A new study just corrected a 30-year measurement error. The fix points straight at a bigger question.
The Number That Didn't Add Up
The Sun appeared to be missing nearly half its silver. Against CI chondrites — primitive meteorites that preserve the original recipe of the solar system — the photosphere registered 0.25 dex less silver than expected. That gap made silver a 2-sigma outlier in the solar abundance data. It sat there, unexplained, for thirty years.
Nobody had a clean explanation. Nobody could agree on where the silver went.
Then a team from Uppsala University looked more carefully at how the measurement itself was being made. Their 2026 paper in Astronomy and Astrophysics did not discover new silver. It found the old tools were reading the wrong number.
The gap shrank from 0.25 dex to 0.06 dex. Most of the missing silver was never actually missing. It was a measurement problem.
For the Liquid Star Model, that matters in a way that goes beyond silver.
What Is a Dex?
Astronomers use "dex" constantly and almost never explain it. Here is what it means.
A dex is a unit on a logarithmic scale. The word comes from "decimal exponent." It counts how many powers of ten separate two numbers.
If one pile of sand holds 10 grains and another holds 100, the difference is 1 dex. Not 90 grains. Not ten times. One dex. Push the second pile to 1,000 grains and the difference becomes 2 dex. The scale compresses enormous spans into numbers a human can actually work with.
The universe needs this compression. The Sun holds vastly more hydrogen than silver. Writing those quantities as ordinary numbers would fill a page with zeros.
So when the paper says the silver gap shrank from 0.25 dex to 0.06 dex, it means the Sun went from appearing to hold roughly 56 percent of the expected silver to holding roughly 87 percent. That is a large correction from a single, better calculation.
Interesting fact: The dex scale works the same way as the earthquake magnitude scale. A magnitude 7 earthquake releases ten times more energy than a magnitude 6. A 1 dex difference in solar abundance means ten times more or less of a given element. Small numbers. Enormous real-world differences.
How Scientists Read the Sun's Chemistry
No spacecraft has scooped a sample from the Sun's photosphere. Scientists read its chemistry a different way.
When sunlight breaks through a prism or diffraction grating, it spreads into a rainbow of colors. Across that rainbow lie thousands of dark lines, each left by a specific element absorbing light at a precise wavelength. Silver leaves two such lines, at 328 and 338 nanometers, in the near-ultraviolet range just past what the human eye detects. The depth and width of those lines tell scientists how much silver the photosphere holds.
The catch is that translating line shape into abundance requires a model of the solar atmosphere. That model must account for temperature, pressure, and atomic behavior across many overlapping layers. Every simplification in the model pushes the final number further from the truth.
For most of the 20th century, solar physicists used one-dimensional models built on a state called local thermodynamic equilibrium, or LTE. LTE treats every layer of gas as perfectly balanced and isolated, radiating at a single stable temperature. Real stellar atmospheres behave nothing like that. The Sun's surface boils, churns, and drives photons outward through layers that are constantly out of equilibrium with each other.
The Better Tool
The Uppsala team, led by S. Caliskan and A.M. Amarsi, built what physicists call a model atom for silver — a quantum mechanical map of every energy state a silver atom occupies, every way it absorbs or emits a photon, and every collision it has with surrounding hydrogen. Nobody had ever built one for silver before.
Getting the silver lines right turned out to be harder than expected. The 328 nm line overlaps with a predicted iron feature at 328.0666 nm whose oscillator strength carries an uncertainty of up to 0.8 dex. The team averaged two equivalent-width measurements — one omitting the iron blend, one including it — and took half their difference as the uncertainty. That single methodological choice contributed 0.06 dex to the total error budget on the 328 nm line alone.
The full 3D simulation they ran accounts for the granular surface of the Sun, where hot plasma rises in bright patches and cooler material sinks in darker lanes. Those temperature contrasts change how silver atoms absorb light across the disk, and a 1D model misses them entirely.
Here is what the simulation found: in the real solar atmosphere, the two silver lines run weaker than LTE predicts. Weaker lines require more silver to produce them. The older models read less silver than was actually there. When the 3D non-LTE correction was applied, the silver abundance rose by 0.19 dex.
Question worth sitting with: Silver was underestimated by 0.19 dex because the atmospheric model was incomplete. How many other heavy elements in the solar spectrum sit on the same kind of miscalibrated scale — and in which direction do their corrections run?
What the Standard Model Says About the Gap
Six hundredths of a dex remain between the Sun's revised silver abundance and the meteoritic value. The standard solar model has an explanation for that residual.
Silver condenses from gas to solid at 699 Kelvin. That puts it in the moderately volatile category — easier to drive off than iron or calcium, harder to hold onto during the high-temperature chaos of early solar system formation. The standard explanation is straightforward: the Sun lost some silver during its formation. The meteorites formed in cooler outer-disk conditions and preserved more.
The paper supports this picture. With the new abundance, silver falls into line with other moderately volatile elements that have received equivalent 3D non-LTE treatment. The pattern holds.
That is a real result. It is not the only way to read the data.
What the Liquid Star Model Sees
The Liquid Star Model agrees the photosphere is plasma. Every solar model does. The photosphere is the bright outer layer where the Sun's light originates, and it is genuinely hot, ionized gas.
What the model proposes lies beneath that layer. Below the photosphere, at a depth suspected to reach approximately 200,000 kilometers where a structural transition called the tachocline sits, the Liquid Star Model places a liquid magma surface — rich in iron, nickel, and the other heavy metals the solar spectrum consistently detects. That buried layer, not the visible plasma above it, drives the chemical and nuclear activity the model proposes as the Sun's primary energy source.
The new silver measurement connects to this framework directly. For thirty years, standard spectroscopic models gave systematically low readings for silver in the photosphere. The 3D non-LTE correction raised the number toward reality. The Liquid Star Model has argued from its beginning that simpler models underestimate heavy element content near the solar surface — not from spectral line modeling, but from the proposed composition of the sub-photospheric liquid layer. Both arguments point in the same direction. The Sun's outer layers hold more heavy metals than older models showed.
One detail from the paper deserves attention here. The team found that 1D models with full non-LTE corrections came closer to the correct answer than 3D models using only LTE corrections. Accounting for quantum atomic behavior, in other words, mattered more than accounting for the Sun's three-dimensional surface structure.
Interesting fact: The team's sensitivity tests showed the model was most vulnerable to uncertainty in the hydrogen collision rates — not in the oscillator strengths or photoionisation data. A factor-of-ten change in hydrogen collision rates shifted the final silver abundance by 0.05 dex. That one input dominates the error budget.
The Silver-Copper Connection
Silver sits at atomic number 47. Copper sits at atomic number 29. Nickel sits at atomic number 28. These three metals occupy the range where a nuclear process called the weak r-process assembles heavy elements from lighter ones.
The Liquid Star Model's most directly testable prediction is that nickel at the sub-photospheric liquid surface absorbs hydrogen and converts to copper through low-energy nuclear reactions. Laboratory LENR experiments have documented this transformation in metallic nickel targets on Earth. At solar scale, running continuously for 4.5 billion years across a surface area 12,000 times that of Earth, the accumulated copper production would leave a fingerprint in the solar wind.
NASA's Genesis spacecraft collected solar wind particles over 27 months and returned them to Earth in 2004. Those samples sit in laboratories right now, available for analysis.
The new silver baseline sharpens the test. The Uppsala team showed that 3D non-LTE corrections raise heavy metal photospheric abundances by measurable amounts. Any copper enrichment in the Genesis samples now gets compared against a revised, more accurate standard model prediction rather than an older underestimate. A signal that might previously have been dismissed as within the old error range could now stand clearly above the corrected baseline.
The solar wind copper test just got more precise.
A Pattern Worth Watching
The 2026 silver paper is not an isolated result. In 2025, the same Uppsala group published a revised sulfur abundance using the same 3D non-LTE approach. Sulfur also came back higher than older measurements showed. Silver follows the same pattern one year later.
Every element that receives a full 3D non-LTE analysis moves toward the meteoritic value. The ones that remain as outliers are almost always the ones still waiting for this treatment.
The Liquid Star Model does not claim credit for this trend. These corrections come from mainstream spectroscopic methods, not from the model's proposals about sub-photospheric structure. But the direction of correction — consistently upward, toward higher heavy metal content — is exactly what a metal-rich liquid layer beneath the photosphere would produce if it were influencing surface composition over geological time.
Patterns this consistent across independent investigations are worth following.
What Comes Next
The main remaining limitation in the silver measurement is the hydrogen collision rate data. Those rates carry an uncertainty of roughly a factor of ten, contributing about 0.05 dex to the total error on the final abundance. That single input now dominates everything else in the error budget.
The SUNRISE UV Spectropolarimeter and Imager, scheduled for future solar observations, could provide the high-resolution centre-to-limb intensity data needed to empirically calibrate the iron blend at 328.0666 nm and push the total uncertainty below 0.05 dex. At that precision, the residual 0.06 dex gap between the Sun and meteorites either disappears into the error bars or holds as a genuine signal.
If it holds, the question of why the Sun is still slightly short on silver becomes the next investigation — and the Liquid Star Model's sub-photospheric chemistry enters that conversation as a live candidate.
The Sun did not lose its silver. The old ruler was short. Switching to a better one changed the reading.
The question now is what else the better ruler will find.