On 3rd September 2015, the Chromospheric Lyman-Alpha SpectroPolarimeter (CLASP) successfully measured the linear polarization produced by scattering processes in the hydrogen Lyman-$\alpha$ line of the solar disk radiation, revealing conspicuous spatial variations in the $Q/I$ and $U/I$ signals. Via the Hanle effect the line-center $Q/I$ and $U/I$ amplitudes encode information on the magnetic field of the chromosphere-corona transition region (TR), but they are also sensitive to the three-dimensional structure of this corrugated interface region. With the help of a simple line formation model, here we propose a statistical inference method for interpreting the Lyman-$\alpha$ line-center polarization observed by CLASP.

We derive a formalism to describe the scattering of polarized radiation over the full spectral range encompassed by atomic transitions belonging to the same spectral series (e.g., the H I Lyman and Balmer series, the UV multiplets of Fe I and Fe II). This allows us to study the role of radiation-induced coherence among the upper terms of the spectral series, and its contribution to Rayleigh scattering and the polarization of the solar continuum. We rely on previous theoretical results for the emissivity of a three-term atom of the $\Lambda$-type taking into account partially coherent scattering, and generalize its expression in order to describe a "multiple $\Lambda$" atomic system underlying the formation of a spectral series. Our study shows that important polarization effects must be expected because of the combined action of partial frequency redistribution and radiation-induced coherence among the terms of the series. In particular, our model predicts the correct asymptotic limit of 100% polarization in the far wings of a \emph{complete} (i.e., $\Delta L=0,\pm 1$) group of transitions, which must be expected on the basis of the principle of spectroscopic stability.

We consider the vector emissivity of the polarized radiation transfer in a $\Lambda$-type atomic transition, which we recently proposed in order to account for both CRD and PRD contributions to the scattered radiation. This expression can concisely be written as $$\varepsilon=\left(\varepsilon^{(1)} -\varepsilon^{(2)}_{\rm f.s.} \right) +\varepsilon^{(2)}\;$$ where $\varepsilon^{(1)}$ and $\varepsilon^{(2)}$ are the emissivity terms describing, respectively, one-photon and two-photon processes in a $\Lambda$-type atom, %the fully and partially redistributed radiations, and where "f.s." means that the corresponding term must be evaluated assuming an appropriate "flat spectrum" average of the incident radiation across the spectral line. In this follow up study, we explicitly consider the expressions of these various terms for the case of a polarized multi-term atom, in order to derive the algebraic forms of the branching ratios between the CRD and PRD contributions to the emissivity. In the limit of a two-term atom with non-coherent lower-term, our results are shown to be in full agreement with those recently derived by Bommier (2017)

We discuss the implementation of physically meaningful branching ratios between the CRD and PRD contributions to the emissivity of a polarized multi-term atom in the presence of both inelastic and elastic collisions. Our derivation is based on a recent theoretical formulation of partially coherent scattering, and it relies on a heuristic diagrammatic analysis of the various radiative and collisional processes to determine the proper form of the branching ratios. The expression we obtain for the emissivity is $\varepsilon=\left[\varepsilon^{\tiny (1)}-\varepsilon^{\tiny (2)}_{\rm f.s.} \right]+\varepsilon^{\tiny (2)}$, where $\varepsilon^{\tiny (1)}$ and $\varepsilon^{\tiny (2)}$ are the emissivity terms for the redistributed and partially coherent radiation, respectively, and where "f.s." implies that the corresponding term must be evaluated assuming a flat-spectrum average of the incident radiation.