Ionized Gas Emission Lines

In the current version, the emission lines of ionized gas in galaxies are modeled exclusively as originating from star-forming HII regions. These are implemented through the spec1d.HII_Region class.

Overview

The synthetic spectrum of ionized gas is constructed by combining multiple emission lines with Gaussian profiles. The base templates come from FSPS-Cloudy simulations (Byler et al. 2017), which tabulate emission line flux ratios under various gas metallicities.

Two flux calibration modes are supported:

  • Hα-calibrated mode: If an observed Hα flux is provided, the spectrum is scaled accordingly.

  • SFR-calibrated mode: If no Hα flux is given but sfr, z, and vpec are provided, the Hα luminosity is inferred via the Kennicutt (1998) relation.

Emission Line Generation

Each emission line is represented by a Gaussian function:

\[\mathcal{E}(\lambda) = \frac{1}{\sigma_{\rm line} \sqrt{2\pi}} \exp\left[ -\frac{(\lambda - \lambda_{\rm line})^2}{2\sigma_{\rm line}^2} \right]\]

where \(\sigma_{\rm line}\) is the line width in wavelength space, computed from the velocity dispersion \(\sigma_v\) via:

\[\sigma_{\rm line} = \frac{\sigma_v}{c} \lambda\]

The total spectrum is given by the weighted sum of all lines:

\[\mathcal{S}(\lambda) = \sum_i \mathcal{L}_i \mathcal{E}_i(\lambda)\]

where \(\mathcal{L}_i\) is the flux of the :math:`i`th line relative to Hα, determined from the model template.

Template Selection

The emission line templates are selected based on gas-phase metallicity (logz). Available models include:

  • 'hii': HII region templates based on FSPS + Cloudy simulations.

  • 'nlr': Narrow-line region templates for AGN (used in other modules).

The nearest metallicity grid point is used; no interpolation is performed. The range of supported metallicity is:

\[\log Z/Z_\odot \in [-2.0, 0.5]\]

Inputs outside this range will be clipped with a warning.

Velocity Broadening

Line broadening due to gas velocity dispersion is applied using a pixel-wise Gaussian convolution:

  • The intrinsic resolution of the line template is deconvolved from the target dispersion.

  • The convolution kernel is constructed per pixel using an adaptive window.

  • This process is implemented in the gaussian_filter1d function.

Dust Attenuation

Dust extinction is applied using the Calzetti et al. (2000) attenuation law:

\[F_{\rm obs}(\lambda) = F_{\rm int}(\lambda) \cdot 10^{-0.4 \cdot E(B-V) \cdot k(\lambda)}\]

where \(k(\lambda)\) is the extinction curve derived from the input ebv and Rv parameters. This reddening is applied to the entire line spectrum prior to flux calibration.

Redshift and Peculiar Velocity

The line-of-sight redshift of the emission spectrum is determined by either:

  • the input vel (if sfr is not used), or

  • the cosmological + peculiar velocity (z, vpec) if using SFR-based calibration:

\[v_{\rm los} = (1 + z) \cdot v_{\rm pec} + z \cdot c\]

The final spectrum is shifted accordingly using spline interpolation in wavelength space.

SFR-Calibrated Mode

If the Hα flux is not given, the model estimates the Hα luminosity using:

\[L({\rm H}\alpha) = 7.9 \times 10^{41} \cdot {\rm SFR}~[\text{erg/s}]\]

This is converted to observed flux using the luminosity distance derived from redshift:

\[F_{\rm obs} = \frac{L({\rm H}\alpha)}{4\pi D_L^2 (1+z)^4}\]

Here, the \((1+z)^4\) term accounts for cosmological surface brightness dimming. The emission line spectrum is then scaled to match this predicted Hα flux.

Output Spectrum

The final synthetic spectrum includes:

  • Proper flux calibration (via Hα or SFR),

  • Dust attenuation,

  • Gaussian broadening by velocity dispersion,

  • Redshift and Doppler shift,

  • Wavelength resampling onto the user-defined grid (config.wave).

All emission lines are modeled together and returned as a 1D array representing the final observed-frame flux.

HII_Region is designed to be used in both single-spectrum and datacube simulations.