Methodology¶
Overview¶
MapLines models emission-line spectra using parametric line profiles combined with Bayesian parameter estimation.
The approach is designed to analyze both integrated spectra and spatially resolved IFU observations.
Spectral models¶
The spectral model can include several components:
Gaussian emission lines
skewed Gaussian profiles
Lorentzian profiles
Voigt profiles
outflow components
power-law continuum
FeII templates
Each component is defined in the configuration file and combined to produce the total model spectrum.
Parameter inference¶
Parameter estimation is performed using Markov Chain Monte Carlo (MCMC)
sampling through the emcee ensemble sampler.
The posterior probability is defined as:
where:
\(L\) is the likelihood
\(P\) is the prior
\(D\) is the observed spectrum
Likelihood function¶
The likelihood assumes Gaussian uncertainties in the observed spectrum.
It is implemented in:
MapLines.tools.priors
Posterior sampling¶
Posterior sampling is performed using the routines in:
MapLines.tools.mcmc
These routines generate chains of model parameters that sample the posterior distribution.
Outputs¶
The fitting procedure produces several products:
best-fit spectra
posterior parameter distributions
parameter maps (for IFU data)
diagnostic plots
These products can then be used to study the physical properties of ionized gas in galaxies and active galactic nuclei.