Preprint tool EoSAlign tries to put diverging megabar pressure scales on one ruler

A preprint posted to arXiv by Sang-Heon Shim introduces EoSAlign, open-source software meant to calculate, compare, and align pressure in diamond-anvil and related extreme-conditions experiments. The problem it targets is not new: published pressure calibrations, from ruby luminescence to metal and oxide equations of state, disagree, and the abstract states that above 100 GPa those scales can diverge by more than 10 GPa. EoSAlign takes laboratory observables — unit-cell volume in cubic angstroms, Raman shift in inverse centimeters, or a luminescence peak in nanometers — and converts them through published calibrations so results can be placed on a common scale.

Methods note: this is a software and reanalysis preprint, not a new high-pressure experiment; the abstract does not report a sample size, and the full methods were not independently re-run here. What the authors say they did is fit and align equations of state for bridgmanite and its high-pressure form, post-perovskite, onto one pressure scale, then use that alignment to shrink differences that come from the choice of calibrant rather than from the mineral itself. That is the right target. Bridgmanite dominates Earth's lower mantle, and the post-perovskite boundary sits near 120 GPa, exactly where scale disagreement is largest.

The correction a software-launch story misses is that alignment is not absolute accuracy. Fei and colleagues, in a 2007 PNAS paper, argued for an internally consistent scale because ruby, gold, platinum, and MgO calibrations pulled the same experiment to different pressures. EoSAlign can make two labs speak one language; it cannot decide which language is true. Shock-reduced isotherms, the thermal-pressure model, and nonhydrostatic stress still set the zero point. A greater-than-10 GPa offset above 100 GPa can move a melting curve or the D-double-prime boundary by an amount later papers have treated as a physical result.

What would strengthen the claim is a public benchmark: the same raw volumes and peak positions run through EoSAlign and through independent codes for the Dewaele, Fei, and Dorogokupets scales, with residuals published. Until the calibration library and versioned defaults are inspectable, open source is a promise in an abstract. The near-term test is practical. If groups re-reduce existing megabar datasets onto one named scale and post-perovskite boundary scatter shrinks without new experiments, the tool has earned its place. If scatter does not shrink, the dominant error was never the spreadsheet.