Synthesize Johnson from Sloan

Science Data Analysis
1

Is there an accepted best (or least bad) way to synthesize Johnson B and V magnitudes from Sloan g’, r’ and i’?

1 Like
2

You might try Kostov & Bonev:

…who give relations using the Pan-STARRS data and Peter Stetson’s large collection of faint-star CCD photometry.

\Brian

3

thx I’ll check it out

4

Interesting how different Kostov and Bonev are from Jester Transformations between SDSS magnitudes and other systems - SDSS-III. At least the signs are the same.

5

Is there a Stetson/Bessell, no Cousins mix here?

6

Stetson provides Johnson-Cousins data on the Landolt system. So the Kostov & Bonev transformations go specifically from the Pan-STARRS flavor of the Sloan system to BVRcIc. What Bessell (two l’s) did was to describe passband systems and filters, and did not define photometric zero-points or colors.

\Brian

7

Spelling corrected, thanks.

So why a difference between Jester and Kostov & Bonev?
Are they using different pass bands?
Did the standard stars change color?

Ray

8

I do not know the answer fully, and it’s a bit of a mess. It is the case, however, that the Sloan systems differ slightly with each data release from the main SDSS survey. I’m under the impression that things were stabilized from DR7 onward, but I haven’t seen any tests. The nominal Sloan ‘standards’ by J. Allyn Smith:

… are on the ‘primed’ system, definitely different than the very large survey telescope catalogues. My feeling is that there aren’t any real standards in the Sloan system. Allyn has visited Lowell recently and said he intends to do re-reductions and extensions of his data, including unpublished stuff.
The original Pan-STARRS system is a bit different again. The more recent ATLAS ‘refcat2’, also derived from the Pan-STARRS data (plus much else) is nominally tied to the GAIA spectrophotometric fluxes, and ought to be the most reliable thing out there. I have been using the Kostov & Bonev transformations with this, and things seem to be OK for V, B-V, and V-R for stars of ordinary color. This catalogue is VizieR item j/apj/867/105 where you can look things up star-by-star. For Sloan r and i, I would adopt ‘refcat2’ values directly until something better comes along. In the south (up to +15 Dec), you might compare also the r and i photometry in SkyMapper DR4:

(VizieR item II/379).

The Kostov & Bonev relations are probably fine for most purposes, but it would have been nice if they’d done some additional statistically rigorous analysis looking at residuals etc. and shown the results.

\Brian

9

Thanks for the links Brian.
I suppose part of the differences is inconsistencies of stars.
Are undiscovered Cepheids masking as standards?
Another is atmospherics and aerosols.
Another is the state of or the cleanliness of the local optics.
The last would be simularly labeled filters with differing passbands.

I have been through a couple sets of filters and they have aged in the filter wheels. I wondered just what my pass bands were.
So I measured them with a spectrophotometer.
It is easy, it is cheap and I don’t know why any individual, let alone any monied and talented scientific consortium, would not do the same for every filter that they can get their hands on and make it part of their documentation.

I have to amend my note after reading the paper. It looks to me that they did an awful lot of work to get it right. They also make mention of lots of sources of error. So I dunno. I knew one of the authors and a couple more that worked on SDSS pipeline software. After many years of looking at comp stars, I doubt that they had time to really study their standards. The observation cadence was fairly sparse. That is a job that the pros could hand off to small telescopes.

Ray