Catching the 2021 γ-ray flare in the blazar TXS 2013+370

TL;DR

TXS 2013+370 is examined during a major GeV flare to locate the γ-ray production site within its inner jet. By combining full-polarimetric mm-VLBI imaging at 22/43/86 GHz with concurrent radio and γ-ray monitoring, the study identifies a new near-core knot, measures an exceptionally high RM, and finds a γ-ray lead of $\Delta t=(102\pm12)$ days over 15 GHz, constraining the γ-ray region to $Δr=(2.71±0.47)$ pc and $r_{\gamma}= -0.66±1.08$ pc (within ~0.42 pc of the jet apex at 1σ). The results support two plausible external-Compton scenarios—seed photons from the BLR ($\sim0.07$ pc) or from the dusty torus (up to ~0.42 pc)—and favor a sub-parsec/parsec-scale γ-ray production region anchored to jet features rather than a simple one-zone model. This work demonstrates the power of mm-VLBI polarimetry to localize high-energy emission in blazars and to test EC models in the inner jet environment, with implications for jet physics and particle acceleration.

Abstract

The $γ$-ray-loud blazar TXS 2013+370, a powerful multiwavelength emitter at $z = 0.859$, underwent an exceptional GeV outburst in late 2020-early 2021. In this work, we present full-polarization VLBI imaging at 22, 43, and 86 GHz together with contemporaneous single-dish monitoring (radio and $γ$-rays) to localize the high-energy dissipation site and probe the inner-jet structure and magnetic field topology. The images revealed a compact near-core knot at $r \simeq 40$-$60~μ\mathrm{as}$ contemporaneously with the GeV flare and a flat, core-dominated spectrum ($α\gtrsim -0.5$). The core has strong linear polarization and exhibits a $\sim 50^{\circ}$ EVPA rotation at 86 GHz; pixel-based and integrated fits yield a high, uniform rotation measure, ${\rm RM} = (7.8 \pm 0.2) \times 10^{4}~{\rm rad~m^{-2}}$, consistent with an external Faraday screen. Cross-correlation of Fermi-LAT and 15 GHz data shows a significant peak with the $γ$ rays leading by $Δt = (102 \pm 12)\,\mathrm{d}$; adopting $β_{\rm app} = 4.2 \pm 0.5$ and $θ= 4.1^{\circ} \pm 0.2^{\circ}$ implies a de-projected separation $Δr_{γ-15} = (2.71 \pm 0.47)\,\mathrm{pc}$ and locates the GeV emission between the jet apex and $\sim 0.42$ pc ($1σ$ range) downstream. Our results do not uniquely pinpoint the emission site; rather, they support two valid scenarios. The $γ$-ray production may occur within the BLR ($\sim 0.07$pc), where external-Compton scattering of optical/UV photons produces the $γ$ rays, or beyond the BLR, reaching $\sim 0.42$ pc ($1σ$) within the inner parsecs, where external-Compton scattering of dusty-torus infrared photons dominates. Both scenarios are compatible within the allowed range of emission distances, while opacity-driven core shifts modulate the observed radio-$γ$ delay without requiring large relocations of the dissipation zone.

Figures (7)

• Figure 1: Total intensity images of TXS 2013+370 from 11 February 2021. Left: 22 GHz; top-right: 43 GHz; bottom-right: 86 GHz. Contours are at 0.5, 1, 2, 4, 8, 16, 32, and 64% of each panel’s peak (22: 1.33 Jy/beam; 43: 0.75 Jy/beam; 86: 0.71 Jy/beam). Restoring beams are shown as gray ellipses (parameters in Table \ref{['table:obs_par']}). Orange circles mark MODELFIT Gaussian centroids. The 43 GHz map resolves a new knot (N2) near the core; at 86 GHz the compact core and components A2, C3, and N2 are clearly detected.
• Figure 2: Light curves of the blazar TXS 2013+370 at different frequencies. From top to bottom: Fermi-LAT 0.1-100 GeV with 1-month binning, flux plotted vs. time for 226 GHz SMA, and 15 GHz OVRO. The vertical dashed lines indicate the estimated ejection times of the components N, N1 refId0, and the new knot N2, respectively, whereas the width of the shadow areas designates the uncertainty of these estimations. For N2, the uncertainty is based on the uncertainty of component A1 as it was found in refId0. The gray dashed line indicates a new flaring activity in the source.
• Figure 3: Spectral index distributions of TXS 2013+370. The contour levels are set to 0.5, 1, 2, 4, 8, 16, 32, and 64% Jy/beam of the peak flux density of the highest frequency map in the pair (see Table: \ref{['table:obs_par']}) and represent the total intensity contours. All the images are convolved with a common beam that was set equal to the equivalent circular beam $b=(b_{max}b_{min})^{1/2}$ of the highest frequency, and each frequency pair was aligned using a 2D cross-correlation analysis. Left: 22-43 GHz frequency pair. Right: 43-86 GHz frequency pair.
• Figure 4: Polarization images of TXS 2013+370 observed on 11 February 2021. Left: 22 GHz polarized-intensity map with overplotted EVPA vectors (white sticks). Middle: 43 GHz polarization image. Right: 86 GHz polarization image. Contours are at 0.5, 1, 2, 4, 8, 16, 32, and 64% of each panel’s peak (22 GHz: 1.31 Jy /beam; 43 GHz: 0.80 Jy /beam; 86 GHz: 0.70 Jy /beam) and represent the total intensity contours. The restoring beam is shown as a gray ellipse in the lower-left corner.
• Figure 5: Faraday rotation results for the blazar TXS 2013+370. Left: Dashed lines correspond to the outermost total intensity contours of the 22 GHz (black) and 43 GHz (red) maps. Solid contours are set to 2, 4, 8, 16, 32, and 64% of the 86 GHz peak polarized intensity (0.70 Jy /beam). All maps are convolved with a common beam, the equivalent circular beam of the 86 GHz map, shown as a gray circle in the bottom-left corner. Middle: Linear fit of Eq. \ref{['eq:rm']} after unwrapping the $n\pi$ ambiguities, showing EVPA rotation from 86 GHz to 22 GHz; the slope yields an integrated RM of $(7.81\pm0.16)\times10^{4}$ rad m$^{-2}$ for the core region. Right: Pixel-based map of RM uncertainties from the covariance matrix of the per-pixel least-squares fits.