Uncertainties in high-$z$ galaxy properties inferred from SED fitting using JWST NIRCam photometry

TL;DR

This paper investigates how uncertainties in SED fitting affect derived properties of high-$z$ galaxies using JWST/NIRCam photometry. By applying Bagpipes to mock Sphinx$^{20}$ galaxies at $z=6$ with variations in star formation histories, metallicity treatment, dust attenuation, and nebular emission, the authors quantify biases in stellar mass $M_\star$, star formation rate $SFR_{10}$, and stellar metallicity $Z_\star$, and assess implications for the stellar mass function and the star formation main sequence. They find substantial biases: even without dust or emission lines, $M_\star$ can be overestimated by ~60%, and $SFR_{10}$ underestimated by ~2×; including dust and nebular emission worsens biases, though adding a line-free medium band like F410M and using optimized fitting statistics (e.g., minimum $\chi^2$) can mitigate some biases. The results highlight the importance of flexible SFHs, careful treatment of dust attenuation, and inclusion of line-free photometric bands to obtain reliable high-$z$ galaxy properties, with direct consequences for interpreting galaxy assembly and evolution in the early Universe.

Abstract

Numerous high-$z$ galaxies have recently been observed with the James Webb Space Telescope (JWST), providing new insights into early galaxy evolution. Their physical properties are typically derived through spectral energy distribution (SED) fitting, but the reliability of this approach for such early systems remains uncertain. Applying {\sc Bagpipes} on simulated SEDs at $z=6$ from the {\sc Sphinx$^{20}$} cosmological simulation, we examine uncertainties in the recovery of stellar masses, star formation rates (SFR$_{10}$), and stellar metallicities from mock JWST/Near-Infrared Camera photometry. Even without dust or emission lines, fitting the intrinsic stellar continuum overestimates the stellar mass by about 60\% on average (and by up to a factor of five for low-mass galaxies with recent starbursts) and underestimates SFR$_{10}$ by a factor of two, owing to inaccurate star formation histories and age-metallicity degeneracies. The addition of dust and nebular emission further amplifies these biases, yielding offsets of approximately +0.3 and -0.4 dex in stellar mass and SFR$_{10}$, respectively, while leaving stellar metallicities largely unconstrained. Incorporating bands free of strong emission lines, such as F410M, helps mitigate stellar mass overestimation by disentangling line emission from older stellar populations. We also find that best-fit or likelihood-weighted estimates are generally more accurate than median posterior values. Although stellar mass functions are reproduced reasonably well, the slope of the star formation main sequence depends sensitively on the adopted fitting model. Overall, these results underscore the importance of careful modelling when interpreting high-$z$ photometry, particularly for galaxies with recent star formation burst and/or strong emission lines, to minimise systematic biases in derived physical properties.

Figures (21)

• Figure 1: Example SED of a Sphinx$^{20}$ galaxy at $z=6$. The black line represents the attenuated spectrum along a random line of sight. Coloured points mark the pivot wavelengths and bandwidths of the six JWST NIRCam filters, with corresponding throughput curves shown below. The orange and yellow lines indicate the intrinsic stellar and nebular continua, respectively. The shaded regions indicate the wavelength range used to measure the UV slope ($\beta$) and the position of the Balmer break, respectively.
• Figure 2: Comparison of the true and derived stellar masses with (left; intr_Zfix) and without (right; intr_Z) fixing the stellar metallicity. Different SFH models are shown in different colours, as indicated in the legend. The right panel also displays the fitted $Z_\star$ values for each SFH model, while the lower panels show the differences relative to the true quantities. When metallicity is treated as a free parameter, $M_\star$ is overestimated by 0.2 dex for both parametric and non-parametric SFH models. This bias reflects the age--metallicity degeneracy, which tends to yield posteriors with underestimated metallicity and overly high fractions of old stars relative to the true values.
• Figure 3: Posterior SFHs corresponding to the median SFR$_\mathrm{100}$ values are displayed together with the true SFH (solid black line) for two representative galaxies: one low-mass ($\log\,M_\star/M_\odot=7.32$, left panels) and one high-mass ($\log\,M_\star/M_\odot=9.67$, right panels). Each row corresponds to a different SFH model used in the fitting, and each coloured line denotes a separate fitting configuration (blue: intr_Z, orange: att_Z, red: full_Z), with the SMC attenuation law applied when a dust model is included. Dashed vertical lines indicate the mass-weighted age. Under intr_Z, the constant SFH model generates the fewest young stars (< 10 Myr) among all models and yields the lowest mass-weighted age. Consequently, it most strongly underestimates SFR$_\mathrm{10}$ and overestimates SFR$_\mathrm{100}$, as most of the stellar mass is concentrated in intermediate-age populations. When dust and nebular components are added, all three SFH models predict larger old stellar populations, increasing the mass-weighted age and leading to an underestimation of SFR$_\mathrm{100}$, particularly in high-mass galaxies.
• Figure 4: Difference between the true and recovered SFRs from the intrinsic SEDs as a function of true $M_\star$ (intr_Z). The upper and lower panels present SFRs averaged over 10 Myr (SFR$_\mathrm{10}$) and 100 Myr (SFR$_\mathrm{100}$), respectively. In the fitting, stellar metallicity is treated as a free parameter. The constant SFH model substantially underestimates SFR$_\mathrm{10}$ but overestimates SFR$_\mathrm{100}$, whereas the double power-law and flexible models recover SFRs to within $\sim$0.5 dex.
• Figure 5: Fitted $M_\star$ and its offset from the true value for fits to the attenuated stellar continuum, with metallicity treated as a free parameter and three dust models applied: SMC (left), Calzetti (middle), and Salim (right). Considerable offsets in $M_\star$ persist even when the normalised true SFH is used, driven mainly by slope mismatches between the true normalised attenuation curve and the assumed model curve. Parametric and non-parametric SFH models yield comparable results. Across all four SFH models, the SMC law yields the closest agreement in $M_\star$, whereas the Calzetti law produces the largest offset.