Science tools: integrated flux units

[Jamie Budynkiewicz]

In ChandraCXC/iris#107, we had fixed the integrated flux calculation so that we did not divide the end result the effective wavelength. This fixed an error where increasing the passband range high end-point resulted in lower flux values, which was incorrect. The result of all integrated fluxes are in erg/s/cm2.

However, now we're stuck with the opposite problem: calculating the flux from a relatively small passband, like any of the SVO filters, isn't really in units of erg/s/cm2; it should still be in flux density units erg/s/cm2/Angstrom.

You can test this adding a SED, interpolating it, calculating the fluxes under some SVO filters, then adding these points to a new SED, and co-plotting the original SED and the integrated fluxes. The fluxes do not lineup with the original SED.

Two possible solutions:

1. have Iris figure out the proper units for the integrated fluxes
2. let the user choose the units after the flux is calculated ("flux" or "flux density" .... or "erg/s/cm2" or "erg/s/cm2/Angstrom"). They would be allowed to select the unit for each integrated point, in case there are both large broadband filters and short photometry filters in the Passband list.

I like choice (2) because it gives the user full freedom over the units, and doesn't force us to come up with a smart way to decide on the units.

[Jamie Budynkiewicz]

Another idea:

• If the filter is a photometric filter, use flux density. If it's a simple flux band, report it in flux units.

And to expand on point (2) above, this would be an option selected as you define the passband/photometric filter