Potential bug in SED unit normalization

[0Navin0]

The following function assumes luminosity distance in meters, as also expected from the docstring. But the code actually returns Mpc units.

skyCatalogs/skycatalogs/utils/sed_tools.py

Line 187 in 240fde4

def dl(self, z):

Following the above assumption, the SEDs are wrongly normalised, which will change the flux/mag calculations because here galsim expects fnu units for the resulting SED.

skyCatalogs/skycatalogs/utils/sed_tools.py

Line 202 in 240fde4

sed_array /= (4.0*np.pi*(self.dl(redshift_hubble))**2)

Is there some mistake in my understanding here?
It seems this routine was written to read from the stored skyCatalog data. Are routines creating the output fluxes also affected?

[sidneymau]

I think most of the details here are wrapped up in the magic factor here: https://github.com/LSSTDESC/skyCatalogs_creator/blob/main/skycatalogs_creator/diffsky_sedgen.py#L36-L41

the dl function should say that it returns the distance in Mpc, which is what that prefactor used in the GalSim SEDs assumes. The distance calculation itself is deferred until the catalogs are read in for simulation and does not appear to be applied to the precomputed SEDs