🪐 A year in 23 hours. The lava world that surprised astronomers

Before a single day ends on Earth, an entire year passes on HD 3167 b. This planet orbits so close to its star that we expect molten rock on its illuminated side. Yet the James Webb Space Telescope revealed a puzzle: it emits less heat than a simple model of a scorching, exposed rock would predict.

The most remarkable part is not just the extreme planet. It is how we can investigate whether a world roughly 154 light-years away has an atmosphere. No landing, no spacecraft orbiting it, and no detailed photograph of its surface. A tiny change in radiation, careful measurements and physics can tell us something extraordinary. [1, 2]

Beyond Pluto, around an entirely different star

HD 3167 b is not part of our Solar System. It orbits the star HD 3167 in the constellation Pisces, approximately 154 light-years away. Its host is a K-type star, cooler than our Sun. The planet’s extremely close orbit, however, more than makes up for that difference. [1, 2, 3]

A light-year measures distance: how far light travels in a year. The radiation arriving from this system has therefore spent roughly 154 years crossing space. That is not a spacecraft travel time. A rocket travels vastly more slowly than light.

Pluto still belongs to our own solar neighbourhood. HD 3167 b belongs to another star. A route drawn as Earth → Jupiter → Pluto → HD 3167 helps establish the idea, but the scale changes enormously between the final two stops. A diagram with evenly spaced objects cannot represent their actual distances.

With about 1.6 times Earth’s radius and around 4.8 times its mass, HD 3167 b is classified as a super-Earth. The term describes a size or mass category. It does not mean a better version of Earth, or imply that the planet is habitable. [3]

What a 23-hour year actually means

A year is the time needed to complete one orbit around a star. HD 3167 b takes approximately 0.96 Earth days, or 23 hours. During one week on Earth, it would complete roughly seven orbits. [1]

The word “Earth” matters. This does not mean sunrise and sunset necessarily repeat every 23 hours. Orbiting a star and rotating on an axis are different motions. On such close orbits, tidal locking is expected: the planet may keep the same hemisphere facing its star, much as the Moon keeps almost the same face towards Earth. [1]

Instead of familiar mornings and evenings, the world could have a permanently illuminated hemisphere and a permanently dark one. That contrast is useful. If the dayside receives an enormous amount of energy, where does that energy go?

A lava world is not a photograph of lava

The label “lava world” comes from the physical conditions. With such intense irradiation, at least part of the dayside may be hot enough to melt rock. The exact melting temperature depends on the material’s composition. [1]

We do not, however, have a photograph of a magma ocean on HD 3167 b. Glowing cracks and orange landscapes in visualisations are artists’ interpretations. They convey an extreme environment, but do not show mapped mountains, coastlines or the planet’s measured colour.

That distinction does not diminish the discovery. It demonstrates how much we can learn without resolving a distant world into a detailed globe. Astronomy often works by decoding radiation rather than producing the kind of close-up photograph we instinctively associate with exploration.

The trick is to let the planet disappear

Imagine two lamps beside one another: one very bright and one faint. From a distance, you measure their combined brightness. Cover the faint lamp while the bright one remains stable, and the difference tells you how much light the faint lamp contributed.

A team led by Brandon Park Coy used a similar principle. Webb’s MIRI instrument observed the system in the mid-infrared, where the planet’s thermal radiation can be measured. When HD 3167 b passed behind its star, its contribution temporarily disappeared. Astronomers call this a secondary eclipse. [1]

It is different from a transit. During a transit, a planet crosses in front of its star, blocking some starlight. During a secondary eclipse, it disappears behind the star, removing the planet’s own radiation from the observed signal. Comparing the signal before and during the eclipse reveals its thermal contribution. [1, 4]

In the infrared band analysed, the drop was approximately 38 parts per million, with an uncertainty of 11 parts per million: about 0.0038%. The observation lasted around 4.3 hours, and the team checked the data using three independent processing pipelines. This is a precision measurement of a tiny difference, not a visible dimming that a human observer could simply spot. [1]

Why roughly a thousand degrees is surprisingly cool

The researchers compared their result with a reference model: a dark body that absorbs incoming starlight and transports no heat to its nightside. Its characteristic dayside temperature would be approximately 2,000°C. The observed radiation instead implies around 1,030°C, with an uncertainty of +180/−200°C. [1]

This does not mean every point on the surface is exactly that hot. The quantity is a brightness temperature: the temperature an ideal emitter would need to produce the corresponding radiation in the observed band. If the planet has an atmosphere, we may be detecting radiation from atmospheric layers rather than directly from the rock beneath them.

Even with that qualification, the difference is substantial. The dayside appears much cooler than the dark, airless reference model predicts. What is reducing its heating, or carrying energy elsewhere?

An atmosphere is a strong explanation, not a complete answer

An atmosphere can help in two ways. Circulating gas can transport heat towards the nightside. Reflective clouds can also send some incoming starlight back into space before it heats the planet. An atmosphere does not therefore follow a universal rule of always warming or always cooling a world. Its properties, and the region and layer being measured, matter. [1]

The authors find that heat redistribution alone cannot explain their result. Even efficient transport would still require substantial reflection. A highly reflective atmosphere with clouds is consequently a compelling possibility. [1]

They also considered an alternative without an atmosphere: an exceptionally reflective surface. In that model, the surface would need to reflect roughly 90% of incoming energy. That is difficult to reconcile with current understanding of lava surfaces, although better laboratory measurements of molten rock are still needed. [1]

The careful conclusion is that the observation provides strong evidence for an atmosphere. It is not a photograph of clouds, nor a complete chemical inventory. Rejecting one simple model is an important result, but it does not automatically settle every detail of the replacement.

What remains unknown, and why this matters

The current spectrum is not precise enough to establish which gases are present. Water vapour and carbon dioxide appear in models the researchers explored, but this study did not confirm them on HD 3167 b. Further observations, including with Webb’s NIRSpec instrument, could help distinguish between the possibilities. [1]

An atmosphere also does not mean life. This is an intensely hot world. Its scientific value lies in testing how rocky planets acquire, retain or lose gaseous envelopes under extreme irradiation.

There is a connection to home, too. The young Earth experienced extensive melting and magma-ocean conditions. A present-day lava world around another star is not an exact copy of early Earth, but it can help researchers investigate interactions between molten rock and an atmosphere. [2]

HD 3167 b offers more than an unusually short year. It shows how an almost imperceptible loss of radiation can become evidence about an entire world. We do not have to reach a planet to learn about it. We have to know what questions to ask of its light.

Sources

1. Coy et al. (2026): Evidence for an Atmosphere on the Ultra-Short Period super-Earth HD 3167 b, particularly Sections II, III.1–III.3 and IV. Published paper in The Astrophysical Journal Letters. The paper is available under CC BY 4.0; this article is an independently written explanation, not a reproduction of the paper.
2. University of Chicago, 23 September 2026: Study led by UChicago scientists finds atmosphere on the coldest lava world to date.
3. NASA: HD 3167 b — Exoplanet Catalog.
4. NASA: How We Find and Characterize Exoplanets.

Temperatures in Celsius are converted and rounded. The approximate 2,000°C reference value is derived from the published 1,300 K temperature and model ratio of 0.57. It is neither the equilibrium temperature nor a direct surface measurement. Facts checked on 13 October 2026.