Planet classification

Planet classification is the sorting of planets into kinds by what can be measured about them: size, mass, density, temperature and, where a spectrum exists, the chemistry of the atmosphere. Astronomy has no single official scheme. There is a formal definition of what a planet is, and a set of working categories (super-Earth, mini-Neptune, hot Jupiter) that grew out of the instruments that found them. Pax Abyssi needs something stricter, because its simulation has to decide what every generated world is before it can compute an atmosphere or draw a surface, so it files each planet under one of 37 coded types. This page explains both: the categories astronomers use, and how the sim's codes map onto them.

What counts as a planet

For the Solar System the rule is the one the International Astronomical Union (IAU) voted on in Prague on 24 August 2006. A planet orbits the Sun, is massive enough for its own gravity to pull it into a nearly round shape, and has cleared the neighbourhood around its orbit. A body that meets the first two tests but not the third, and is not a moon, is a dwarf planet, which is why Pluto and Eris are now classed as dwarf planets ^[1].

Planets around other stars need a different test, because nobody can check whether a world 100 light years away has cleared its orbit. The IAU's working definition of an exoplanet, amended in 2018, sets an upper mass: below about 13 Jupiter masses, the point at which an object of solar composition can fuse deuterium in its core. It must orbit a star, a brown dwarf or a stellar remnant, and its mass must be less than about one twenty-fifth of its host's ^[2]. The 13 Jupiter-mass line is a convention more than a physical wall; the deuterium-burning threshold moves between about 11 and 16 Jupiter masses with composition ^[3]. Objects above it are brown dwarfs.

As of 25 September 2026 the NASA Exoplanet Archive listed 6,372 confirmed exoplanets ^[4].

How astronomers sort planets

By size and mass

Most known exoplanets were found by one of two methods, and each measures one number. A transit, the small dip in starlight as a planet crosses its star, gives the planet's radius. A radial-velocity survey, which detects the star's wobble, gives a minimum mass (the true mass times the sine of the orbit's unknown tilt). The first working classes were therefore classes of size. The Kepler mission sorted its candidates into Earth-size (below 1.25 Earth radii), super-Earth-size (1.25 to 2), Neptune-size (2 to 6) and Jupiter-size (6 to 15 Earth radii) ^[5]. NASA's public categories follow the same logic: terrestrial planets, super-Earths, Neptune-like planets (including mini-Neptunes) and gas giants ^[6].

The surprise of the Kepler era was that the most common planets around Sun-like stars, at least on orbits shorter than about a year, fall between Earth and Neptune in size, a kind the Solar System lacks ^{[7] [8]}. Inside that range sits the Radius valley, a shortage of planets between about 1.5 and 2 Earth radii that divides rocky super-Earths from gas-wrapped mini-Neptunes ^[9].

Where both mass and radius are known, the pair itself defines classes. Fitting the mass-radius relation of 316 well-measured objects from dwarf planets to low-mass stars, Chen and Kipping (2017) found the slope changes at three places: near 2 Earth masses, where solid "Terran" worlds give way to "Neptunian" worlds with thick envelopes; near 0.41 Jupiter masses (about 130 Earth masses), where "Jovian" worlds begin and radius stops growing with mass; and near 0.08 solar masses, where hydrogen fusion turns a body into a star ^[10].

By density and composition

Mass and radius together give the bulk density,

$$\rho = \frac{3M}{4\pi R^3} \approx 5.51\ \mathrm{g\,cm^{-3}} \times \frac{M/M_\oplus}{(R/R_\oplus)^3},$$

and density is the first clue to composition. Earth averages 5.51 g/cm³, Mercury 5.43, water ice about 1, and Saturn 0.69. Planets are compressed by their own weight, so the comparison is made against model curves for pure iron, Earth-like rock, rock mixed with water, and rock with a hydrogen envelope rather than against a single number ^[11]. The curves overlap: a planet of a given mass and radius can often be matched by more than one recipe, a problem called compositional degeneracy. That is why the same measured world can be argued to be a water world or a rocky core under a thin hydrogen layer (see Ocean world and Mini-Neptune).

By temperature

The second axis is warmth. A planet's equilibrium temperature is the temperature a black body would reach if it absorbed the starlight falling on it and radiated the same power back to space, with the heat spread over the whole sphere:

$$T_\mathrm{eq} = T_\star \sqrt{\frac{R_\star}{2a}}\,(1 - A_B)^{1/4} \approx 278\ \mathrm{K}\,(1 - A_B)^{1/4} \left(\frac{L_\star}{L_\odot}\right)^{1/4} \left(\frac{a}{1\ \mathrm{AU}}\right)^{-1/2}$$

Here $T_\star$, $R_\star$ and $L_\star$ are the star's temperature, radius and luminosity, $a$ is the orbital distance and $A_B$ is the Bond albedo, the fraction of all incoming light the planet reflects. For Earth, with $A_B \approx 0.3$, the formula gives 255 K, well below the measured mean surface temperature of about 288 K; the 33 K difference is the greenhouse effect. It is still the one temperature that can be computed for any planet with a known orbit, so it is the standard axis for words like "hot", "temperate" and "cold". A dayside that cannot share its heat with the night side runs hotter than $T_\mathrm{eq}$; see Lava world.

By atmosphere

For giant planets, temperature controls which clouds can condense, and so how the planet looks. Sudarsky, Burrows and Pinto (2000) turned that into five theoretical classes, from ammonia-cloud giants below about 150 K to silicate-cloud giants above about 1,400 K ^[12]. The scheme predates any measured spectrum of a cool exoplanet giant and is best read as a model; see Sudarsky classification. For small planets, spectra from the James Webb Space Telescope are only now starting to separate bare rock from atmosphere, so atmospheric classes of rocky exoplanets remain provisional.

The Pax Abyssi scheme

The sim's taxonomy is built on the same axes (mass, radius, temperature, composition, atmosphere), but it has to be complete: every generated world must land in exactly one class. It has three levels. The type (for example TOW, an ocean world) decides which physics engine runs. The subtype (TOW-AR, a young Archean-style ocean world) is a physics-derived refinement that every planet carries. Some subtypes have a third level, the sub-subtype, a visual variant that decides what the planet looks like (for example TBB-FB, a flood-basin basalt world). Codes are mnemonic: T for terrestrial, B for barren, and so on. Older codes such as T2-O survive in the sim's internals and are listed as aliases on each class page.

There are 37 types in nine series:

Most types split further. The gas and ice giants divide by temperature, then by chemistry, internal heat or haze, then by weather, into 41 types, shown with pictures on the family pages linked from Gas giant; the lava worlds by dayside temperature and whether they are tidally locked, the mini-Neptunes by temperature bands, and the habitable-zone worlds by atmosphere and ocean cover, giving roughly 100 subtypes in all. The class pages give each type's subtypes and the numbers that trigger them.

How a planet gets its type

A generated system is built outward from its star. The system's architecture (compact, resonant, hot-Jupiter and so on) lays out orbital slots, each slot is labelled by how much starlight it receives, and a type is drawn from that zone's probability table. The table is tilted by the chemistry of the star's disc: a carbon-rich disc makes carbon planets more likely, an iron-rich one allows iron planets, and a volatile-rich one favours ocean and ice worlds. Temperature guards then catch impossible draws, so an ice type placed too close to its star becomes a barren or lava world.

The type's physics engine then computes the planet: mass, radius, atmosphere, albedo, surface temperature. The engine may reclassify the world if the physics disagrees with the label. A world whose rock is melting becomes a lava world; a greenhouse world that ends up below the boiling point of water becomes a mixed world; an ocean world that freezes over becomes a subsurface ocean world.

Real exoplanets in the sim's catalogue are typed differently, because for them the mass is known and the architecture is not. A decision tree works down from mass: rocky branches below about 2 Earth masses, lava worlds and mini-Neptunes from 2 to 10 depending on temperature and radius, hot mini-Neptunes or ice giants to 50, and gas giants beyond, split into hot, temperate and cold by equilibrium temperature.

How the codes relate to the literature

Pax Abyssi's types are built on the literature's axes and go further, because a simulation that draws every world needs a finer grain than astronomy's working categories. The top levels follow the literature closely: the Super-Earth and Mini-Neptune pages use its definitions, and our giant families line up with the Sudarsky classification, as that page shows. The finer levels are our own, named for what a pilot sees: a "Phosphine-Ruddy" cold giant is one whose phosphorus-rich chemistry gives it a rust haze, and a "dry habitable world" is a world with air and liquid water but little ocean. Each class page says which parts come from the literature and which are ours, and carries a science-status banner: observed, model, sim or speculative.

Image pendingFile:Mass_radius_diagram_sim_classes.svgA log-log plot of planet mass against radius, with confirmed exoplanets as dots, four composition curves, and shaded regions for the sim's planet series.

See also

• Radius valley
• Super-Earth and Mini-Neptune
• Habitable zone
• Sudarsky classification
• Gas giant and Ice giant
• Star system generation
• Planetary system archetypes
• Planet occurrence
• Sol

References

1. 1International Astronomical Union (2006). IAU 2006 General Assembly: Result of the IAU Resolution votes. IAU press release iau0603. www.iau.org/IAU/Iau/News/PR2006/iau-2006-general-assembly-resolution-votes.aspx
2. 2Lecavelier des Etangs, A. and Lissauer, J. J. (2022). The IAU working definition of an exoplanet. New Astronomy Reviews 94, 101641. doi:10.1016/j.newar.2022.101641
3. 3Spiegel, D. S., Burrows, A. and Milsom, J. A. (2011). The deuterium-burning mass limit for brown dwarfs and giant planets. The Astrophysical Journal 727, 57. doi:10.1088/0004-637X/727/1/57
4. 4NASA Exoplanet Science Institute (2026). NASA Exoplanet Archive. Caltech/IPAC. exoplanetarchive.ipac.caltech.edu/
5. 5Borucki, W. J. et al. (2011). Characteristics of planetary candidates observed by Kepler. II. Analysis of the first four months of data. The Astrophysical Journal 736, 19. doi:10.1088/0004-637X/736/1/19
6. 6NASA Science. Exoplanet types. NASA Science: Exoplanets. science.nasa.gov/exoplanets/planet-types/
7. 7Fressin, F. et al. (2013). The false positive rate of Kepler and the occurrence of planets. The Astrophysical Journal 766, 81. doi:10.1088/0004-637X/766/2/81
8. 8Petigura, E. A., Howard, A. W. and Marcy, G. W. (2013). Prevalence of Earth-size planets orbiting Sun-like stars. Proceedings of the National Academy of Sciences 110, 19273-19278. doi:10.1073/pnas.1319909110
9. 9Fulton, B. J. et al. (2017). The California-Kepler Survey. III. A gap in the radius distribution of small planets. The Astronomical Journal 154, 109. doi:10.3847/1538-3881/aa80eb
10. 10Chen, J. and Kipping, D. (2017). Probabilistic forecasting of the masses and radii of other worlds. The Astrophysical Journal 834, 17. doi:10.3847/1538-4357/834/1/17
11. 11Zeng, L. et al. (2019). Growth model interpretation of planet size distribution. Proceedings of the National Academy of Sciences 116, 9723-9728. doi:10.1073/pnas.1812905116
12. 12Sudarsky, D., Burrows, A. and Pinto, P. (2000). Albedo and reflection spectra of extrasolar giant planets. The Astrophysical Journal 538, 885-903. doi:10.1086/309160