Spite Plateau in Metal-Poor Stars

• Spite Plateau is the observed constant lithium abundance (~2.1–2.4 dex) in warm, metal-poor dwarf stars, questioning standard Big Bang predictions.
• Spectral synthesis of the Li I 6707.8 Å line, enhanced by NLTE and 3D corrections, refines measurements and uncovers deviations at extreme low metallicities.
• Its universality across Galactic and extragalactic environments and consistency with stellar-depletion models inform debates on primordial lithium and transport processes.

The Spite Plateau is the near-constant lithium abundance observed in warm, metal-poor halo dwarf stars, first identified by Spite and Spite in 1982. In standard abundance notation, lithium is written as $A(\mathrm{Li})=\log_{10}(N_{\rm Li}/N_{\rm H})+12$, while metallicity is measured as $$[\mathrm{Fe}/\mathrm{H}]=\log_{10}\!\big((N_{\rm Fe}/N_{\rm H})_{\rm star}\big)-\log_{10}\!\big((N_{\rm Fe}/N_{\rm H})_{\odot}\big)$$. Historically, the plateau is associated with unevolved Population II dwarfs over a broad low-metallicity interval, commonly around $-3<[{\rm Fe/H}]<-1$, with observed $A(\mathrm{Li})$ values near $2.1$--$2.4$ dex. Its significance derives from the fact that these photospheric abundances are substantially lower than the primordial lithium abundance predicted by standard Big Bang nucleosynthesis, creating the cosmological lithium problem (Simpson et al., 2020).

1. Definition, notation, and the cosmological discrepancy

In its classical form, the Spite Plateau denotes the empirical flatness of lithium abundances in old, warm, metal-poor dwarf stars. The plateau is usually defined from stars hot enough to avoid strong convective depletion and unevolved enough to preserve a relatively simple surface-abundance history. Several compilations place it at $A(\mathrm{Li})\approx2.1$--$2.3$ dex, with small intrinsic scatter, over metallicities extending at least to $[\mathrm{Fe}/\mathrm{H}]\lesssim-1$ (Matteucci et al., 2021, Mucciarelli et al., 2022).

The cosmological importance of the plateau is set by its offset from standard Big Bang nucleosynthesis. One homogeneous GALAH-based analysis adopts $A(\mathrm{Li})_{\rm BBN}=2.75\pm0.02$ and measures a plateau near $$[\mathrm{Fe}/\mathrm{H}]=\log_{10}\!\big((N_{\rm Fe}/N_{\rm H})_{\rm star}\big)-\log_{10}\!\big((N_{\rm Fe}/N_{\rm H})_{\odot}\big)$$0 in warm metal-poor dwarfs, implying a discrepancy of about $$[\mathrm{Fe}/\mathrm{H}]=\log_{10}\!\big((N_{\rm Fe}/N_{\rm H})_{\rm star}\big)-\log_{10}\!\big((N_{\rm Fe}/N_{\rm H})_{\odot}\big)$$1 dex (Simpson et al., 2020). Other summaries place the primordial prediction in the range $$[\mathrm{Fe}/\mathrm{H}]=\log_{10}\!\big((N_{\rm Fe}/N_{\rm H})_{\rm star}\big)-\log_{10}\!\big((N_{\rm Fe}/N_{\rm H})_{\odot}\big)$$2--$$[\mathrm{Fe}/\mathrm{H}]=\log_{10}\!\big((N_{\rm Fe}/N_{\rm H})_{\rm star}\big)-\log_{10}\!\big((N_{\rm Fe}/N_{\rm H})_{\odot}\big)$$3, again about a factor of $$[\mathrm{Fe}/\mathrm{H}]=\log_{10}\!\big((N_{\rm Fe}/N_{\rm H})_{\rm star}\big)-\log_{10}\!\big((N_{\rm Fe}/N_{\rm H})_{\odot}\big)$$4 above the plateau in linear $$[\mathrm{Fe}/\mathrm{H}]=\log_{10}\!\big((N_{\rm Fe}/N_{\rm H})_{\rm star}\big)-\log_{10}\!\big((N_{\rm Fe}/N_{\rm H})_{\odot}\big)$$5 (Matteucci et al., 2021). The basic observational fact is therefore robust even though the quoted plateau zero-point varies with sample definition, temperature scale, atmosphere model, and NLTE treatment.

A persistent question is whether the plateau should be interpreted as a preserved primordial abundance or as a depleted surface value. Several of the cited studies explicitly reject the first interpretation. Mass-dependent depletion, atomic diffusion, turbulent mixing, pre-main-sequence processing, and related transport physics have all been advanced as mechanisms capable of lowering surface lithium while preserving a narrow plateau-like morphology (Melendez et al., 2010, Fu et al., 2015, Borisov et al., 2024).

2. Empirical characterization in halo dwarfs

Observationally, plateau work is anchored in the Li I resonance doublet at $$[\mathrm{Fe}/\mathrm{H}]=\log_{10}\!\big((N_{\rm Fe}/N_{\rm H})_{\rm star}\big)-\log_{10}\!\big((N_{\rm Fe}/N_{\rm H})_{\odot}\big)$$6, often quoted as $$[\mathrm{Fe}/\mathrm{H}]=\log_{10}\!\big((N_{\rm Fe}/N_{\rm H})_{\rm star}\big)-\log_{10}\!\big((N_{\rm Fe}/N_{\rm H})_{\odot}\big)$$7. Modern analyses treat this line with spectral synthesis and increasingly with NLTE or 3D NLTE radiative transfer, because neutral lithium is a trace species whose line formation is sensitive to the radiation field and to atmospheric structure (Simpson et al., 2020, Wang et al., 2021).

Different studies report somewhat different plateau values because they adopt different stellar selections. A broad summary of representative determinations is useful.

Regime	Selection	Reported $$[\mathrm{Fe}/\mathrm{H}]=\log_{10}\!\big((N_{\rm Fe}/N_{\rm H})_{\rm star}\big)-\log_{10}\!\big((N_{\rm Fe}/N_{\rm H})_{\odot}\big)$$8
Classical warm halo dwarfs	$$[\mathrm{Fe}/\mathrm{H}]=\log_{10}\!\big((N_{\rm Fe}/N_{\rm H})_{\rm star}\big)-\log_{10}\!\big((N_{\rm Fe}/N_{\rm H})_{\odot}\big)$$9 and $-3<[{\rm Fe/H}]<-1$0	$-3<[{\rm Fe/H}]<-1$1--$-3<[{\rm Fe/H}]<-1$2 dex
GALAH warm dwarfs	$-3<[{\rm Fe/H}]<-1$3 and $-3<[{\rm Fe/H}]<-1$4	$-3<[{\rm Fe/H}]<-1$5
“Broken plateau” high-metallicity branch	$-3<[{\rm Fe/H}]<-1$6	$-3<[{\rm Fe/H}]<-1$7 with $-3<[{\rm Fe/H}]<-1$8
“Broken plateau” low-metallicity branch	$-3<[{\rm Fe/H}]<-1$9	$A(\mathrm{Li})$0 with $A(\mathrm{Li})$1

The GALAH DR3 study used 588,571 stars observed with HERMES at $A(\mathrm{Li})$2, selected dwarfs with $A(\mathrm{Li})$3, reliable Gaia astrometry, $A(\mathrm{Li})$4 in the red arm, and zero-quality flags for the stellar and lithium solutions. Among 251 warm dwarfs with $A(\mathrm{Li})$5, $A(\mathrm{Li})$6, and measured lithium, the overall plateau is $A(\mathrm{Li})$7 with dispersion $A(\mathrm{Li})$8 (Simpson et al., 2020). In that dataset the plateau is described as “basically flat” with $A(\mathrm{Li})$9 over the GALAH metallicity range.

A different high-precision study of 88 halo stars found that the plateau is not a single constant level but is “broken” into two flat regimes separated near $2.1$0. For “less-depleted” stars selected with a metallicity-dependent temperature cutoff, it reported $2.1$1 with $2.1$2 below $2.1$3 and $2.1$4 with $2.1$5 above that break, with slopes consistent with zero within each regime (Melendez et al., 2010). This directly counters older claims of a monotonic decline across the full metallicity range, because those claims can arise when stars from both regimes are combined or when the temperature scale is biased.

The empirical location of the plateau also depends strongly on the adopted $2.1$6 scale. A later critique emphasizes that literature $2.1$7 scales can differ by as much as $2.1$8--$2.1$9, which propagates into $2.4$0 shifts of order $2.4$1--$2.4$2 dex. After homogenization to an IRFM scale, that study obtained a high-quality plateau estimate of $2.4$3 with $2.4$4 for 35 stars in $2.4$5 (Norris et al., 2023). This illustrates that the existence of a plateau is not in dispute, whereas its exact zero-point remains method-dependent.

3. Universality across Galactic and extragalactic environments

A major recent development is the use of accreted stellar populations to test whether the plateau depends on galactic formation environment. The GALAH analysis of Gaia-Enceladus-Sausage (GSE) stars used orbital actions to isolate 93 GSE dwarfs and compared them to retrograde halo, prograde halo, and disk samples. In the plateau regime defined by $2.4$6 and $2.4$7, the subsample means are statistically indistinguishable: GSE, $2.4$8; retrograde halo, $2.4$9; prograde halo, $A(\mathrm{Li})\approx2.1$0; disk, $A(\mathrm{Li})\approx2.1$1 (Simpson et al., 2020). The differences are no larger than $A(\mathrm{Li})\approx2.1$2 dex, much smaller than the intrinsic dispersions.

This result is astrophysically important because the GSE stars formed in an extra-Galactic environment and are chemically distinguishable from in situ Milky Way populations; more metal-rich GSE stars are $A(\mathrm{Li})\approx2.1$3-poor relative to in situ disk stars, yet their lithium lies on the same plateau (Simpson et al., 2020). The study therefore argues that the lithium problem is not a consequence of formation environment.

Chemical-evolution modeling of dwarf spheroidal and ultra-faint dwarf galaxies reaches a similar conclusion from a different direction. In those models, lithium is produced mainly by novae and cosmic rays, with smaller contributions from low- and intermediate-mass stars, and the system evolves according to

$A(\mathrm{Li})\approx2.1$4

When compared with observations, the upper envelopes in the Milky Way halo, Sculptor, Sagittarius/M54, $A(\mathrm{Li})\approx2.1$5 Centauri, Gaia–Enceladus, and the S2/Slygr streams define the same plateau level, suggesting that the Spite Plateau could be a universal feature (Matteucci et al., 2021). In that framework, late lithium production from novae and cosmic rays occurs too late to affect most metal-poor stars in dwarf systems, so environmental differences in galactic chemical evolution do not naturally generate different plateau levels.

A plausible implication is that lithium is a poor discriminator between in situ and accreted origin at low metallicity. That implication is explicitly drawn in the dwarf-galaxy modeling paper, which notes that mixed in-situ and ex-situ halo stars all lie on the same plateau (Matteucci et al., 2021).

4. Breakdown, meltdown, and chemically peculiar departures

Although the plateau is thin over much of parameter space, departures from a single universal level occur at the lowest metallicities. One major observational theme is the “meltdown” below about $A(\mathrm{Li})\approx2.1$6 to $A(\mathrm{Li})\approx2.1$7. In metal-poor dwarfs, this refers to both a decrease in mean lithium and a sharp increase in star-to-star scatter (Mucciarelli et al., 2022, Sbordone et al., 2010).

A VLT/UVES study of 28 halo dwarfs with $A(\mathrm{Li})\approx2.1$8 established that below $A(\mathrm{Li})\approx2.1$9 the plateau does not persist as a simple constant. It found a positive correlation of $2.3$0 with $2.3$1, with slopes of about $2.3$2--$2.3$3 dex per dex depending on the temperature scale, and showed that the scatter roughly doubles below $2.3$4 (Sbordone et al., 2010). An especially striking result is that no star with $2.3$5 lies above the plateau; the distribution fans out only toward lower $2.3$6.

A detailed chemical investigation of 11 very metal-poor dwarfs reinforced this picture. Excluding strong outliers, that sample yielded $2.3$7, but also included a low-lithium star at $2.3$8 with $2.3$9 and a likely blue straggler with only an upper limit $[\mathrm{Fe}/\mathrm{H}]\lesssim-1$0 (Pinto et al., 2021). Its central result is that lithium-poor stars do not show anomalies in other measured elements, which argues against a simple interpretation in terms of unusual natal abundances of many species.

One recent critique interprets the low-metallicity behavior in terms of chaotic early star formation and astration of primordial lithium. After IRFM homogenization, that work reported decreasing mean $[\mathrm{Fe}/\mathrm{H}]\lesssim-1$1 and increasing dispersion toward lower metallicity, with $[\mathrm{Fe}/\mathrm{H}]\lesssim-1$2 and $[\mathrm{Fe}/\mathrm{H}]\lesssim-1$3 in the highest halo-metallicity bin, but $[\mathrm{Fe}/\mathrm{H}]\lesssim-1$4 and $[\mathrm{Fe}/\mathrm{H}]\lesssim-1$5 in the lowest bin (Norris et al., 2023). It proposed a toy model in which C-rich, Li-poor early gas coexists and coalesces with later C-normal gas, producing the observed “meltdown.” That is an interpretation rather than a consensus result, but it frames one current line of argument.

At the opposite extreme, a handful of lithium-rich metal-poor dwarfs are known. In the GALAH GSE and retrograde-halo samples, four such stars have $[\mathrm{Fe}/\mathrm{H}]\lesssim-1$6--$[\mathrm{Fe}/\mathrm{H}]\lesssim-1$7 (Simpson et al., 2020). Proposed origins in that study include post-formation accretion from an AGB companion, ISM accretion, or early novae contributions, but the reported $[\mathrm{Fe}/\mathrm{H}]\lesssim-1$8-process abundances do not show a uniform correlation.

5. Stellar-depletion models and related physical interpretations

The dominant line of explanation in the cited literature is that the Spite Plateau is a depleted surface abundance rather than an undepleted primordial value. One influential study finds significant mass-dependent depletion among plateau stars and shows that models with atomic diffusion plus turbulent mixing reproduce both the plateau structure and the inferred primordial abundance. In that work, the best agreement is obtained with a T6.25 turbulent-mixing prescription and an initial $[\mathrm{Fe}/\mathrm{H}]\lesssim-1$9 using MARCS atmospheres, or $A(\mathrm{Li})_{\rm BBN}=2.75\pm0.02$0 with Kurucz overshooting atmospheres (Melendez et al., 2010).

Globular clusters provide an especially strong test because stars with the same birth composition can be observed across multiple evolutionary phases. A synthesis of fifteen years of cluster work reports systematic heavy-element abundance differences between turn-off stars and giants, ranging from about $A(\mathrm{Li})_{\rm BBN}=2.75\pm0.02$1 dex in M4 to about $A(\mathrm{Li})_{\rm BBN}=2.75\pm0.02$2 dex in M30, with turn-off stars generally more depleted than giants (Korn, 2021). The paper argues that any stellar-physics solution to the lithium problem must reproduce these heavy-element trends as well as lithium. Atomic diffusion moderated by additional mixing does so, and diffusion-corrected birth-cloud lithium abundances around $A(\mathrm{Li})_{\rm BBN}=2.75\pm0.02$3 are inferred, leaving only a small nominal offset to $A(\mathrm{Li})_{\rm BBN}=2.75\pm0.02$4 from BBN (Korn, 2021).

A complementary constraint comes from lower red giant branch stars. A sample of 58 LRGB stars spanning roughly $A(\mathrm{Li})_{\rm BBN}=2.75\pm0.02$5 exhibits a thin lithium plateau at $A(\mathrm{Li})_{\rm BBN}=2.75\pm0.02$6--$A(\mathrm{Li})_{\rm BBN}=2.75\pm0.02$7 with $A(\mathrm{Li})_{\rm BBN}=2.75\pm0.02$8--$A(\mathrm{Li})_{\rm BBN}=2.75\pm0.02$9, but no sign of the dwarf-star meltdown (Mucciarelli et al., 2022). Standard models with only convection or diffusion cannot reproduce this from $$[\mathrm{Fe}/\mathrm{H}]=\log_{10}\!\big((N_{\rm Fe}/N_{\rm H})_{\rm star}\big)-\log_{10}\!\big((N_{\rm Fe}/N_{\rm H})_{\odot}\big)$$00, but models including additional main-sequence transport can. The same mixing efficiency that explains the dwarf plateau also produces an almost flat LRGB lithium distribution. This strongly favors a stellar-transport solution and implies that the dwarf meltdown need not reflect a true drop in birth lithium.

Other proposed mechanisms remain active in the literature. Pre-main-sequence overshooting plus residual accretion regulated by EUV photo-evaporation can reproduce a plateau at $$[\mathrm{Fe}/\mathrm{H}]=\log_{10}\!\big((N_{\rm Fe}/N_{\rm H})_{\rm star}\big)-\log_{10}\!\big((N_{\rm Fe}/N_{\rm H})_{\odot}\big)$$01 for $$[\mathrm{Fe}/\mathrm{H}]=\log_{10}\!\big((N_{\rm Fe}/N_{\rm H})_{\rm star}\big)-\log_{10}\!\big((N_{\rm Fe}/N_{\rm H})_{\odot}\big)$$02--$$[\mathrm{Fe}/\mathrm{H}]=\log_{10}\!\big((N_{\rm Fe}/N_{\rm H})_{\rm star}\big)-\log_{10}\!\big((N_{\rm Fe}/N_{\rm H})_{\odot}\big)$$03 stars starting from $$[\mathrm{Fe}/\mathrm{H}]=\log_{10}\!\big((N_{\rm Fe}/N_{\rm H})_{\rm star}\big)-\log_{10}\!\big((N_{\rm Fe}/N_{\rm H})_{\odot}\big)$$04 (Fu et al., 2015). Cold protostellar accretion with very small seed mass and radius can also deplete lithium, but only in a restricted and fine-tuned region of parameter space; the same study finds negligible depletion for $$[\mathrm{Fe}/\mathrm{H}]=\log_{10}\!\big((N_{\rm Fe}/N_{\rm H})_{\rm star}\big)-\log_{10}\!\big((N_{\rm Fe}/N_{\rm H})_{\odot}\big)$$05 stars under conditions that strongly deplete $$[\mathrm{Fe}/\mathrm{H}]=\log_{10}\!\big((N_{\rm Fe}/N_{\rm H})_{\rm star}\big)-\log_{10}\!\big((N_{\rm Fe}/N_{\rm H})_{\odot}\big)$$06 stars (Cassisi et al., 2020). Mass-loss models with rates of about $$[\mathrm{Fe}/\mathrm{H}]=\log_{10}\!\big((N_{\rm Fe}/N_{\rm H})_{\rm star}\big)-\log_{10}\!\big((N_{\rm Fe}/N_{\rm H})_{\odot}\big)$$07 can mimic turbulence models and match some observations, but the required rates are described as unlikely compared with the solar value (Vick et al., 2013). More recent rotating models with gravitational settling, diffusion, rotation, and magnetic fields reproduce both a dwarf plateau at $$[\mathrm{Fe}/\mathrm{H}]=\log_{10}\!\big((N_{\rm Fe}/N_{\rm H})_{\rm star}\big)-\log_{10}\!\big((N_{\rm Fe}/N_{\rm H})_{\odot}\big)$$08--$$[\mathrm{Fe}/\mathrm{H}]=\log_{10}\!\big((N_{\rm Fe}/N_{\rm H})_{\rm star}\big)-\log_{10}\!\big((N_{\rm Fe}/N_{\rm H})_{\odot}\big)$$09 and an RGB plateau near $$[\mathrm{Fe}/\mathrm{H}]=\log_{10}\!\big((N_{\rm Fe}/N_{\rm H})_{\rm star}\big)-\log_{10}\!\big((N_{\rm Fe}/N_{\rm H})_{\odot}\big)$$10 dex, again with an initial abundance compatible with BBN (Yang et al., 26 Feb 2026). STAREVOL calculations similarly calibrate a parametric turbulence law with $$[\mathrm{Fe}/\mathrm{H}]=\log_{10}\!\big((N_{\rm Fe}/N_{\rm H})_{\rm star}\big)-\log_{10}\!\big((N_{\rm Fe}/N_{\rm H})_{\odot}\big)$$11 and recover a nearly constant plateau from an initial $$[\mathrm{Fe}/\mathrm{H}]=\log_{10}\!\big((N_{\rm Fe}/N_{\rm H})_{\rm star}\big)-\log_{10}\!\big((N_{\rm Fe}/N_{\rm H})_{\odot}\big)$$12 (Borisov et al., 2024).

Taken together, these studies do not identify a single unique depletion mechanism, but they converge on the view that stellar transport and depletion physics can produce the observed plateau while starting from a primordial lithium abundance close to the BBN prediction.

6. Spectroscopic diagnostics, isotopes, and unresolved controversies

The dominant abundance diagnostic is the Li I $$[\mathrm{Fe}/\mathrm{H}]=\log_{10}\!\big((N_{\rm Fe}/N_{\rm H})_{\rm star}\big)-\log_{10}\!\big((N_{\rm Fe}/N_{\rm H})_{\odot}\big)$$13 doublet, but its interpretation is not trivial. NLTE corrections are standard in current work, and 3D NLTE profile fitting has become central for both absolute abundance work and isotopic studies (Simpson et al., 2020, Wang et al., 2021). The sensitivity of $$[\mathrm{Fe}/\mathrm{H}]=\log_{10}\!\big((N_{\rm Fe}/N_{\rm H})_{\rm star}\big)-\log_{10}\!\big((N_{\rm Fe}/N_{\rm H})_{\odot}\big)$$14 to temperature scale, reddening, and model atmosphere remains a major source of systematic uncertainty.

An important alternative line-formation hypothesis is chromospheric back-radiation. In a model with a thin gray chromospheric slab, enhanced UV irradiation overionizes Li I, weakens the $$[\mathrm{Fe}/\mathrm{H}]=\log_{10}\!\big((N_{\rm Fe}/N_{\rm H})_{\rm star}\big)-\log_{10}\!\big((N_{\rm Fe}/N_{\rm H})_{\odot}\big)$$15 line by a factor of about $$[\mathrm{Fe}/\mathrm{H}]=\log_{10}\!\big((N_{\rm Fe}/N_{\rm H})_{\rm star}\big)-\log_{10}\!\big((N_{\rm Fe}/N_{\rm H})_{\odot}\big)$$16--$$[\mathrm{Fe}/\mathrm{H}]=\log_{10}\!\big((N_{\rm Fe}/N_{\rm H})_{\rm star}\big)-\log_{10}\!\big((N_{\rm Fe}/N_{\rm H})_{\odot}\big)$$17, and can lower the inferred abundance by about $$[\mathrm{Fe}/\mathrm{H}]=\log_{10}\!\big((N_{\rm Fe}/N_{\rm H})_{\rm star}\big)-\log_{10}\!\big((N_{\rm Fe}/N_{\rm H})_{\odot}\big)$$18--$$[\mathrm{Fe}/\mathrm{H}]=\log_{10}\!\big((N_{\rm Fe}/N_{\rm H})_{\rm star}\big)-\log_{10}\!\big((N_{\rm Fe}/N_{\rm H})_{\odot}\big)$$19 dex when analyzed with conventional atmospheres (Takeda, 2019). That study further predicts a metallicity-dependent trend and calls for UV spectrophotometry as the critical test. This does not deny the plateau observationally; rather, it contests whether the standard line-formation analysis recovers the true atmospheric lithium abundance.

A separate controversy concerned the presence of $$[\mathrm{Fe}/\mathrm{H}]=\log_{10}\!\big((N_{\rm Fe}/N_{\rm H})_{\rm star}\big)-\log_{10}\!\big((N_{\rm Fe}/N_{\rm H})_{\odot}\big)$$20 in Spite-plateau stars. Ultra-stable ESPRESSO spectra analyzed with 3D NLTE radiative transfer find no $$[\mathrm{Fe}/\mathrm{H}]=\log_{10}\!\big((N_{\rm Fe}/N_{\rm H})_{\rm star}\big)-\log_{10}\!\big((N_{\rm Fe}/N_{\rm H})_{\odot}\big)$$21 in HD 84937, HD 140283, or LP 815-43, with $$[\mathrm{Fe}/\mathrm{H}]=\log_{10}\!\big((N_{\rm Fe}/N_{\rm H})_{\rm star}\big)-\log_{10}\!\big((N_{\rm Fe}/N_{\rm H})_{\odot}\big)$$22 upper limits of $$[\mathrm{Fe}/\mathrm{H}]=\log_{10}\!\big((N_{\rm Fe}/N_{\rm H})_{\rm star}\big)-\log_{10}\!\big((N_{\rm Fe}/N_{\rm H})_{\odot}\big)$$23, $$[\mathrm{Fe}/\mathrm{H}]=\log_{10}\!\big((N_{\rm Fe}/N_{\rm H})_{\rm star}\big)-\log_{10}\!\big((N_{\rm Fe}/N_{\rm H})_{\odot}\big)$$24, and $$[\mathrm{Fe}/\mathrm{H}]=\log_{10}\!\big((N_{\rm Fe}/N_{\rm H})_{\rm star}\big)-\log_{10}\!\big((N_{\rm Fe}/N_{\rm H})_{\odot}\big)$$25 for $$[\mathrm{Fe}/\mathrm{H}]=\log_{10}\!\big((N_{\rm Fe}/N_{\rm H})_{\rm star}\big)-\log_{10}\!\big((N_{\rm Fe}/N_{\rm H})_{\odot}\big)$$26, respectively (Wang et al., 2021). This removes the so-called second cosmological lithium problem and supports the conclusion that earlier percent-level claims were artifacts of inadequately modeled convective asymmetries.

The present status is therefore asymmetric. The existence of the Spite Plateau itself is secure; its extension across Galactic substructures and external systems is also well supported (Simpson et al., 2020, Matteucci et al., 2021). What remains unresolved is the exact physical pathway from primordial lithium to the observed surface abundances. The strongest recurring interpretation in the cited literature is that the plateau is a universal depleted value, shaped primarily by stellar transport and destruction processes rather than by environment-specific initial conditions or by a failure of standard cosmology. Alternative views remain in circulation, especially at the most metal-poor end and in line-formation physics, but they operate against an observational background in which the plateau is real, persistent, and cosmologically anomalous.