The abundance discrepancy between collisionally-excited and recombination lines in ionized nebulae: origin of the temperature-fluctuation parameter t^2 and its impact on the absolute metallicity scale

Chemical abundances of ionized gaseous nebulae (H II regions and planetary nebulae) derived from bright collisionally excited lines (CELs) are systematically lower, typically by a factor of about two in H II regions and by much larger factors in planetary nebulae, than abundances derived from faint optical recombination lines (RLs) of the same ions. This is the abundance discrepancy problem, quantified by the abundance discrepancy factor (ADF). A long-standing proposed explanation invokes real temperature inhomogeneities across the nebula, parameterized by the mean-square temperature fluctuation t^2 introduced by Peimbert, which bias the exponential CEL diagnostics and hence the electron-temperature-based 'direct method'. Despite major recent observational efforts — including the 2023 Nature analysis of CEL/RL ratios across a large database of deep spectra, direct far-infrared versus optical t^2 measurements in the galaxy Mrk 71, and spatially resolved integral-field mapping — the community has not converged: it remains unsettled whether temperature inhomogeneities are the (sole) cause of the ADF, what their physical origin is, and how large the resulting correction to the absolute metallicity scale of the Universe, including James Webb Space Telescope (JWST) high-redshift determinations, should be. The problem remains open and actively contested as of 2026.

Background

Ionized gaseous nebulae — H II regions (clouds photoionized by recently formed massive stars) and planetary nebulae (PNe, the ejected envelopes of dying low-mass stars) — emit an optical spectrum dominated by recombination lines (RLs) of H and He and collisionally excited lines (CELs) of metal ions. CELs are 'forbidden' transitions driven by electron collisions; their intensities depend exponentially on the electron temperature $T_e$ , whereas RL intensities are nearly temperature-independent. The 'direct method' of abundance determination therefore measures $T_e$ from the ratio of a faint auroral CEL (e.g. [O III] $`\lambda`$ 4363) to strong nebular CELs (e.g. [O III] $`\lambda`$ 5007) and converts observed CEL fluxes into ionic abundances; ionization correction factors (ICFs) then recover the abundances of unobserved ionization stages. This method underpins the metallicity scale of star-forming galaxies across cosmic time, now including detections of the auroral [O III] $`\lambda`$ 4363 line at redshifts beyond 3 with the James Webb Space Telescope (JWST). Since Peimbert's 1967 work, it is known that if the gas temperature is not uniform, the $T_e$ measured from CEL ratios is biased high and the CEL-based abundances are biased low; the effect is parameterized by $t^2$ , the mean-square fractional temperature fluctuation over the nebula. Recombination lines, being nearly temperature-independent, are almost immune to this bias. Abundances of the same ion (e.g. O $^{2+}$ ) derived from the two line families then disagree: RL abundances exceed CEL abundances by a factor of about 2 (an abundance discrepancy factor, ADF, of roughly 0.2–0.3 dex) in typical H II regions and by factors from several to over 100 in some planetary nebulae. Standard one-dimensional photoionization models of static, chemically homogeneous, constant-density H II regions predict $t^2$ between 0.000 and about 0.015, whereas several observational diagnostics of the best-studied objects indicate $t^2$ around 0.02–0.1; Peimbert, Peimbert and Delgado-Inglada (2017) call reconciling this difference an open problem with at least ten proposed explanations in the literature. Alternative resolutions of the ADF that do not require large $t^2$ include chemical inhomogeneities (metal-rich cold clumps, invoked especially for the extreme ADFs of PNe, possibly tied to close-binary central stars), deviations of the electron velocity distribution from Maxwellian ( $\kappa$ distributions, now largely disfavored), and unaccounted radiative-transfer effects such as deviations from 'Case B' recombination (the standard assumption that certain H and He recombination photons are reabsorbed on the spot). The Pérez-Montero (2017) tutorial of the direct method lists temperature fluctuations, ionization-structure and density effects, neutral zones, photon escape, and calibration discrepancies among the unresolved limitations of the method; the coherent and central thread among them, recognized by the field as a named open problem, is the ADF together with the reality, magnitude, and physical origin of $t^2$ . Its resolution matters because if $t^2$ is large, CEL-based metallicities — the basis of the direct method and of most strong-line calibrations — are systematically underestimated, with the largest corrections at low metallicity, precisely the regime probed for high-redshift galaxies; if instead the ADF has another origin, the direct-method scale stands but the RL diagnostics require a different interpretation.

Problem Statement

For photoionized gaseous nebulae, determine the origin of the abundance discrepancy between collisionally excited lines and recombination lines of heavy ions, and in particular settle whether real spatial temperature inhomogeneities across the ionized gas — the Peimbert temperature-fluctuation parameter $t^2$ , exceeding the values ( $t^2 \lesssim 0.015$ ) predicted by standard one-dimensional photoionization models of static, chemically homogeneous H II regions — are the cause of the factor-of-about-two (and larger) excess of recombination-line over CEL abundances. A complete resolution must (i) identify the physical mechanism or combination of mechanisms — temperature inhomogeneities, density variations, chemical inhomogeneities, radiative-transfer effects such as Case B deviations — that quantitatively accounts for the observed abundance discrepancy factors and for the different temperature diagnostics (CEL ratios, Balmer-continuum temperatures — derived from the Balmer jump, the discontinuity in the hydrogen recombination continuum near 3646 Angstrom — He I recombination temperatures, far-infrared fine-structure lines) within individual objects such as Orion, Mrk 71, and the Lagoon Nebula; and (ii) establish the resulting corrected absolute gas-phase metallicity scale, stating how much CEL-based direct-method and strong-line abundances, including those now being measured for high-redshift galaxies with JWST, are biased at a given metallicity. Conversely, a demonstrated absence of the required temperature inhomogeneities, together with a quantitatively established alternative mechanism for the CEL/RL discrepancy, would resolve the problem in the other direction.

References

E. Pérez-Montero 2017, 'Ionized Gaseous Nebulae Abundance Determination from the Direct Method', Publications of the Astronomical Society of the Pacific 129, 043001, doi:10.1088/1538-3873/aa5abb, https://doi.org/10.1088/1538-3873/aa5abb (source tutorial; preprint arXiv:1702.04255)
M. Peimbert, A. Peimbert, G. Delgado-Inglada 2017, 'Nebular Spectroscopy: A Guide on H ii Regions and Planetary Nebulae', Publications of the Astronomical Society of the Pacific 129, 082001, doi:10.1088/1538-3873/aa72c3, https://doi.org/10.1088/1538-3873/aa72c3
J. E. Méndez-Delgado, C. Esteban, J. García-Rojas, K. Kreckel, M. Peimbert 2023, 'Temperature inhomogeneities cause the abundance discrepancy in H ii regions', Nature 618, 249-251, doi:10.1038/s41586-023-05956-2, https://doi.org/10.1038/s41586-023-05956-2 (preprint arXiv:2305.11578)
Y. Chen, T. Jones, R. Sanders, D. Fadda, J. Sutter, R. Minchin, E. Huntzinger, P. Senchyna, D. Stark, J. Spilker, B. Weiner, G. Roberts-Borsani 2023, 'Accurate oxygen abundance of interstellar gas in Mrk 71 from optical and infrared spectra', Nature Astronomy 7, 771-778, doi:10.1038/s41550-023-01953-7, https://doi.org/10.1038/s41550-023-01953-7 (preprint arXiv:2304.09898)
J. E. Méndez-Delgado, C. Esteban, J. García-Rojas, K. Kreckel, M. Peimbert 2024, 'Effects of density and temperature variations on the metallicity of Mrk 71', Nature Astronomy 8, 275-277, doi:10.1038/s41550-024-02198-8, https://doi.org/10.1038/s41550-024-02198-8 (Matters Arising; preprint arXiv:2310.01197, whose arXiv version bears the different title 'Temperature inhomogeneities in Mrk71 can not be discarded')
J. E. Méndez-Delgado, E. D. Skillman, E. Aver, C. Morisset, C. Esteban, J. García-Rojas, K. Kreckel, N. S. J. Rogers, F. F. Rosales-Ortega, K. Z. Arellano-Córdova, S. R. Flury, E. Reyes-Rodríguez, M. Orte-García, S. Tan 2025, 'Generalized Te([O iii])–Te(He i) Discrepancies in Ionized Nebulae: Possible Evidence of Case B Deviations and Temperature Inhomogeneities', The Astrophysical Journal 986, 74, doi:10.3847/1538-4357/adc67a, https://doi.org/10.3847/1538-4357/adc67a
K. Z. Arellano-Córdova, M. Rodríguez 2020, 'The Te[N ii]–Te[O iii] temperature relation in H ii regions and the reliability of strong-line methods', Monthly Notices of the Royal Astronomical Society 497, 672-686, doi:10.1093/mnras/staa1759, https://doi.org/10.1093/mnras/staa1759
L. J. Kewley, S. L. Ellison 2008, 'Metallicity Calibrations and the Mass-Metallicity Relation for Star-forming Galaxies', The Astrophysical Journal 681, 1183-1204, doi:10.1086/587500, https://doi.org/10.1086/587500
J. García-Rojas 2020, 'Physical Conditions and Chemical Abundances in Photoionized Nebulae from Optical Spectra', Reviews in Frontiers of Modern Astrophysics, 89-121, doi:10.1007/978-3-030-38509-5_4, https://doi.org/10.1007/978-3-030-38509-5_4
R. A. Riffel, O. L. Dors, A. C. Krabbe, C. Esteban 2021, 'Electron temperature fluctuations in Seyfert galaxies', Monthly Notices of the Royal Astronomical Society: Letters 506, L11-L15, doi:10.1093/mnrasl/slab064, https://doi.org/10.1093/mnrasl/slab064
A. Singh, G. A. Blanc, N. Kumari, J. E. Méndez-Delgado, S. F. Sánchez, C. Morisset, E. Congiu, K. Kreckel, A. Roman-Lopes, O. Egorov, N. Drory, R. Sankrit, A. Mejía-Narváez, E. Egorova, A. M. Jones, D. Bizyaev, N. Sattler, E. J. Johnston, D. Minniti, R. de J. Zermeño, J. G. Fernández-Trincado, J. A. Kollmeier 2026, 'SDSS-V LVM: A Spatially Resolved Study of the Physical Conditions and the Chemical Abundance Discrepancy in the Lagoon Nebula (M 8)', The Astrophysical Journal 1001, 238, doi:10.3847/1538-4357/ae563a, https://doi.org/10.3847/1538-4357/ae563a
M. Curti 2025, 'The Chemical Evolution of Galaxies', arXiv:2504.08933, https://arxiv.org/abs/2504.08933 (single-author review chapter for the Encyclopedia of Astrophysics, ed. I. Mandel, to be published by Elsevier as a Reference Module)