and share a uniform reddish colour, consistent with formation from a single locally collapsing pebble cloud. The contact-binary morphology and the absence of a high-energy impact crater at the neck are signatures of a gentle low-velocity merger ($\lesssim$ few m s$^{-1}$). From Figure 2 of {cite:t}`Stern2019`.
```

### The other dwarf planets

The IAU recognises three other trans-Neptunian dwarf planets: **Eris**, a scattered-disk object ({numref}`fig:l12-eris`); **Haumea**, a fast-rotating ellipsoid with rings; and **Makemake**, a classical KBO.

Candidate dwarf planets include Gonggong, Quaoar, Orcus and Salacia.
**Sedna** belongs to a different class: its detached orbit (perihelion 76 AU, aphelion about 900 AU) may point to an inner Oort cloud population or to an early stellar encounter, and it is one of a handful of extreme trans-Neptunian objects whose orbits appear to cluster in orientation ({numref}`fig:l12-planet-nine`), the observation that motivates the **Planet 9** hypothesis of an undiscovered distant planet {cite:p}`Batygin2019`.

```{figure} figures/sedna_orbits_batygin.avif
:name: fig:l12-planet-nine
:width: 80%
:align: center

Orbits of the distant trans-Neptunian objects, including Sedna, projected on the ecliptic (main panel; the bar marks 250 AU), with arrows giving the direction of each perihelion and the inset showing the orbit-pole (angular momentum) vectors. The perihelion directions and the poles are not spread evenly around the sky, which is the apparent clustering that motivates the Planet 9 hypothesis. Reproduced from {cite:t}`Batygin2019`.
```

```{figure} figures/eris.avif
:name: fig:l12-eris
:width: 360px
:align: center

Eris and its moon Dysnomia imaged by the Hubble Space Telescope. Eris is approximately the same size as Pluto but more massive, hence denser. Its discovery in 2005 directly precipitated the IAU's 2006 planet definition. Image credit: NASA/ESA/M. Brown, public domain.
```

{numref}`fig:l12-haumea` shows the occultation light curve that revealed the ring of Haumea.

```{figure} figures/haumea_ring.avif
:name: fig:l12-haumea
:width: 480px
:align: center

Stellar occultation light curve of the dwarf planet Haumea on 21 January 2017, recorded with the 1 m telescope at Konkoly Observatory (Hungary). The deep central drop is the body of Haumea; the two narrower symmetric dips on either side reveal a $\sim 70$ km wide, $\sim 2{,}287$ km radius ring, the first ring system discovered around a trans-Neptunian object {cite:p}`Ortiz2017`. Figure adapted from {cite:t}`Sicardy2024`.
```
### The Oort cloud

The **Oort cloud** is the most distant component of the solar system: a roughly spherical shell of icy bodies at heliocentric distances of approximately $2{,}000$ to $50{,}000$ AU.
Jan Oort inferred its existence in 1950 from the orbital statistics of long-period comets, whose orbital energies concentrate at nearly bound values ({numref}`fig:l12-oort-schematic`) {cite:p}`Oort1950`.
Incoming comets arrive isotropically from all directions with semimajor axes clustering near $20{,}000$ to $50{,}000$ AU, implying a steady-state reservoir of $\sim 10^{11}$ to $10^{12}$ objects with a total mass of $1$ to $10\,\Mearth$ {cite:p}`Brasser2013`.

```{figure} figures/oort_cloud_kaib.avif
:name: fig:l12-oort-schematic
:width: 600px
:align: center

Final structure of the simulated Oort cloud and scattered disk after 4.5 Gyr of evolution from the solar system's birth in an embedded star cluster. Top row: perihelion $q$ (AU) vs semimajor axis $a$ for two simulations of Oort cloud formation. Bottom row: orbital inclination vs $a$ for the same two simulations. The transition from the planetary region (small $a$) through the scattered disk and into the inner ($a \sim 10^{3}$ AU) and outer ($a \sim 10^{4}$ AU) Oort cloud is visible as a continuous distribution. Inner-cloud bodies retain a wide range of inclinations imposed by the cluster phase, while outer-cloud bodies are nearly isotropic. The Oort cloud has never been directly observed; its existence is inferred from this kind of dynamical model and from the orbital statistics of long-period comets. From Figure 13 of {cite:t}`Kaib2008`.
```

Giant planets scattered planetesimals between Jupiter and Neptune onto eccentric orbits, where passing stars and the galactic tide lifted their perihelia into the outer cloud {cite:p}`Dones2004`.

### Comets: composition, structure, and activity

A **comet** is a small icy body that becomes active when its perihelion brings it close enough to the Sun for surface ices to sublimate.
Whipple's "dirty snowball" model {cite:p}`Whipple1950` describes a comet as a mixture of water ice, refractory dust, and organic molecules.
Measured comets exhibit high porosity ($50$-$75\%$) and a low bulk density of order $0.5$ g cm$^{-3}$.

```{figure} figures/comet_diagram.svg
:name: fig:l12-comet-anatomy
:width: 600px
:align: center

Schematic anatomy of an active comet. The solid **nucleus** (1--30 km across) is surrounded by an extended **coma** of sublimated gas and dust. Two distinct tails point in different directions: the **ion tail** (blue) is shaped by the solar wind and points anti-sunward, while the **dust tail** (yellow) is shaped by radiation pressure on dust particles and lags behind the nucleus along its orbit. Image credit: Wikimedia Commons, public domain.
```

Far from the Sun, a comet is an inert **nucleus** (a dark body 1--30 km across with albedo $\sim 4\%$).
Inside roughly 5 AU, ice sublimation creates an extended **coma** of gas and dust ($10^4$-$10^6$ km across), an **ion tail** swept anti-sunward by the solar wind, and a **dust tail** pushed by radiation pressure ({numref}`fig:l12-comet-anatomy`).

Halley's Comet ({numref}`fig:l12-halley`) is the prototypical example with a 76-year orbital period ({numref}`fig:l12-halley1986`).
Its 1986 apparition revealed an irregular $15 \times 8 \times 8$ km nucleus whose active jets are confined to small surface fractions.

```{figure} figures/halley.avif
:name: fig:l12-halley
:width: 500px
:align: center

Comet 1P/Halley imaged by the Giotto Halley Multicolour Camera (HMC) on 13-14 March 1986, the first close-up view of a cometary nucleus. The nucleus is approximately $15 \times 8 \times 8$ km, very dark (albedo $\sim 0.04$), and irregular; bright jets of gas and dust emerge from a few discrete active regions on the sunward (left) side. Image credit: ESA / Halley Multicolour Camera Team / MPAe Lindau, public domain.
```

### Short-period and long-period comets

**Short-period comets** ($P < 200$ yr) have low-inclination, prograde orbits near the ecliptic plane.
They divide into **Jupiter-family comets** (JFCs; $P < 20$ yr, $T_J > 2$) and intermediate-period **Halley-type comets**.

The **Tisserand parameter** is an approximate constant of motion in the restricted three-body problem, derived from the Jacobi integral.
Because $T_J$ is conserved across encounters with Jupiter while individual orbital elements vary, it is the standard cometary classifier.
For semimajor axis $a$, eccentricity $e$, and inclination $i$, it is given by

$$
T_J = \frac{a_J}{a} + 2\sqrt{\frac{a}{a_J}\,(1 - e^2)} \, \cos i,
$$

where $a_J = 5.20$ AU is Jupiter's semimajor axis.
Main-belt asteroids have $T_J > 3$ (Ceres, $T_J \approx 3.3$), Jupiter-family comets have $2 < T_J < 3$ (67P/Churyumov-Gerasimenko, $T_J \approx 2.7$), and long-period comets have $T_J < 2$ (1P/Halley, $T_J \approx -0.6$).

Jupiter-family comets originate in the trans-Neptunian scattered disk, where objects perturbed inward by Neptune are captured into Jupiter-controlled orbits {cite:p}`Volk2008`.
Most JFCs survive in these orbits for $\sim 10^4$ to $10^5$ yr before ejection or collision.

**Long-period comets** ($P > 200$ yr) have isotropic orbital orientations and eccentricities approaching 1.
They originate in the **Oort cloud**, deflected inward by the galactic tide and passing stars.
Many make a single passage before ejection, while some are captured into Halley-type orbits.

```{figure} figures/halley2.avif
:name: fig:l12-halley1986
:width: 500px
:align: center

Comet 1P/Halley photographed against the Milky Way on 21 March 1986 during its most recent apparition. The bright coma surrounds the nucleus and the long tail extends across many degrees of sky. Halley is the only short-period comet bright enough to be visible to the unaided eye. Image credit: R. Haefner / European Southern Observatory, CC BY 4.0 (Wikimedia Commons).
```

### The D/H ratio of cometary water

Earth's ocean water has a **deuterium-to-hydrogen ratio** (D/H) of $\sim 1.56 \times 10^{-4}$ (the standard mean ocean water value, or VSMOW), which is closely matched by carbonaceous chondrites.
Comets span a factor of 3 in D/H with no clean separation between dynamical classes: 103P/Hartley 2 ({numref}`fig:l12-hartley2`) is near VSMOW {cite:p}`Hartogh2011`, whereas 67P/Churyumov-Gerasimenko is about three times VSMOW {cite:p}`Altwegg2015`.

```{figure} figures/hartley2.avif
:name: fig:l12-hartley2
:width: 480px
:align: center

Comet 103P/Hartley 2 imaged by NASA's EPOXI mission during the close flyby on 4 November 2010. The peanut-shaped nucleus is approximately 2.2 km long with two distinct lobes joined by a smoother waist; bright jets of gas and dust emerge from both ends. Hartley 2 is notable for having a D/H ratio in its water consistent with Earth's oceans, in contrast to most other measured comets. Image credit: NASA/JPL-Caltech/UMD, public domain.
```

Mixing models show that carbonaceous chondrite water must dominate Earth's volatile budget by mass, with only a small cometary contribution {cite:p}`Alexander2017`.
Earth's water thus reflects inner solar system processing of mostly carbonaceous-chondrite-like material ({ref}`Lecture 9 <lecture09>`).

## Part 3: Messengers and visitors

This final part of the lecture surveys recent results from in situ space missions, sample return, and interstellar visitors.

### Rosetta and 67P/Churyumov-Gerasimenko

The European Space Agency's **Rosetta** mission was the first to orbit a comet, escorting 67P/Churyumov-Gerasimenko through perihelion before a controlled descent in 2016 ({numref}`fig:l12-67p`).

In November 2014, the lander **Philae** made the first soft landing on a cometary nucleus.

```{figure} figures/67p_nucleus.avif
:name: fig:l12-67p
:width: 600px
:align: center

The nucleus of comet 67P/Churyumov-Gerasimenko imaged by Rosetta. The bilobed "duck" shape is one of the most distinctive features of the comet; the two lobes are joined at a narrow neck region (the bright collar visible in this image). The bilobed morphology is interpreted as a contact binary formed from the gentle merger of two primordial cometesimals, much like Arrokoth on a smaller scale. Image credit: ESA/Rosetta/MPS for OSIRIS Team MPS/UPD/LAM/IAA/SSO/INTA/UPM/DASP/IDA (ESA Standard Licence).
```

Rosetta's headline scientific findings include:

- The bilobed nucleus is a **contact binary**, formed by the gentle merger of two primordial cometesimals at low velocity {cite:p}`Massironi2015`.
- Bulk density of $0.533 \pm 0.006$ g cm$^{-3}$ implies a porosity of $\sim 70$-$80\%$ in a fluffy aggregate of dust and ice.
- Cometary water D/H of $(5.3 \pm 0.7) \times 10^{-4}$ is approximately three times the terrestrial ocean value, not matching Earth's water {cite:p}`Altwegg2015`.
- Detection of glycine, phosphorus, methylamine, and other prebiotic molecules in the coma by the ROSINA mass spectrometer {cite:p}`Altwegg2016`.

### Asteroid sample return: Hayabusa, Hayabusa2, OSIRIS-REx

**Sample-return missions** deliver physical material from small bodies back to Earth, establishing direct links between laboratory meteorite groups and known asteroid parent bodies.

Hayabusa returned microscopic grains from S-type asteroid (25143) Itokawa in 2010 ({numref}`fig:l12-itokawa`), confirming a link to LL-class ordinary chondrites {cite:p}`Nakamura2011`.
Itokawa, Ryugu and Bennu are all **rubble piles**: aggregates of boulders and gravel held together by weak self-gravity rather than monolithic rock, with bulk densities near $1.2$ to $1.9$ g cm$^{-3}$ and macroporosities of tens of per cent ({ref}`Lecture 8 <lecture08>`).

```{figure} figures/itokawa_full.avif
:name: fig:l12-itokawa
:width: 500px
:align: center

The S-type near-Earth asteroid (25143) Itokawa imaged by Hayabusa AMICA in 2005. Itokawa is approximately 535 m long with a distinctive bilobed sea-otter profile: two rougher lobes joined by the smoother "Muses Sea" neck region (the central tan area). The complete absence of large craters and the high boulder fraction are signatures of a **rubble-pile** internal structure: a body held together by self-gravity and inter-fragment friction rather than by internal cohesion. Itokawa is the parent body of the LL-class ordinary chondrites. Image credit: ISAS / JAXA, CC BY 4.0.
```

Hayabusa2 visited the carbonaceous (Cb-type) asteroid (162173) Ryugu and returned 5.4 g of sample in 2020 ({numref}`fig:l12-ryugu`).
The returned material is rich in water-bearing phyllosilicates, carbonates, and organic molecules, including amino acids and the nucleobase uracil, with a bulk composition matching CI chondrites {cite:p}`Yokoyama2023,Naraoka2023,Oba2023`.

```{figure} figures/ryugu.avif
:name: fig:l12-ryugu
:width: 480px
:align: center

The Cb-type near-Earth asteroid (162173) Ryugu imaged by Hayabusa2 ONC-T on 12 July 2018, showing an approximately 900 m diameter "spinning top" shape produced by past rapid rotation. Image credit: JAXA / Hayabusa2 ONC team, processed by Kevin M. Gill, CC BY 2.0.
```

OSIRIS-REx visited the carbonaceous (B-type) asteroid (101955) Bennu and returned 121 g of sample in 2023 ({numref}`fig:l12-bennu`).
Analyses revealed hydrated phyllosilicates, carbonates, and organic matter with a CI-chondrite-like bulk composition, recording extensive aqueous alteration on an outer-solar-system precursor {cite:p}`Lauretta2024`.

```{figure} figures/bennu_lauretta.avif
:name: fig:l12-bennu
:width: 380px
:align: center

(101955) Bennu and the OSIRIS-REx sample collection site at three nested zoom levels, showing the $\sim 490$ m rubble-pile asteroid (top) down to the 1.4 m field of view at Nightingale (bottom). From Figure 1 of {cite:t}`Lauretta2024`.
```

### Lucy: a tour of the Trojans

**Jupiter Trojans** are asteroids librating around the L4 and L5 Lagrange points in Jupiter's orbit, $60^{\circ}$ ahead of and behind the planet.

Their D-type spectra resemble Kuiper Belt objects, which suggests capture from distant planetesimals during the Nice-model instability {cite:p}`Morbidelli2005` ({ref}`Lecture 2 <lecture02>`).
NASA's Lucy mission (launched October 2021) is a 12-year tour visiting eight Trojans across both swarms between 2027 and 2033.
En route to the Trojans, Lucy encountered the inner main-belt asteroid (152830) Dinkinesh, discovering that its orbiting moonlet Selam is a contact binary ({numref}`fig:l12-dinkinesh`).

```{figure} figures/dinkinesh_levison.avif
:name: fig:l12-dinkinesh
:width: 600px
:align: center

The Dinkinesh-Selam system imaged by NASA's Lucy mission during its flyby on 1 November 2023. (a-f) Three stereo pairs of the inner main-belt asteroid (152830) Dinkinesh, about 720 m across, on approach, near closest approach and on departure; yellow dots trace its longitudinal trough, red dots its equatorial ridge, and the moonlet appears behind the primary in (d). (g) Dinkinesh with the trough section (cyan) shifted back into place. (h-l) The moonlet Selam, a contact binary of two near-equal lobes of about 210 and 230 m, and (m) the pair at their true separation (scale bars 200 m and 1 km). Reproduced from {cite:t}`Levison2024`, Fig. 1.
```

### Psyche: the metal world

(16) Psyche ($\sim 222$ km effective diameter) is the largest M-type asteroid, interpreted as the **exposed core** {cite:p}`ElkinsTanton2020` of a differentiated planetesimal whose silicate mantle was stripped by a giant impact.
{numref}`fig:l12-psyche` shows the shape that radar and adaptive-optics imaging give it: a flattened body whose outline departs from an ellipsoid at three depressions.

```{figure} figures/psyche_shape_shepard.avif
:name: fig:l12-psyche
:width: 600px
:align: center

Shape of (16) Psyche seen from above its south pole, from a combined radar and adaptive-optics model. Left: the best-fit ellipsoid (dashed) over the shape model, with the regions Alpha, Bravo and Charlie where the body falls short of the ellipsoid. Right: a rounded-rectangle outline that follows the model more closely at Bravo and Charlie; Alpha West marks a further depression. Longitudes 0°, 180° and 270° are indicated. Reproduced from {cite:t}`Shepard2021`.
```

### Interstellar visitors

**Interstellar objects** are bodies originating outside the solar system on unbound, hyperbolic orbits.
Discovered in October 2017, 1I/'Oumuamua was the first confirmed interstellar visitor, a small ($\sim 100$–$200$ m long), highly elongated body with non-gravitational acceleration but no detectable coma {cite:p}`Micheli2018`.
Its detection implies a much higher space density of interstellar objects than pre-discovery models predicted ({numref}`fig:l12-oumuamua`).

```{figure} figures/oumuamua_iso_density.avif
:name: fig:l12-oumuamua
:width: 600px
:align: center

Inferred space number density of interstellar objects in pc$^{-3}$ implied by the discovery of 1I/'Oumuamua, broken down by assumed population type along the horizontal axis (asteroidal, comets from giant-planet ejection, comets from white-dwarf disruption, free-floating planetary fragments, two-population mixed models). The vertical axis is the implied space density. The Pan-STARRS detection of a single ISO already implies densities orders of magnitude larger than pre-discovery predictions for any reasonable assumed population, and survey statistics now imply $\sim 10^{4}$ ISOs larger than $\sim 100$ m within Neptune's orbit at any given time. From Figure 2 of {cite:t}`OumuamuaTeam2019`.
```

Discovered in August 2019, 2I/Borisov was the first confirmed interstellar comet ({numref}`fig:l12-borisov`).
Spectroscopy revealed a CO abundance of at least 173% relative to water, compared to typical solar-system cometary values near 4% {cite:p}`Bodewits2020`.
A CO/H$_2$O ratio above unity indicates that Borisov formed beyond the CO snowline in a colder region of its host system.

```{figure} figures/borisov_composition_bodewits.avif
:name: fig:l12-borisov
:width: 700px
:align: center

Composition of volatiles in the coma of 2I/Borisov compared with comets in our solar system. The vertical axis is the elemental abundance ratio with respect to atomic oxygen; the horizontal axis is the atomic mass ratio of the element to oxygen, which spreads the four ratios H/O, C/O, N/O, and S/O across the plot. Each ratio has five symbols: average solar-system comets (blue circle), 67P/Churyumov-Gerasimenko (orange triangle), comet C/2009 P1 (Garradd) averaged (filled red square) and after perihelion (open red square), and 2I/Borisov (green triangle). Error bars are $1\sigma$, and a small arrow below a symbol marks an upper limit, as for S/O in Borisov. Borisov has C/O $= 0.71 \pm 0.35$, about seven times the average cometary value of $0.098 \pm 0.058$, while its H/O of $0.58 \pm 0.36$ is about three times lower than the cometary average of $1.97 \pm 0.13$. Both differences point to a carbon-rich, water-poor ice inventory and therefore to formation beyond the CO snowline of its host planetary system. From Figure 3 of {cite:t}`Bodewits2020`.
```

### Open questions and frontier topics

Several key questions remain open across small-body science:

- **Structure of the Oort cloud**: The Oort cloud remains an inferred population with no direct observations, imaging, or samples.
- **Origin of Earth's water**: The relative contributions of carbonaceous chondrites and comets remain under active investigation, connecting to {ref}`Lecture 9 <lecture09>` and {ref}`Lecture 14 <lecture14>`.
- **Dynamically new comets**: ESA's Comet Interceptor (launch 2029) will wait at the Sun-Earth L2 point for a comet entering the inner solar system for the first time {cite:p}`Snodgrass2019`.
- **Interstellar visitors**: Each interstellar object samples the formation chemistry of another planetary system, providing a comparative cosmochemistry probe across stars.

The multi-spacecraft architecture of ESA's Comet Interceptor will observe an incoming pristine or interstellar target from multiple directions simultaneously ({numref}`fig:l12-comet-interceptor`).

```{figure} figures/comet_interceptor.avif
:name: fig:l12-comet-interceptor
:width: 600px
:align: center

Schematic of the planned ESA Comet Interceptor mission, showing the multi-spacecraft flyby geometry. The main spacecraft and two smaller probes will sample a single target from multiple directions, allowing the first 3D reconstruction of an active comet's nucleus, coma, and plasma environment. Image credit: ESA, CC BY-SA 3.0 IGO.
```

## Summary and takeaways

Small bodies are the **formation fossils** of the solar system: everything that did not become a planet.
The inner small bodies (main-belt asteroids and the meteorites they deliver) are dominantly rocky leftovers; the outer small bodies (Kuiper Belt, scattered disk, Oort cloud, comets) are dominantly icy leftovers.

Kirkwood gaps in the main belt, the bimodal inclination distribution of the classical Kuiper Belt, the existence of a populated 3:2 resonance with Neptune (the plutinos including Pluto), and the trapped Trojan swarms at Jupiter's L4 and L5 Lagrange points are all fossils of giant-planet migration.

**Meteorites** are the only solar-system samples that can be studied in a terrestrial laboratory, and they provide direct ground-truth on ages and compositions that no spacecraft instrument can match.
The Pb-Pb age of CAIs at $4567.30 \pm 0.16$ Myr anchors time zero, with parallel uranium decay chains making the isochron method self-calibrating.

The **NC-CC dichotomy** is one of the most important open problems in cosmochemistry.
Three competing physical interpretations (Jupiter as an early gravitational barrier, snow-line migration with pebble isolation, and a temporal-epoch model with no spatial barrier) each match the existing data and each make different predictions for how the solar system's planetesimal reservoirs were structured in time and space.

**Recent missions** have transformed our understanding of small bodies.
Rosetta gave us cometary geophysics and the closest direct sample of comet structure.
Hayabusa2 and OSIRIS-REx returned carbonaceous material from Ryugu and Bennu, linking spectral types to chondrite groups, while DART demonstrated kinetic asteroid deflection.

The new and most unexpected category is **interstellar visitors**.
They are the first physical samples of other planetary systems directly observable in our solar neighbourhood, linking small bodies to the exoplanet science of {ref}`Lecture 13 <lecture13>`.

## References

```{bibliography}
:filter: docname in docnames
```