GJ 523b gives 'Mega-Earth' a number: 23 Earth masses inside 2.5 Earth radii

A group at the University of Wisconsin-Madison has characterized GJ 523b, a planet carrying 23 times Earth's mass inside a radius about 2.5 times Earth's, and has posted the paper to arXiv while it is under review ^[2][1][4]. It is the first exoplanet discovered and cataloged by researchers at the Wisconsin Center for Origins Research, and its usefulness is arithmetic rather than aesthetic: a label astronomers have used loosely for over a decade now has a specific object attached to it ^[5][6].

Take the two headline numbers at face value and the consequences follow directly. Twenty-three Earth masses in 2.5 Earth radii is a bulk density about 1.5 times Earth's ^[21], and a surface gravity about 3.7 times Earth's ^[22]. The team reports the interior as mostly dense rock with a massive core, at roughly 60 percent of Neptune's size ^[10]. "Dense planets like this aren't uncommon, but they're usually small rocky planets similar to Earth or Mercury," says lead author Max Kroft, a graduate student in the lab of assistant professor of astronomy Thomas Beatty ^[15][3].

The formation problem is the interesting part. The standard sequence has a rocky, metal-rich core accreting a hydrogen-dominated envelope, and the paper's framing notes that Jupiter and Saturn acquired their envelopes once they reached roughly 20 Earth's worth of material ^[16]. GJ 523b is past that threshold and did not ^[17]. Two candidate explanations appear in the paper: the atmosphere was stripped when the planet orbited too close to its star, or the object began as two planets that collided and lost the gas in the heat of the impact ^[18]. "A planet can't hold on to its atmosphere if it's really hot, and so you could be left with this big glob of rock made by these two planets with very little atmosphere," Kroft says ^[19]. Note that the popular account slides between mass and size at the 20-Earth mark; the comparison only holds in mass ^[16][17].

The paper's more durable contribution is a concession. According to Beatty, defining a Mega-Earth "isn't something we astronomers can really do by ourselves: We need geologists who understand how iron and rock behave at pressures no laboratory on Earth can reach, and atmospheric scientists who can tell us how much of what we measured is rock at all" ^[7]. That is the right order of operations. A mass and a radius give a mean density and nothing more; whether that average describes rock throughout, or rock plus a thin light envelope inflating the measured radius, is a modeling question with an equation-of-state input that no terrestrial lab currently supplies ^[7].

The detection path was conventional: a TESS candidate, then follow-up with the WIYN telescope in Arizona and its high-resolution spectrograph, plus James Webb data for density and atmosphere ^[13][14][11]. That pipeline is where the supply of comparison objects sits. TESS has flagged more than 8,000 candidates and fewer than a quarter are confirmed ^[12], so under roughly 2,000 have been validated ^[23]. Kroft expects more dense, oversized outliers to emerge from the backlog ^[24].

Worth watching: whether review holds the 23-mass figure, whether Webb spectra reveal a light envelope that would shrink the rocky fraction, and whether WiCOR, which launched in 2024 across seven UW-Madison departments and was hunting Hycean worlds when it found this one, gets a second data point ^[8][9][10]. The reported system age of nearly 170 million years is a constraint the loss scenarios will have to fit ^[20].