Somewhere around 50 light-years from Earth, a dim red dwarf star called LHS 1903 hosts four planets arranged in an order that, according to decades of planet formation theory, should not be possible. The innermost world is rocky. The two middle planets are swollen with thick hydrogen-helium atmospheres. And the outermost world, the one farthest from the star where gas should be easiest to hold onto, is rocky again.
That sequence, confirmed through observations by the European Space Agency’s CHEOPS space telescope and published in the journal Science in June 2025, upends one of the most reliable patterns in exoplanet science: rocky planets close in, gas-rich planets farther out. The discovery is now pushing theorists to rethink how planets assemble around the most common type of star in the galaxy.
A solar system running in reverse
In our own solar system, the architecture follows a tidy gradient. Mercury, Venus, Earth, and Mars sit close to the Sun, all made of rock and metal. Jupiter, Saturn, Uranus, and Neptune occupy the outer reaches, bloated with gas and ice they captured from the cooler, volatile-rich regions of the disk that birthed them. Thousands of confirmed exoplanets around other stars have broadly reinforced that pattern. LHS 1903 breaks it.
The research team, led by Luisa Maria Serrano of the University of Bern, measured each planet’s size using CHEOPS transit photometry and pinned down masses with radial velocity data from ground-based telescopes. Combining those two measurements yields bulk density, the single best indicator of whether a planet is rocky or gas-rich. Their results are unambiguous: LHS 1903 b (the innermost) and LHS 1903 e (the outermost) are dense, rocky bodies, while LHS 1903 c and d, sandwiched between them, have densities low enough to require substantial gaseous envelopes.
To put the sizes in perspective, the two rocky worlds are roughly Earth-sized, while the gaseous middle planets are closer to mini-Neptunes, a few times Earth’s radius and far puffier. All four orbit much closer to their star than Mercury does to the Sun, completing their years in a matter of days. But even within that compact arrangement, the density flip between the middle and outer planets is stark and statistically significant.
Why this layout is so hard to explain
Standard planet formation theory, known as core accretion, holds that solid cores grow first, then sweep up gas from the surrounding protoplanetary disk. Farther from the star, the disk is cooler, gas lingers longer, and growing planets have more time and more raw material to build thick atmospheres. That is why gas giants tend to form in the outer disk and rocky worlds end up closer in.
LHS 1903 e defies that logic. It sits beyond two planets that successfully captured gas, yet it appears to have none. The leading explanation, outlined in the Science paper and in an accompanying open-access preprint, is that the outermost planet formed late. Red dwarfs like LHS 1903 go through an extended, luminous adolescence before settling into their long, quiet main-sequence lives. During that hyperactive youth, intense radiation and stellar winds may have blown residual gas out of the outer disk before LHS 1903 e finished growing its core. By the time the planet was massive enough to grab an atmosphere, there was no atmosphere left to grab.
But that hypothesis rests on knowing the star’s age with some precision, and the published data do not yet nail it down tightly enough to rule out alternatives. Other possibilities include late-stage atmospheric stripping, where high-energy radiation from the star gradually boiled away a once-thick envelope, or inward migration of the gaseous planets from originally wider orbits, which would mean the current arrangement is not the one in which the planets formed.
What the data can and cannot tell us
The density contrast between the middle and outer planets is the strongest piece of evidence, and it comes from two independent measurement techniques. Radius is derived from how much starlight each planet blocks during a transit. Mass comes from the tiny gravitational tug each planet exerts on the star, detected as a wobble in the star’s spectrum. When both methods converge on a gas-poor composition for LHS 1903 e while confirming gaseous envelopes for c and d, the finding is about as robust as exoplanet characterization gets with current instruments.
The peer-reviewed publication record confirms the full author list and institutional affiliations, showing that the collaboration spans multiple observatories and research groups across Europe and beyond. That breadth matters: it means the result does not hinge on a single telescope’s calibration or a single team’s data pipeline.
What the data cannot yet reveal is the chemical makeup of the gaseous envelopes around planets c and d. Transmission spectroscopy with the James Webb Space Telescope could, in principle, probe those atmospheres and test whether the gas they retained matches what formation models predict for a disk that was running out of material. No such observations have been announced as of June 2026, but the system is a natural candidate for future proposals.
Transit timing variations, subtle shifts in when each planet crosses the star caused by gravitational nudges from its neighbors, could also expose hidden dynamics. If a fifth, non-transiting planet lurks in the system, its gravitational fingerprint might explain some of the architectural oddities. The existing CHEOPS data may already contain hints, but extracting them will require careful reanalysis.
What LHS 1903 means for the bigger picture
Red dwarfs account for roughly 70 percent of all stars in the Milky Way, and they host a disproportionate share of the exoplanets discovered so far, largely because their small size makes transiting planets easier to detect. Many of those worlds sit in compact, multi-planet configurations broadly similar to LHS 1903. If a rocky outer planet can emerge beyond gas-rich neighbors in one such system, similar architectures may be hiding in existing survey data, overlooked because nobody expected to find them.
For theorists, the system acts as a stress test. Any successful formation model must now reproduce not just individual planet masses and radii but also their relative ordering and spacing. That means tracking how disk temperature, ionization, and turbulence evolve over millions of years, how quickly planetary cores grow at different distances, and how stellar radiation preserves or destroys gaseous envelopes. Subtle differences in formation timing, migration history, or local disk structure could all play roles, and simulations will need to explore each one.
For observers, the next steps are concrete. Tighter radial velocity monitoring could shrink the error bars on LHS 1903 e’s mass, testing whether it is truly as rocky as current data suggest or whether a thin volatile layer remains possible. Spectroscopic follow-up of the gaseous middle planets could reveal whether they formed from the same reservoir of material or captured gas at different stages of the disk’s life. And broader surveys of red dwarf systems, armed with the knowledge that this configuration exists, can start looking for siblings.
A pattern the textbooks did not predict
Planet formation has always been messier than the clean diagrams in introductory astronomy courses suggest. Hot Jupiters, planets in retrograde orbits, and worlds squeezed into resonant chains have all forced revisions to the standard story over the past three decades. LHS 1903 is the latest entry in that tradition, and perhaps the most pointed. It does not just add a new type of planet to the catalog. It scrambles the expected order of an entire system, placing a rocky world exactly where theory says it has no business being.
Whether LHS 1903 turns out to be a rare fluke or the first recognized member of a hidden population, it has already done its job: it has shown that the rules governing how planets arrange themselves around stars are less settled than most astronomers assumed. The next generation of observations will determine which of those rules need rewriting.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
