Exoplanet Demographics Eval

---
name: exoplanet-demographics-eval
description: Evaluates a cosmological galaxy simulation framework by comparing its synthetic exoplanet population demographics against real observational catalogs from the NASA Exoplanet Archive and Kepler mission. Use when the user wants to benchmark on NASA Exoplanet Archive, Kepler observations, or asks about evaluating this task. Reports planet type fraction.
metadata:
  skill_kind: dataset_eval
  source_arxiv: 2511.14704
  bibtex_key: padois2025glxsimu
  confidence: high
---

# exoplanet-demographics-eval

> Modelling the Milky Way's exoplanet population based on cosmological galaxy simulations — Chlo\'e Padois et al. (arXiv:2511.14704, 2025)

## What this evaluates

Evaluates a cosmological galaxy simulation framework by comparing its synthetic exoplanet population demographics against real observational catalogs from the NASA Exoplanet Archive and Kepler mission.

## Datasets

- **NASA Exoplanet Archive** — total ?; splits: (unstated)
- **Kepler observations** — total ?; splits: (unstated)

## Metrics

- `planet type fraction` **(primary)** — range: percent
  - Percentage of simulated planets falling into categories: Earth-like, super-Earths/Neptunes, and giants/sub-giants.
- `CHZ occurrence rate ($\eta_{\oplus}$)` — range: percent
  - Fraction of simulated planets located within the circumstellar habitable zone, defined using optimistic recent Venus and early Mars instellation limits ($S/S_{\oplus}=1.78$ and $0.32$).
- `distribution matching` — range: other
  - Visual and qualitative comparison of simulated mass-period and effective temperature vs. instellation distributions against observed exoplanet populations.

## Input / output format

**Input**: Stellar particle properties (mass, metallicity, age, birth radius) from cosmological simulation snapshots, processed through planet formation and occurrence rate prescriptions.

**Output**: Catalog of simulated exoplanets with properties: host star parameters, planet mass, orbital period, instellation, and CHZ classification.

## Scoring recipe

```python
Calculate fraction of planets per type: count(type) / total_planets * 100.
Determine CHZ fraction: count(in_CHZ) / total_planets * 100.
Compare simulated distributions (mass-period, Teff vs instellation) against observational catalogs (NASA, Kepler) using visual overlay and qualitative assessment of parameter space coverage.
```

## Common pitfalls

- Observational selection effects strongly bias real catalogs against faint M-dwarf hosts, making direct numerical comparison without correction misleading.
- CHZ boundaries vary significantly across literature; the paper uses an optimistic definition (Kopparapu2013) which inflates the reported habitable zone fraction compared to conservative estimates.

## Evidence (verbatim from paper)

> Simulating exoplanets only around single stars as described in Sect. [2.3], we end up with a total sample of 22.6 million planets, organised in 2.9 million planetary systems and composed of around 52.5% of Earth-like planets, 44% of super-Earths and Neptunians, and around 3.5% of giants and sub-giants. About 23% of them are located in the circumstellar habitable zone (CHZ; see Sect. [3.4]).

## Citation

```bibtex
@misc{padois2025glxsimu,
  title={Modelling the Milky Way's exoplanet population based on cosmological galaxy simulations},
  author={Chlo\'e Padois et al.},
  year={2025},
  note={arXiv:2511.14704}
}
```

- arXiv: 2511.14704