This code implements a model to predict nebular emission in galaxies. The code performs an optimal bilinear interpolation on a multi-dimensional grid of photo-ionization models to retrieve multiple emission lines for objects with a given input set of properties. The code is written in Python and it was conceived to generalise the results that were published in Orsi et al. 2014 so that users can incorporate this model easily.
This code is pretty straightforward to install and use. After cloning the repository, add it to your python path (e.g. export PYTHONPATH="${PYTHONPATH}:/home/your-path-to-get_emlines/") and use it as in the example below.
Make sure you have the following packages installed:
- numpy
- matplotlib
- scipy
The grid dataset is included in this repository.
import get_emlines as lines
sfr = 0.1 # Msun/yr
Z = 0.01 # M_metals/M_gas
lums = lines.get_emlines(sfr,Z)
# OIII 5007 luminosity:
print lums['OIII_5007']
# All luminosities available should retrieve something like this:
lums.dtype
Out[3]: dtype([('Lyalpha', '<f4'), ('Hbeta', '<f4'), ('Halpha', '<f4'), ('OII_3727', '<f4'), ('OII_3729', '<f4'),
('OIII_5007', '<f4'), ('OIII_4959', '<f4'), ('OI_6300', '<f4'), ('NII_6548', '<f4'), ('NII_6584', '<f4'),
('SII_6717', '<f4'), ('SII_6731', '<f4'), ('NeIII_3870', '<f4'), ('CII_158um', '<f4'), ('NII_205um', '<f4')])
# Another example with an array of dummy galaxies:
import numpy as np
sfr = np.logspace(-2,2,1e3) ; z = np.linspace(1e-3,1e-1,1e3)
# Testing the verbose keyword as well
lums = lines.get_emlines(sfr,z,verbose=True)The messages printed in the above example (verbose activated) should look something like this:
DEBUG - get_emlines(): verbose output activated
DEBUG - get_lumlines(): Rootdir:
INFO - get_lumlines(): Computing emission lines for 1000 galaxies
DEBUG - get_2dfunc(): Constructing grid function for all lines
DEBUG - get_2dfunc(): Done constructing 2D grid function
DEBUG - get_2dfunc(): Constructing grid function for Halpha
DEBUG - get_2dfunc(): Done constructing 2D grid function
DEBUG - get_lumlines(): Running lines with g0=-1.300000
INFO - get_emlines(): Luminosities computed OK
# some output luminosities:
print lums['Halpha'].min(), lums['Hbeta'].max()
39.2671 42.8018You can access the list of lines available in, e.g.:
HIImodels/Lines/LineInfo_Levesque10
By default, the list of lines available (and their names) are:
# Emission Lines contained in file LineData_Levesque10
# id Emission Line Central Wavelength
0 Lyalpha 1216.0
1 Hbeta 4861.0
2 Halpha 6563.0
3 OII_3727 3727.0
4 OII_3729 3729.0
5 OIII_5007 5007.0
6 OIII_4959 4959.0
7 OI_6300 6300.0
8 NII_6548 6548.0
9 NII_6584 6584.0
10 SII_6717 6717.0
11 SII_6731 6731.0
12 NeIII_3870 3870.0
13 CII_158um 1.5800
14 NII_205um 2.0500
The central wavelgnth of the last two FIR lines is in um, the rest in Angstroms.
This code computes galaxy line luminosities log(L [erg s-1]) based on an input SFR [M_sun/yr] and metallicity Z. To assign emission lines to galaxies, first we assign an ionization parameter q, which is assumed to be related to the gas-phase metallicity by a power-law:
Then the code makes use of the grid of photo-ionization models of Levesque et al. 2010 to perform a bilinear interpolation over q and Z, leaving the electron density fixed. Finally, to scale line fluxes to galaxy-wide luminosities, the code uses the input SFR to infer the H-alpha luminosity, and then use its predicted flux to infer all other line luminosities.
. You can change the slope g0 or normalization q0 from their standard values (from Orsi+14) by doing:
lum_disk = lines.get_emlines(sfrdisk, zdisk, g0 = -1.5,q0 = 3.5e7)In the above the slope and normalization changed to -1.5 and 3.5e7, respectively.