Stellar-to-Nebular Dust Attenuation Ratio

• Stellar-to-nebular dust attenuation ratio quantifies the differential extinction affecting the stellar continuum versus nebular emission lines in galaxies.
• It is derived by comparing SED-based stellar color excesses with Balmer decrement measurements to accurately correct observed fluxes.
• Empirical studies using JWST and IFU data reveal that the ratio varies significantly with galaxy properties like SFR, metallicity, and dust geometry.

The stellar-to-nebular dust attenuation ratio quantifies the relative extinction suffered by the stellar continuum compared to nebular emission lines in galaxies. This ratio is fundamental for accurate correction of observed fluxes in extragalactic surveys, precise derivation of star formation rates, and for constraining dust/star geometry and ISM properties. It is typically parameterized as the ratio of color excesses, $$f \equiv E(B-V)_{\mathrm{star}} / E(B-V)_{\mathrm{neb}}$$, or equivalently as the ratio of attenuations $A_{V,\mathrm{star}}/A_{V,\mathrm{neb}}$, and is found to be a strong function of galaxy physical conditions, particularly SFR, metallicity, gas/dust geometry, and the evolutionary status.

1. Fundamental Measurement Techniques and Definitions

Determining the attenuation ratio requires independent estimates of extinction for the stellar and nebular components. Stellar attenuation is usually derived from SED fitting, which incorporates multiwavelength broadband photometry (UV–NIR), with outputs such as $E(B-V)_{\mathrm{star}}$ or $A_{V,\mathrm{star}}$. Nebular attenuation is most commonly measured from hydrogen recombination line ratios (e.g., Balmer or Paschen decrements), where the observed flux ratio (e.g., $F_{\mathrm{H}\alpha}/F_{\mathrm{H}\beta}$) is compared to the dust-free intrinsic value (e.g., 2.86 for H$\alpha$/H$\beta$, assuming Case B recombination at $T=10^4\,$K and $n_e=100\,$cm$^{-3}$), and the color excess $E(B-V)_{\mathrm{neb}}$ is inferred via: $$E(B-V)_{\mathrm{neb}} = \frac{-2.5}{k(\mathrm{H}\beta) - k(\mathrm{H}\alpha)}\,\log_{10} \left[ \frac{2.86}{F_{\mathrm{H}\alpha}/F_{\mathrm{H}\beta}} \right]$$ where $k(\lambda)$ is the adopted attenuation curve (e.g., Cardelli, Calzetti).

The ratio $f$ is then computed as: $$f = \frac{E(B-V)_{\mathrm{star}}}{E(B-V)_{\mathrm{neb}}}$$ In some works, $A_{V,\mathrm{star}}/A_{V,\mathrm{neb}}$ or $A_{V,\mathrm{star}}/A_{V,\mathrm{gas}}$ is used, with $A_{V} = R_V\,E(B-V)$.

Recent work (e.g. ALPINE-CRISTAL-JWST Survey (Tsujita et al., 21 Oct 2025)) employs spatially resolved JWST/NIRSpec spectroscopy for direct pixel-by-pixel SED and emission-line fitting, while others (e.g. (Rodríguez-Muñoz et al., 2021)) utilize statistical cross-calibration of SFR tracers corrected for dust to constrain $f$.

2. Empirical Trends and Canonical Values

Global Measurements and Typical Values

Classic studies of local starbursts (e.g., Calzetti et al.) found $f\approx 0.44$, implying that nebular lines are, on average, attenuated by more than twice the amount suffered by the integrated stellar continuum. This has been substantiated in SDSS galaxies by direct comparison of $A_{V,\mathrm{star}}$ and $A_{V,\mathrm{neb}}$ (Zahid et al., 2017, Koyama et al., 2018, Paspaliaris et al., 1 Sep 2025), with modern resolved studies showing that the ratio can depend on subgalactic location and is more accurately described as a function of local ISM properties and star-forming activity (Lin et al., 2019, Hemmati et al., 2015).

The most recent high-redshift analyses using JWST (e.g., ALPINE-CRISTAL-JWST (Tsujita et al., 21 Oct 2025)) report $f=0.51^{+0.04}_{-0.03}$ for $z\sim4.4-5.7$ main-sequence galaxies. This is slightly higher than the canonical local value, but still requires a nebular correction about twice that of the stellar continuum.

The following table summarizes typical $f$ values as a function of redshift and sample selection:

Sample/Redshift	$$f \equiv E(B-V)_{\mathrm{star}}/E(B-V)_{\mathrm{neb}}$$	Notes/Method
Local (Calzetti)	0.44	Canonical starburst value
SDSS spirals (PHANGS, (Paspaliaris et al., 1 Sep 2025))	$0.59\pm0.05$	Balmer decrement vs. SED fit (young stars)
SHARDS/CANDELS (0.3–1.5, (Rodríguez-Muñoz et al., 2021))	$0.55$ ([H$\alpha$]), $0.69$ ([OII])	SED vs. emission lines, IR+UV SFR normalization
MOSDEF ($z\sim2$, (Shivaei et al., 2020, Reddy et al., 2020))	$\sim0.5$ (low-$Z$), $\sim1$ (high-$Z$)	SED+Balmer decrements, metallicity dependent
ALPINE-CRISTAL-JWST ($z\sim4-6$, (Tsujita et al., 21 Oct 2025))	$0.51^{+0.04}_{-0.03}$	Prospector SED fits + Balmer decrement

In resolved studies (e.g. (Hemmati et al., 2015)), median $$E(B-V)_{\mathrm{neb}}/E(B-V)_{\mathrm{star}} \sim 2.5$$ for $z\sim0.4$ emission-line galaxies, but with significant (factor of ~1) intrinsic scatter.

Galaxy Property Trends

• Stellar Mass and SFR: $f$ tends to decrease with increasing stellar mass and/or SFR (Koyama et al., 2018, Lin et al., 2019, Lorenz et al., 2023); that is, higher mass and higher SFR galaxies exhibit greater differential attenuation (higher nebular vs stellar extinction).
• Specific SFR and Stellar Age: $f$ rises with sSFR and for younger mass-weighted stellar ages (Tsujita et al., 21 Oct 2025), indicating less pronounced differential attenuation in systems with more recent or ongoing star formation.
• Metallicity: The attenuation ratio depends strongly on gas-phase metallicity. In low-metallicity systems, $$E(B-V)_{\mathrm{neb}} \sim 2 \times E(B-V)_{\mathrm{star}}$$; in high-metallicity galaxies, nebular and stellar reddening are similar ($f\sim1$) (Shivaei et al., 2020).

3. Physical Origins and Dust Geometry

The differential attenuation arises from the relative spatial distributions and geometric relationship of young star-forming regions, older stellar populations, and the dust responsible for extinction. The widely-adopted two-component model (e.g. Charlot & Fall 2000) invokes:

• Dense “birth clouds”: Nebular emission, arising from the ionized gas around massive young stars, is attenuated by both the dust in these clouds and the diffuse ISM.
• Diffuse ISM: The integrated stellar continuum, dominated by longer-lived, older stars, samples the entire galaxy and is attenuated chiefly by the more pervasive, less optically thick ISM.

Spatially resolved IFU and pixel-by-pixel SED studies (Hemmati et al., 2015, Lin et al., 2019, Paspaliaris et al., 1 Sep 2025) confirm that:

• Clumpy, centrally concentrated regions of high stellar mass surface density exhibit increased nebular color excess.
• In galaxies with high sSFR or young stellar ages, the youngest massive stars have not migrated far from their natal clouds, so the difference between nebular and continuum attenuation is small; as the population ages or SFR drops, differential attenuation is enhanced.
• Inclination effects can modulate the observed $f$ due to path length changes through foreground dust (Hemmati et al., 2015, Lin et al., 2019), but are subdominant compared to the physical drivers above.

4. Attenuation Curves and their Influence

The choice of attenuation curve is critical in determining $E(B-V)_{\mathrm{star}}$ and $E(B-V)_{\mathrm{neb}}$ and thus $f$. The Milky Way, SMC, and Calzetti starburst curves each differ in slope, normalization, and UV bump structure. Notably:

• At low metallicity ($Z$), the SMC curve is steeper, and used in $z\sim2$–5 low-mass, low-metallicity galaxies (Shivaei et al., 2020, Alavi et al., 1 Oct 2025), while at high $Z$ (and at high mass), galaxies show Calzetti-like or even Milky Way-like attenuation properties (Shivaei et al., 2020, Sanders et al., 9 Aug 2024).
• In young, high-EW systems, deep JWST spectroscopy (Sanders et al., 9 Aug 2024) reveals cases where the stellar and nebular components both see the same dust screen, and thus $f$ approaches unity.
• Application of incorrect attenuation laws to emission lines (e.g., using a stellar-based curve for nebular emission) can bias corrected line fluxes and physical parameter derivations by 10–40% or more (Tsujita et al., 21 Oct 2025, Pahl et al., 14 Oct 2025, Lin et al., 7 Oct 2024).

5. Methodological Advances and Modern Constraints

Current approaches combine SED fitting with pixel-by-pixel and spatially-integrated emission-line analysis (Tsujita et al., 21 Oct 2025, Paspaliaris et al., 1 Sep 2025). JWST NIRSpec and NIRCam offer the sensitivity and spatial resolution to measure multiple HI recombination lines in individual high-$z$ galaxies, enabling direct empirical derivation of nebular attenuation curves on an object-by-object basis (Sanders et al., 9 Aug 2024, Pahl et al., 14 Oct 2025, Tsujita et al., 21 Oct 2025).

Recent studies deploy comparison of independent SFR tracers (UV+IR vs emission lines) to calibrate $f$ statistically in diverse galaxy populations (Rodríguez-Muñoz et al., 2021, Koyama et al., 2018). Systematic biases associated with assumptions about the relative $E(B-V)$ between components are now understood to impact reionization models and measurements of the ionizing photon production efficiency ($\xi_\mathrm{ion}$) by up to 0.2–0.3 dex (Pahl et al., 14 Oct 2025, Tsujita et al., 21 Oct 2025).

6. Implications, Limitations, and Future Directions

The stellar-to-nebular attenuation ratio is central to accurate physical modeling of galaxies. Adopting $f=1$ (no differential attenuation) in high-redshift SED analysis leads to systematic underestimation of line luminosities and $\xi_\mathrm{ion}$ by $3-36\%$ and $4-46\%$, respectively (Tsujita et al., 21 Oct 2025). For integrated measurements (i.e. large photometric apertures or unresolved galaxies), outshining effects from luminous young populations can bias stellar mass estimates downward by $>0.2$ dex unless pixel-based SED fitting is used (Tsujita et al., 21 Oct 2025).

The convergence in attenuation curves for nebular lines (Milky Way-like, (Sanders et al., 9 Aug 2024, Lin et al., 7 Oct 2024, Rezaee et al., 2021)) and the systematic variation of $f$ with physical galaxy parameters is now robust observationally. Ongoing JWST surveys, combined with deep ground-based IFU data and multiwavelength SED coverage, enable the community to refine models for dust geometry and radiative transfer for both nearby and high-redshift star-forming galaxies.

Key open problems include:

• Disentangling the contribution of dust-star spatial geometry versus dust grain evolution on the observed $f$ ratio.
• Developing universally applicable attenuation models for spatially unresolved high-redshift galaxies.
• Quantifying the evolution of the stellar-to-nebular attenuation ratio across cosmic time in the context of ISM phase structure, dust-to-gas ratio evolution, and starburst age.

A plausible implication is that future models and SED-fitting procedures should treat $f$ as a parameter dependent on metallicity, SFR, and galaxy evolutionary stage, rather than as a universal constant.

7. Summary Table of Empirical $f$ Measurements

Survey/Sample	Redshift	Typical $f$	Correlations with Properties	Remarks
Local starbursts (Calzetti)	0	0.44	Weak	Canonical; much used
SDSS/PHANGS spirals (Paspaliaris et al., 1 Sep 2025)	0	0.59$\pm$0.05	$f$ increases for young, declines for old	SED-based, spatially resolved
SHARDS/CANDELS (Rodríguez-Muñoz et al., 2021)	0.3-1.5	0.55-0.69	$f$ increases with A(UV), weak with $M_*$	SFR UV+IR calibration
MOSDEF (low-$Z$) (Shivaei et al., 2020)	1.4-2.6	$\sim$0.5	$f$ up with $Z$	High/low-$Z$ dichotomy
MOSDEF (high-$Z$) (Shivaei et al., 2020)	1.4-2.6	$\sim$1	Strong metallicity dependence	Calzetti-like curve
ALPINE-CRISTAL-JWST (Tsujita et al., 21 Oct 2025)	4.4-5.7	0.51$^{+0.04}_{-0.03}$	$f$ with sSFR, age	JWST/NIRSpec+NIRCam
JWST/AURORA (Pahl et al., 14 Oct 2025)	1.5-6.9	Variable, always $<1$	Increases with sSFR, lower in dusty galaxies	Individual nebular curves

These measurements reinforce the conclusion that the stellar-to-nebular dust attenuation ratio is not universal but varies systematically with physical, chemical, and geometric properties of galaxies and should be incorporated in all extragalactic analyses requiring accurate emission-line and SED corrections.