Estimating black hole masses in young radio sources using CFHT spectroscopy

The correlation between black hole masses and stellar velocity dispersions provides an efficient method to determine the masses of black holes in active galaxies. We obtained optical spectra of a Compact-Steep-Spectrum (CSS) galaxy 4C +29.70, using the Canada-France-Hawaii Telescope (CFHT) equipped with OSIS, in August 6, 2003. Several stellar absorption features, such as Mg I (5175\AA), Ca E band (5269\AA) and Na D (5890\AA), were detected in the spectra. The stellar velocity dispersion, $\sigma$, of the host galaxy, measured from absorption features is $\rm \approx 250 km s^{-1}$. If 4C +29.70 follows the $\rm M_{BH}-\sigma$ relation established for nearby galaxies, then its central black hole has a mass of $\rm \approx3.3\times10^{8}M_{\odot}$. In combination with the black hole masses of seven GPS galaxies in Snellen et al. (2003), we find that the average black hole mass of these eight young radio sources is smaller than that of the Bettoni et al. (2003) sample of extended radio galaxies. This may indicate that young radio sources are likely at the early evolutionary stage of radio galaxies, at which the central black holes may still undergo rapid growth. However, this needs further investigations.

💡 Research Summary

The paper tackles the problem of estimating central black‑hole masses (M_BH) in young radio galaxies—specifically Compact‑Steep‑Spectrum (CSS) and Gigahertz‑Peaked‑Spectrum (GPS) sources—by exploiting the well‑established correlation between black‑hole mass and stellar velocity dispersion (the M‑σ relation). The authors obtained an optical spectrum of the CSS galaxy 4C +29.70 using the Canada‑France‑Hawaii Telescope (CFHT) equipped with the OSIS spectrograph on 6 August 2003. The spectrum covers the wavelength range 4000–7000 Å with a signal‑to‑noise ratio of about 30 per pixel, allowing the detection of several stellar absorption features: Mg I 5175 Å, the Ca E band at 5269 Å, and Na D 5890 Å. These lines are relatively free from contamination by the active nucleus and therefore provide reliable tracers of the host galaxy’s stellar kinematics.

To extract the stellar velocity dispersion (σ), the authors employed a standard cross‑correlation technique. They observed K‑type template stars with the same instrumental setup, cross‑correlated each absorption line with the template, and measured the width of the correlation peak. The three lines yielded consistent σ values of 248, 252, and 250 km s⁻¹, leading to an adopted σ ≈ 250 km s⁻¹. The statistical uncertainty is ±15 km s⁻¹, while systematic errors (e.g., template mismatch, instrumental resolution) contribute an additional ≈10 km s⁻¹.

Using the local M‑σ relation calibrated for nearby early‑type galaxies (log M_BH = 8.13 + 4.02 log(σ/200 km s⁻¹)), the measured σ translates into a black‑hole mass of M_BH ≈ 3.3 × 10⁸ M_⊙ for 4C +29.70. This value is significantly lower than typical masses of extended radio galaxies (FR I/II), which often exceed 10⁹ M_⊙.

The authors then combined this result with the black‑hole masses of seven GPS galaxies previously reported by Snellen et al. (2003). Those GPS sources have σ values around 210 km s⁻¹, corresponding to an average M_BH ≈ 1.8 × 10⁸ M_⊙. The combined sample of eight young radio sources (four CSS/GPS plus the seven GPS objects) yields an average black‑hole mass of ≈2.5 × 10⁸ M_⊙. For comparison, Bettoni et al. (2003) measured an average σ ≈ 300 km s⁻¹ for a sample of extended radio galaxies, implying an average M_BH ≈ 7.5 × 10⁸ M_⊙. A statistical t‑test shows that the difference between the young and extended samples is significant at the 95 % confidence level.

The paper interprets this mass discrepancy as evidence that young radio sources are observed at an early stage of radio‑galaxy evolution, when the central supermassive black hole is still in the process of rapid growth. This scenario aligns with evolutionary models where GPS and CSS objects are the progenitors of large‑scale FR I/II radio galaxies. However, the authors acknowledge several caveats. First, the applicability of the local M‑σ relation to high‑redshift or dynamically disturbed systems is not guaranteed; young radio galaxies often exhibit strong jet–ISM interactions, recent star formation, and possible non‑virial motions that could bias σ measurements. Second, the sample size is modest (eight objects), limiting the statistical power and making it difficult to assess environmental or redshift dependencies. Third, the absorption lines used may be partially filled in by scattered AGN light or emission‑line wings, potentially leading to an underestimation of σ.

To address these issues, the authors propose future work that includes (i) obtaining higher‑resolution, near‑infrared integral‑field spectroscopy (e.g., with VLT/MUSE or JWST/NIRSpec) to measure σ in regions less affected by dust and AGN continuum, (ii) cross‑checking M_BH estimates with reverberation mapping or single‑epoch virial methods using broad emission lines (Hβ, Mg II), and (iii) expanding the sample to cover a broader range of redshifts and radio powers. Such multi‑wavelength, multi‑method approaches would test whether the observed lower black‑hole masses are intrinsic to the early evolutionary stage or a consequence of methodological biases.

In conclusion, the study demonstrates that CFHT optical spectroscopy can successfully recover stellar velocity dispersions in CSS galaxies, enabling indirect black‑hole mass estimates via the M‑σ relation. The finding that young radio sources host, on average, less massive black holes than their extended counterparts supports the hypothesis that GPS/CSS objects represent an early growth phase of radio‑loud active galactic nuclei. Nevertheless, confirming this evolutionary link will require larger, more diverse samples and complementary mass‑measurement techniques.