Input Files

MCSED is designed to model the SEDs of a set of galaxies with a common set of photometric and spectroscopic data. Since MCSED can accept a wide range of constraints, with dozens of possible filter combinations, emission line fluxes, and/or absorption line spectra indices, the format of the input file has some flexibility. But the basic file structure is simple: the data are entered in a simple, space-delimited ascii file, with the first line of the file labeling the file’s columns. The rest of the file gives the measured flux densities and/or emission-line fluxes and/or absorption line spectral indices for each object, one object per line.

Required Columns

The input data file has three required columns, which must have these exact labels: Field, ID, and z. In other words, an object’s identification consists in two parts: a string which contains the name of the field in which the object is found, and a unique integer ID which is specific to that field. If both field and ID are unnecessary, one can simply enter a placeholder for one of the entries. Redshifts must be specified for every source.

The remaining columns in the input file should be pairs of numbers representing photometric flux densities (and their \(1\,\sigma\) errors), emission line fluxes (and their errors), and/or absorption line indices (and their errors). Since the quoted uncertainties for photometric observations often do not include systematic and/or external errors, MCSED also allows the user to specify a minimum fractional uncertainty for any type of observation. The defaults for the minimum errors can be found in config.py, and by default are set to phot_floor_error = 0.05 (for photometric errors), emline_floor_error = 0.05 (for errors in emission-line fluxes), and absindx_floor_error = 0.05 (for errors in absorption line indices). These defaults can be changed by editing the above parameters in config.py.

Photometry

Using Skelton et al. (2014)

MCSED was originally written to analyze galaxies in the five CANDELS fields (AEGIS, COSMOS, GOODS-N, GOODS-S, and UDS), hence there are special commands built into the program to handle the PSF-matched photometry from Skelton et al. (2014). If the user’s sources are in the Skelton catalog, the objects can be specified by their field (i.e., aegis, cosmos, goodsn, goodss, or uds) and the unique Skelton ID number. This links the input line directly to the object’s photometry in the files provided by Skelton et al. (2014). Momcheva et al. (2016) provide grism redshifts (and emission line fluxes) for all Skelton photometry. Users with Skelton sources are encouraged to use the grism redshifts for the redshift column.

Additional photometry not included in the Skelton catalogs can be specified in the input file in the same way as general photometry as discussed in General Case.

General Case

If the input objects are not associated with the Skelton et al. (2014) catalog (identified via the Field and ID columns described above), or if users wish to supplement this catalog with additional photometry, the input file must include additional columns. Photometric measurements should be given as flux densities with \(1\,\sigma\) uncertainties associated with each measurement (null value \(=-99\)). The columns containing these data in the input file should be labeled f_filter_name and e_filter_name, where filter_name is the name of a .res file in the FILTERS directory. (In other words, columns named f_hst_acsF606W and e_hst_acsF606W should refer to the flux densities (not magnitudes!) and uncertainties taken through the filter defined in FILTERS/hst_acsF606W.res.) Following Skelton et al. (2014), the units for flux density are scaled to an AB magnitude of 25, so \(1.00 = 3.63 \times 10^{-30}\) ergs cm\(^{-2}\) s\(^{-1}\) Hz\(^{-1}\) (e.g., if the user’s flux densities are in \(\mu\)Jy, the values must be multiplied by \(10^{0.4(25-23.9)} \approx 2.754\)).

Emission Lines

MCSED can include emission line fluxes in the likelihood function. To do this, the user first specifies the line’s name (keyword Name), rest-frame wavelength (in Angstroms), and relative weight in the config.py emission-line dictionary. A weight of 1.0 means the line contributes just as much weight to the likelihood function as a photometric data point; a weight of 0.0 implies that the line is ignored. The user then provides the objects’ emission line strengths and \(1\,\sigma\) error bars by entering the data in the input file and labeling the columns as Name_FLUX and Name_ERR, where Name is the line’s keyword listed in the emline_list_dict dictionary defined in config.py. The emission line fluxes and errors must be specified in units of \(10^{-17}\) ergs cm\(^{-2}\) s\(^{-1}\), unless a different multiplication factor to the base unit of ergs cm\(^{-2}\) s\(^{-1}\) is specified by the keyword emline_factor in config.py. The emission lines currently included in config.py are given below. Additional lines can be added by expanding the emline_list_dict in config.py.

Emission Lines Definitions
Line Name Wavelength (Å) Weight
H\(\beta\) Hb 4861 1.0
H\(\alpha\) Ha 6563 1.0
[O III] OIII 5007 0.5
[O II] OII 3727 0.5
[N II] NII 6583 0.5

Currently, MCSED cannot fit blended emission lines.

Absorption Line Indices

Absorption line indices can also be used in MCSED’s likelihood function. These measurements are input in a similar way as additional photometry or emission line fluxes are included. In the input file, the columns containing an absorption line index and its uncertainty are labeled as Name_INDX and Name_Err, where Name is the line’s keyword, as listed in the absorption_index_dict dictionary defined in config.py. The indices that are pre-defined in MCSED are listed in the table below. As one can see from the table, the indices are defined via their wavelength ranges, the units they are quoted in, and a relative weight similar to that defined for the emission lines.

Absorption Line Indices Definitions
  Index Band (Å) Blue Continuum (Å) Red Continuum (Å)  
Name Weight Blue Red Blue Red Blue Red Units¹
Lick_CN1 1.0 4142.125 4177.125 4080.125 4117.625 4244.125 4284.125 1
Lick_CN2 1.0 4142.125 4177.125 4083.875 4096.375 4244.125 4284.125 1
Lick_Ca4227 1.0 4222.250 4234.750 4211.000 4219.750 4241.000 4251.000 0
Lick_G4300 1.0 4281.375 4316.375 4266.375 4282.625 4318.875 4335.125 0
Lick_Fe4383 1.0 4369.125 4420.375 4359.125 4370.375 4442.875 4455.375 0
Lick_Ca4455 1.0 4452.125 4474.625 4445.875 4454.625 4477.125 4492.125 0
Lick_Fe4531 1.0 4514.250 4559.250 4504.250 4514.250 4560.500 4579.250 0
Lick_Fe4668 1.0 4634.000 4720.250 4611.500 4630.250 4742.750 4756.500 0
Lick_Hb 1.0 4847.875 4876.625 4827.875 4847.875 4876.625 4891.625 0
Lick_Fe5015 1.0 4977.750 5054.000 4946.500 4977.750 5054.000 5065.250 0
Lick_Mg1 1.0 5069.125 5134.125 4895.125 4957.625 5301.125 5366.125 1
Lick_Mg2 1.0 5154.125 5196.625 4895.125 4957.625 5301.125 5366.125 1
Lick_Mgb 1.0 5160.125 5192.625 5142.625 5161.375 5191.375 5206.375 0
Lick_Fe5270 1.0 5245.650 5285.650 5233.150 5248.150 5285.650 5318.150 0
Lick_Fe5335 1.0 5312.125 5352.125 5304.625 5315.875 5353.375 5363.375 0
Lick_Fe5406 1.0 5387.500 5415.000 5376.250 5387.500 5415.000 5425.000 0
Lick_Fe5709 1.0 5696.625 5720.375 5672.875 5696.625 5722.875 5736.625 0
Lick_Fe5782 1.0 5776.625 5796.625 5765.375 5775.375 5797.875 5811.625 0
Lick_NaD 1.0 5876.875 5909.375 5860.625 5875.625 5922.125 5948.125 0
Lick_TiO1 1.0 5936.625 5994.125 5816.625 5849.125 6038.625 6103.625 1
Lick_TiO2 1.0 6189.625 6272.125 6066.625 6141.625 6372.625 6415.125 1
Lick_Hd_A 1.0 4083.500 4122.250 4041.600 4079.750 4128.500 4161.000 0
Lick_Hg_A 1.0 4319.750 4363.500 4283.500 4319.750 4367.250 4419.750 0
Lick_Hd_F 1.0 4091.000 4112.250 4057.250 4088.500 4114.750 4137.250 0
Lick_Hg_F 1.0 4331.250 4352.250 4283.500 4319.750 4354.750 4384.750 0
D4000 1.0 …… …… 3750.000 3950.000 4050.000 4250.000 2
¹Unit codes: 0 = Å; 1 = mag; 2 = ratio

These definitions come from Bruzual (1983) and Worthey et al. (1994); they are calculated by finding the average value of \(F_{\lambda}\) within the blue and red continuum bands, interpolating a line through these values to estimate the continuum, \(F_C\), and then computing equivalent width via

\[{\rm EW} = \int_{\lambda_1}^{\lambda_2} \left( 1 - \frac{F_{\lambda}}{F_C} \right) d\lambda\]

Important Note: absorption line indices are defined for a specific spectral resolution. MCSED makes no attempt to match this resolution: it uses the SSP spectra as is. The user should consider this carefully before deciding on the utility of this feature.