3D-Herschel: Constraining Dust Emission with Panchromatic Modeling of 3D-HST Galaxies

We present 3D-Herschel, a new 0.3-350$μ$m photometric catalog that combines deblended Herschel far-infrared (FIR) imaging with the CANDELS/3D-HST legacy fields to probe the dust-obscured universe. Using the 17-parameter fitting code Prospector-$β$, a Bayesian inference framework, we model 41,387 galaxies spanning 0.5 $< z <$ 2.5 to measure stellar and dust properties with realistic error bars. Comparing fits with and without FIR constraints, we find that for the 3.2$%$ of galaxies with $>3 σ$ detections in at least two Herschel bands, UV-MIR-only models (0.3-24$μ$m) recover robust stellar ages, SFRs, and stellar masses (50-70$%$ within the median 1$σ$ error). Consequently, the star-forming sequence shows no systematic offset, with an average deviation of 0.1$\pm$0.07 dex at fixed stellar mass for FIR-detected sources at all redshifts. However, the use of rigid log-average IR templates with fixed dust emission parameters ($γ$, $U_{\mathrm{min}}$, $Q_{\mathrm{PAH}}$) in UV-MIR modeling corresponds to an unevolving MIR-to-IR luminosity ratio and cold dust temperatures. By contrast, fits that include Herschel photometry, with added freedom in the FIR, yield dust temperatures that are $\sim$7K warmer at all redshifts, with $\sim$0.2 dex higher IR-to-7.7$μ$m luminosity ratios at the low-mass end of a Herschel-detected sample (log($M_{\star}$) $\sim$9.6 $M_{\odot}$). These results demonstrate that MIR-to-IR conversions depend on stellar mass, cautioning against $L_{\mathrm{IR}}$-independent templates without FIR data. For galaxies with $<10^{11} \ M_{\odot}$ at $z>1.5$, even with state-of-the-art analysis, Herschel can at best provide upper limits due to source confusion; next-generation FIR telescopes will be essential to fully characterize dust emission in distant galaxies.

💡 Research Summary

The paper introduces “3D‑Herschel,” a new photometric catalog that spans the full 0.3–350 µm wavelength range by combining deblended Herschel far‑infrared (FIR) imaging with the existing CANDELS/3D‑HST legacy data. Using forced‑photometry techniques (MOPHONGO) that employ high‑resolution HST priors, the authors extract reliable FIR fluxes from the low‑resolution PACS and SPIRE maps, increasing the number of FIR‑detected galaxies by roughly 50 % compared with traditional cross‑matching. The final sample comprises 41 387 galaxies in the redshift interval 0.5 < z < 2.5, each with up to ~50 photometric bands covering UV, optical, near‑IR, mid‑IR (Spitzer/IRAC, MIPS 24 µm) and FIR (Herschel 70–350 µm).

For the physical‑parameter inference the authors adopt Prospector‑β, a state‑of‑the‑art Bayesian SED‑fitting framework with 17 free parameters. The model includes a non‑parametric star‑formation history, a two‑component dust attenuation model (birth‑cloud + diffuse ISM), and a flexible dust‑emission component based on the Draine & Li (2007) prescription, allowing the PAH fraction (Q_PAH), minimum radiation field (U_min) and the fraction of dust heated by intense radiation (γ) to vary. Energy balance is enforced so that the energy absorbed in the UV–optical is re‑emitted in the IR.

Two fitting configurations are compared: (1) “UV‑MIR‑only” fits that use data from 0.3 to 24 µm (i.e., no FIR constraints) and (2) “FIR‑included” fits that also incorporate the Herschel photometry. Only 3.2 % of the sample (≈ 1 300 galaxies) have >3σ detections in at least two Herschel bands, but this subset is sufficient to test the impact of FIR data on derived galaxy properties.

Key findings:

1. Stellar masses, SFRs, and ages are robust without FIR – For the FIR‑detected subsample, the UV‑MIR‑only fits recover stellar mass, star‑formation rate, and mass‑weighted age within the median 1σ uncertainties for 50–70 % of objects. Consequently, the star‑forming main sequence (SFMS) shows virtually no systematic offset: the average deviation between FIR‑included and UV‑MIR‑only SFRs at fixed mass is only 0.1 ± 0.07 dex across all redshifts. This demonstrates that, for galaxies bright enough in the FIR, the UV‑to‑MIR SED already contains enough information to constrain the bulk of the stellar population.

2. Fixed IR templates introduce biases – When FIR data are omitted, the authors adopt a “log‑average” IR template with fixed dust‑emission parameters (γ, U_min, Q_PAH). This assumption forces a constant MIR‑to‑total‑IR luminosity ratio and a static dust temperature across the galaxy population. By contrast, the FIR‑included fits reveal that the average dust temperature is ≈ 7 K higher than the template value at all redshifts, and that low‑mass galaxies (log M★ ≈ 9.6 M⊙) have IR‑to‑7.7 µm (PAH) luminosity ratios higher by ~0.2 dex. Thus, MIR‑to‑IR conversion factors depend on stellar mass (and implicitly on redshift), invalidating the use of universal, FIR‑independent IR templates.

3. Limits of Herschel at high‑z, low mass – For galaxies with M★ < 10¹¹ M⊙ at z > 1.5, source confusion in Herschel maps prevents reliable flux extraction; the catalog can only provide upper limits. The authors argue that next‑generation FIR facilities (e.g., SPICA, Origins Space Telescope) with higher angular resolution and sensitivity will be required to break this barrier and fully characterize dust emission in typical high‑z, low‑mass systems.

Methodologically, the paper validates the deblended Herschel photometry against existing PEP and GOODS‑H catalogs, showing consistent fluxes and reduced scatter. The Bayesian posterior distributions illustrate reduced degeneracies between dust attenuation, stellar age, and metallicity when FIR data are included, confirming the long‑standing expectation that FIR constraints help break the “dust–age” degeneracy.

In the discussion, the authors place their results in the context of previous work that relied on UV+IR SFR estimators (e.g., Daddi et al. 2007; Kennicutt & Evans 2012). They note that while UV+IR methods work well for strongly star‑forming systems, they can overestimate SFRs in more quiescent galaxies by up to an order of magnitude. Prospector‑β’s energy‑balance approach, especially with flexible dust emission, yields SFRs that are systematically lower (by 0.1–1 dex) for such galaxies, aligning with earlier findings (Leja et al. 2019b).

The paper concludes that the combination of high‑dimensional Bayesian SED fitting and robust FIR deblending provides a powerful framework for future galaxy‑evolution studies. It emphasizes that while UV‑MIR data alone can recover bulk stellar properties for FIR‑bright galaxies, accurate dust‑temperature measurements, PAH strength variations, and reliable total IR luminosities require direct FIR observations. The 3D‑Herschel catalog thus represents a valuable resource for the community, and its methodology sets a benchmark for upcoming surveys that will exploit JWST, ALMA, and future FIR missions to push these analyses to fainter, higher‑redshift regimes.