Review sky brightness calculation

[dkirkby]

We currently apply extinction to the sky brightness before converting it to elec/sec via Survey.get_flux. Is this the right thing to do?

[dkirkby]

The relevant code from survey.py is:

        # Calculate the mean sky background level in detected electrons per pixel.
        self.mean_sky_level = self.get_flux(self.sky_brightness)*self.pixel_scale**2

and:

    def get_flux(self,ab_magnitude):
        """Convert source magnitude to flux.

        The calculation includes the effects of atmospheric extinction.

        Args:
            ab_magnitude(float): AB magnitude of source.

        Returns:
            float: Flux in detected electrons.
        """
        ab_magnitude += self.extinction*(self.airmass - 1.)
        return self.exposure_time*self.zero_point*10**(-0.4*(ab_magnitude-24))

[fjaviersanchez]

If I change the zero_points and I write ab_magnitude += self.extinction*(self.airmass - 1.2) it is slightly better than: ab_magnitude += self.extinction*(self.airmass - 1.0) but the flux is still ~0.5% off recomputing the zero_points for the new airmass conditions.

[fjaviersanchez]

I think we can't recompute the zero-points for the different conditions since we have the model for LSST but this might be (probably will be) different for DES and CFHTLS. What do you think @dkirkby?

[dkirkby]

Our DES & CFHTLS zero points are simply rescaled from the LSST zero points, so should have the same airmass scaling. In other words, I think we want to use:

self.extinction*(self.airmass - zeropt_airmass)

Since the zero-point already includes extinction at zeropt_airmass.

This reminds me that we need to update the DES & CFHTLS zeropoints, now that the LSST ones have changed!

[fjaviersanchez]

I updated these zero-points in PR #20. I also added DECAM g and DECAM z filters.