ExoWorld Habitability Tuner

Live demo: eagnespuerto.github.io/exoworld-tuner

ExoWorld offers insight into the possible habitability of any exoplanet you can imagine. Be it for research or for science fiction writing, this tool is handy in ensuring you are working with a planet that may or may not be habitable. It can also be used as an interactive teaching tool for science educators looking to share knowledge on how worlds can or cannot be habitable.

A single-file static site — open index.html in any modern browser, or host it on GitHub Pages. The tuner is a UI port of the VetStar / STEHM habitability scoring engine in eagnespuerto/vetstar (backend/app/habitability.py).

How to use

Drag the sliders to set a hypothetical planet (orbital distance, radius, mass) and pick a host star spectral class plus pipeline/disposition flags. The Habitability Chance Index (HCI) recomputes live, breaking the result into six weighted sub-scores and a density modifier.

Presets are included for an Earth twin, the full TRAPPIST-1 system, a hot Jupiter, a super-Earth, and an eclipsing-binary false positive.

Themes

The header has a four-way theme switch:

Theme	Use case
Night	Default — observatory dark palette
Black	OLED-friendly pure black
Warm	Sepia / paper-lamp tones for low-light reading
Light	Daytime / high-contrast monitors

The chosen theme is persisted in localStorage. The system-diagram stage stays dark in every theme — space doesn't have a light mode, and locking it preserves the meaning of the spectral colour ramp.

Host star

Three ways to set the host:

1. Spectral-class buttons (M / K / G / F / A) — uses a class-average Teff and R★ (good for quick what-ifs).
2. Custom — type an arbitrary Teff (K) and R★ (R⊙). The HZ ring, stellar-type sub-score, and the star's rendered colour all update from Teff.
3. Known star dropdown — picks a real host's measured Teff and R★ from the literature (Sun, Proxima Centauri, TRAPPIST-1, Alpha Centauri A/B, Tau Ceti, the Kepler/TOI hosts, etc.) and switches the selector to Custom. This is the right choice when you want to speculate about hypothetical planets in a real system: pick the host, then drag the orbit/radius sliders to place a fictional world around it.

Planet presets also auto-populate their real host star — picking TRAPPIST-1e now sets Teff = 2566 K, R★ = 0.119 R⊙ (the actual ultracool dwarf), not the generic M-dwarf class average. This makes the HZ edges and stellar sub-score noticeably more honest for compact systems.

Whenever you change just the star (a spectral button or a known-star pick — not a full planet preset), the orbital-distance slider snaps to the centre of that star's conservative HZ. So the diagram opens on a habitable orbit by default, and you can drag inward/outward from a sensible starting point.

Preset auto-revert

The "known exoplanet preset" dropdown is a snapshot. The moment you tweak any planetary observable (orbit, radius, mass, disposition, vetting flag, sectors) the dropdown reverts to "— select a template —", so the UI never claims you're still looking at TRAPPIST-1e once you've moved the sliders.

System diagram

The mini-system panel above the gauge is a top-down view of the orbital plane. The star sits at the centre of the projection, the orbit and the conservative habitable zone are drawn as concentric circles (the HZ band literally curves around the star), and the planet is rendered as a half-lit disc with its day side facing the star. The dashed green circles mark the inner and outer conservative HZ edges from Kopparapu et al.

---

The HCI calculation

The HCI is a weighted average of six independent sub-scores (each on [0, 1]), scaled to a 0–100 index, then nudged by a bulk-density modifier. A hard override caps obvious false positives.

HCI = (Σ wᵢ · sᵢ) / (Σ wᵢ) · 100  +  density_modifier

Component	Weight	What it measures
Planet size	30%	STEHM atmosphere-retention regime
Habitable zone	25%	Insolation relative to Kopparapu HZ edges
Stellar type	15%	How well STEHM calibration transfers
TOI disposition	15%	ExoFOP confidence level
Vetting flags	10%	Pipeline verdict (planet / EB / blend)
Multi-sector	5%	Detection consistency across TESS sectors

1. Planet size (30%) — STEHM

Uses thresholds from Hill et al. (2026), STEHM, arXiv:2605.00170:

• Rp > 2.2 R⊕ → s = 0.05 (likely sub-Neptune, not rocky)
• Rp < 0.5 R⊕ → s = 0.02 (atmosphere retention ≈ 0)
• Rp ≥ 0.8 R⊕ → s = 0.75 + 0.25 · min((Rp − 0.8)/0.2, 1) (favourable; can retain long-term CO₂)
• 0.7 ≤ Rp < 0.8 → s = 0.35 + 0.40 · (Rp − 0.7)/0.1 (marginal)
• 0.5 ≤ Rp < 0.7 → s = 0.05 + 0.30 · (Rp − 0.5)/0.2 (unlikely)

If mass is provided, the size score is multiplied by a density-verdict factor:

• ρ/ρ⊕ ≥ 0.6 → ×1.1 (rocky-consistent)
• ρ/ρ⊕ < 0.4 → ×0.35 (volatile-rich)
• otherwise → ×0.7 (intermediate)

2. Habitable zone (25%) — Kopparapu et al. (2013/2014)

The HZ edges scale with stellar luminosity L★ = R★² · (Teff / 5778 K)⁴:

limit(au) = S₀(Teff) · √L★

where S₀(Teff) is a 4th-order polynomial in (Teff − 5780) for each of:

• recent_venus and runaway_greenhouse → optimistic & conservative inner edges
• maximum_greenhouse and early_mars → conservative & optimistic outer edges

Scoring, with iO, iC, oC, oO denoting the four edges:

• a < iO → s = 0.05 (too hot, runaway greenhouse)
• a > oO → s = 0.10 (too cold, frozen)
• iO ≤ a < iC → s = 0.30 + 0.25 · (a−iO)/(iC−iO) (warm OHZ edge)
• oC < a ≤ oO → s = 0.65 + 0.30 · (oO−a)/(oO−oC) (cool OHZ edge)
• iC ≤ a ≤ oC → s = min(0.75 + 0.20 · (1 − |f − 0.5|·2), 1) with f = (a−iC)/(oC−iC) (CHZ, peaks mid-zone)

3. Stellar type (15%)

Coarse buckets by Teff:

Regime	Teff (K)	Score
G dwarf (solar analog)	5000 – 6000	0.90
F dwarf	6000 – 7500	0.80
K dwarf	3700 – 5000	0.65
Hot star (A or earlier)	> 7500	0.40
M dwarf	< 3700	0.30

STEHM was calibrated for Sun-like stars; lower scores for M dwarfs reflect harsher XUV and flaring.

4. TOI disposition (15%) — ExoFOP

Code	Meaning	Score
CP / KP	Confirmed / Known planet	1.00
PC / APC	Planet candidate / ambiguous	0.75
TOI (unc.)	Unclassified TOI	0.55
(none)	No TOI designation	0.50
FP / FA	False positive / false alarm	0.05

5. Vetting flags (10%) — pipeline verdict

Verdict	Score
Planet candidate — on-target centroid	0.85
Planet candidate — centroid unclear	0.55
Ambiguous	0.45
No significant signal	0.40
Eclipsing-binary candidate / blend	0.05

6. Multi-sector (5%)

Let obs = sectors observed, det = sectors showing the dip:

• obs = 1 → s = 0.40 (single sector, period unconstrained)
• det = 0 → s = 0.30 (no detections — inconsistent with a transit)
• otherwise → s = min(0.40 + 0.60 · det/obs, 1)

Bulk-density modifier (±10 pts, post-average)

After the weighted average is scaled to 0–100:

• ρ/ρ⊕ ≥ 0.6 (terrestrial, ρ ≳ 3.3 g/cm³) → +10
• ρ/ρ⊕ < 0.4 (gas-giant, ρ ≲ 2.2 g/cm³) → −10
• otherwise → 0

Hard override

If the pipeline verdict is eclipsing_binary_candidate or false_positive_blend, the final HCI is capped at 12 regardless of other inputs.

Tier mapping

HCI	Tier
≥ 70	Promising
45 – 69	Marginal
20 – 44	Unlikely
< 20	Very unlikely

---

Caveats

• STEHM models a pure-CO₂ stagnant-lid planet as a best case. Non-thermal escape, magnetic fields, and plate tectonics are excluded — treat the HCI as a first-order estimate, not a vetting verdict.
• The M-dwarf regime is intentionally penalised: STEHM is not calibrated for M-dwarf XUV environments, and the safe-size threshold may exceed 0.8 R⊕ there.
• Disposition and vetting components encode observational confidence, not intrinsic habitability — a confirmed hot Jupiter still scores poorly because size and zone dominate.

References

• Hill et al. (2026), STEHM, arXiv:2605.00170
• Kopparapu et al. (2013, 2014) — habitable-zone limits
• TESS / ExoFOP TOI dispositions
• Source scoring: backend/app/habitability.py in eagnespuerto/vetstar