PyEMILI: A New Generation Computer-aided Spectral Line Identifier -- II. Emission-line Identification and Plasma Diagnostics of a Sample of Gaseous Nebulae

TL;DR

PyEMILI II expands a Python-based spectral-line identification framework to robustly identify faint lines in deep nebular spectra and to perform plasma diagnostics from optical recombination lines. The paper adds a comprehensive atomic transition database via Kurucz Line Lists and a new recombination-line fitting module that uses MCMC to derive $T_e$, $N_e$, and ionic abundances from ORLs, aided by an emissivity cube for key ions. Application to 34 deep spectra across PNe, H II regions, and HH objects shows high identification agreement with literature IDs (often >90%), demonstrates corrections and enhancements to prior line identifications, and provides new ORL-based abundances. The ORL diagnostics reveal an anti-correlation between $T_e$(O II) and the abundance discrepancy factor (ADF), supporting a two-component nebular model with a cold, metal-rich phase dominating ORLs. Overall, PyEMILI proves to be a powerful tool for both line identification and advanced plasma diagnostics in deep nebular spectroscopy, with publicly available data and a path toward broader applicability.

Abstract

In order to test the robustness and reliability of the new generation spectral-line identifier PyEMILI, as initially introduced in Paper I, in line identification and establish a reference/benchmark dataset for future spectroscopic studies, we run the code on the line lists of a selected sample of emission-line nebulae, including planetary nebulae (PNe), HII regions, and Herbig-Haro (HH) objects with deep high-dispersion spectroscopic observations published over the past two decades. The automated line identifications by PyEMILI demonstrate significant improvements in both completeness and accuracy compared to the previous manual identifications in the literature. Since our last report of PyEMILI, the atomic transition database used by the code has been further expanded by cross-matching the Kurucz Line Lists. Moreover, to aid the PyEMILI identification of numerous faint optical recombination lines (ORLs) of CII, NII, OII and NeII, we compiled a new dataset of effective recombination coefficients for these nebular lines, and created a new subroutine in the code to generate theoretical spectra of heavy-element ORLs at various electron temperature and density cases; these theoretical spectra can be used to fit the observed recombination spectrum of a PN to obtain the electron temperature, density and ionic abundances using the Markov-Chain Monte Carlo (MCMC) method. We present MCMC-derived parameters for a sample of PNe. This work establishes PyEMILI as a robust and versatile tool for both line identification and plasma diagnostics in deep spectroscopy of gaseous nebulae.

Figures (16)

• Figure 1: Number distribution of atomic transitions, both the newly added (6154 transitions, in dark blue) and the entire atomic transition database ($\sim$891,000 transitions, in light blue) used by PyEMILI, classified by elements.
• Figure 2: Emissivities of the @series O ii O ii O ii$\lambda$4649.13 M1 2p$^{2}$3s $^{4}$P$_{5/2}$ -- 2p$^{2}$3p $^{4}$D$^{\rm o}_{7/2}$ (top-left) and $\lambda$4089.29 M48a 2p$^{2}$3d $^{4}$F$_{9/2}$ -- 2p$^{2}$4f G[5]$^{\rm o}_{11/2}$ (top-right), and @series N ii N ii N ii$\lambda$5679.56 M3 2p3s $^{3}$P$^{\rm o}_{2}$ -- 2p3p $^{3}$D$_{3}$ (bottom-left) and $\lambda$4041.31 M39b 2p3d $^{3}$F$^{\rm o}_{4}$ -- 2p4f GG[9/2]$_{5}$ (bottom-right) nebular lines, as a function of electron temperature and density; these four lines are the strongest transitions of the 3s--3p and 3d--4f arrays of the two ions. Emissivity $\epsilon$ is defined by Eq 1 using the effective recombination coefficients calculated by NIIcoeNIIcoe_corr and OIIcoe. The color bar of each panel indicates the emissivity value in logarithm, $\log{\epsilon(\lambda)}$+26.
• Figure 3: VLT/UVES echelle spectrum of NGC 6153 showing the faint nebular emission lines identified by PyEMILI (this work, labeled with red IDs) and also detected and identified in the deep spectroscopy of 2000MNRAS.312..585L. The black "?" indicates the line was unidentified in 2000MNRAS.312..585L. The spectrum has been velocity-corrected through the measurements of the @series H i H i H i Balmer lines. Sky emission lines are marked by blue vertical-dashed lines.
• Figure 4: Comparison of the observed spectrum (black) of NGC 6153 with the theoretical spectra (red) of @series O ii O ii O ii and @series N ii N ii N ii obtained using the specified electron temperature, electron densities, and ionic abundances.
• Figure 5: Electron temperature derived with the O ii recombination lines detected in PNe as a function of ADF in logarithm scale. Black-filled circles are the sample of PNe from Table \ref{['DIAGNOSTICS']}, whose $\log{T_{\rm e}}$(@series O ii O ii O ii) values were obtained through MCMC sampling. Blue squares are the four PNe (NGC 7009, NGC 6778, M 1-42, and Ou 5; see text for the references) whose temperatures were derived from the observed O ii$\lambda$4649/$\lambda$4089 line ratio, using the O ii effective recombination coefficients calculated by OIIcoe; the blue downward arrow on Ou 5 indicates the O ii temperature of PN is an upper limit. ADF values were adopted from literature (see Table \ref{['DIAGNOSTICS']} and description in the text). The grey dashed line is a linear fit to the whole sample (see the legend on top of the figure).