Detailed Description of Parameters

There are two parameters files - parameters_master and parameters_model. The distinction between where a parameter goes is somewhat arbitrary, but is based on the following. Parameters_master tends to contain parmaeters that will likely be shared amongst all snapshots in a given galaxy run, while parameters_model are parameters that might change from run to run (like the snapshot name or the galaxy center).

The parameters files are (somewhat non-traditionally) written as python code files because this enables you to embed little snippets in them that may be particular to your galaxy run. For example, for gadget snapshot naming conventions, it can be useful to have a snippet along the lines of:

Gadget_snap_num = 20
if Gadget_snap_num < 10:
   snapnum_str = '00'+str(Gadget_snap_num)
elif Gadget_snap_num >= 10 and Gadget_snap_num <100:
   snapnum_str = '0'+str(Gadget_snap_num)
   snapnum_str = str(Gadget_snap_num)

Gadget_snap_name = 'snapshot_'+snapnum_str+'.hdf5'

That puts the correct number of 0’s in front of the snapshot name.


Resolution Keywords

oref:Over Refinement of the Octree - 1 means each data holding cell (a False) gets split into 8 one more time. Very heavy on the memory. Default is 0.
n_ref:Threshold number of particles to refine over. When nparticles > n_ref the octree refines further. Default is 64.
zoom_box_len:Side length to zoom in on. Is +/- zoom_box_len from the center. Units are proper kpc. So, a grid centered on [0,0,0] with zoom_box_len = 200 would extend from [-200,200] kpc in physical units at the redshift of the simulation.
bbox_lim:Initial bounding box of grid for SPH simulations (+/- bbox_lim). Units are kpc. This must encompass all of the particles in a simulation currently. This just has to be a big number, but you want to be careful of making too large as precision limitations only allow for up to 20 levels of refinement.


n_processes:Number of MPI processes to run the radiative transfer on. Note, the stellar population synthesis will only run on as many processors as are on a core since its parallelization is (not MPI)

RT Information

For all photon counts, a decent rule of thumb is 10-100x the number of grid cells that you have, though of course you should check the convergence properties of your simulation.

 Number of photons to use in main iterations (for the whole grid) for specific energy and temperature calculations.
 Number of photons to use for the SED/image calculation
 If raytracing is set (which is the default hard-coded into the code), number of raytracing photons to use for source emission.
 Similar to n_photons_raytracing_sources but for dust emission.

Dust Information

dustdir:String. Path to where your dust files are. String format - (e.g. ‘/home/desika/hyperion-dust-0.1.0/dust_files/’)
dustfile:String. Name of your main dust file. String format - (e.g. ‘d03_3.1_6.0_A.hdf5’)
PAH:Boolean - True means use model for PAHs, False means don’t.
 Dust mass to metals mass ratio
 Boolean. False ensures energy conservation. But the emisivities may not be strictly correct if the energy in a cell is out of range of the emissivities. True modifies the energy in the simulation, but ensures that the emissivities are consistent with the energy. See: <>

Hydro Code Units

Currently these are actually not used in powderday). They remain in the parameters file as a placeholder though as we may need them as an over-ride if we find some HDF5 files don’t contain this information.

unit_mass:Mass code units for galaxy simulation. Units: Msun/h
unit_length:Length code unit for galaxy simulation. Units: kpc/h
unit_age:Stellar age units. Units: Gyr/h
unit_velocity:Velocity code unit for galaxy simulation. Units: cm/s

Stellar SEDs Info


Boolean. True means force binning of the stellar SEDs (in bins of age and metallicity). False means don’t. False results in an exact solution since the stellar SEDs are individually represented (as opposed to broken up into bins). This said, this can be very slow to run, and extremely hard on the memory.


Boolean. True means this is a cosmological simulation, False means idealized galaxy simulation.


IMF parameter for stellar pops calculations.

  1. Salpeter
  2. Chabrier
  3. Kroupa
  4. Van Dokkum
  5. Dave

Though note options 3 and 4 are currently not supported.


Weight given to post AGB stars. 1 is the default.


Boolean. If set to True, then enables the Charlot & Fall birthcloud models for all stars with age younger than birth_cloud_clearing_age.


Stars with age < birth_cloud_clearing_age have Charlot & Fall birthclouds (if CF_on == True). Meaningless if CF_on == False. Units: Gyr.


Forced metallicity increase in the newstar particles. Useful for idealized galaxy simulations where the stars can form out of pristine gas. Units are absolute (so 0.02 = Solar). Setting to 0 (default) means that you use the stellar metallicities as they come in the simulation (i.e. for Cosmological simulations).


Age in Gyr of disk stars for idealized simulations. Meaningless for cosmological simulations. Note, if this is <=7, then these will live in Charlot & Fall birthclouds (if CF_on = True).

Note, for Gadget simulations, stars are divided into newstars, disk stars and bulge stars. For Tipsy outputs, the stars initalized with the simulation are auto-detected by their nonsensical ages, and assigned as disk stars. So, if there are stars initalized with your Tipsy simulation, assign their ages (and metallicities below) as disk stars.


As disk_stars_age but for bulge stars.


Metallicity of disk stars in FSPS metallicity units. See last page of FSPS manual for numbers. (e.g. 20 = Solar for Padova + BaSeL tracks). Meaningless for cosmological simulations.


As disk_stars_metals but for bulge stars.


Number of bins to bin the stellar ages in (boundaries are the oldest and youngest star particles; linear bins in log(age)).


Meaningless parameter; place holder for future code additions.


String. Location of the metallicity maps in FSPS for the stellar libraries you use. Currently Padova2007 is the default (hard coded into powderday), so this should point to something like: “/Users/desika/fsps/ISOCHRONES/Padova/Padova2007/zlegend_basel.dat”

Images and SED Parameters

NTHETA:Number of polar angles to view galaxy at


 Boolean. False means the simulation automatically centers on the cell with the highest gas density peak. True means you center on x_cent,y_cent,z_cent as given in parameters_model.


You should probably never touch any of these.


 String - currently the snapshot name of your galaxy run. (Naming will change as other front ends built).
hydro_dir:Location of snapshots
PD_output_dir:String - location of where powderday output files should go.
Auto_TF_file:String - name of the TF logical file to be written (doesn’t need a path - will go into PD_output_dir)
 String - name of the dust density ascii file to be written (doesn’t need a path - will go into PD_output_dir)
inputfile:String - name of the input HDF5 (rtin) file for powderday to write before radiative transfer begins.
outputfile:String - name of the output HDF5 (rtout) file after radiative transfer
x_cent:Location in grid coordinates of the x-coordinate of the center of your galaxy. Only pertinenet if MANUAL_CENTERING==True. Otherwise ignored by powderday.
y_cent:As x_cent but for the y-coordinate
z_cent:As x_cent but for the z-coordinate.