While pondering optimal ways of getting atmospheric extinction in the course all-sky photometry, it occurred to me that it ought to be more convenient to use fields at sec z ~2.2 - 2.5 _due_south_ at intermediate Declinations, thus between about -25 and -30 Dec for mid-northern observers (similarly in the north for southern-hemisphere observers). This instead of using the usual equatorial Landolt fields at large E-W hour-angles, which some telescopes complain about. The advantage, it would seem to me, is that the fields don't change airmass rapidly, so that you have at least half an hour on either side of the meridian to get the data. With the equatorial fields, you have to time things rather closely, and there is inevitably a significant range in airmass in the data if you do multiple images/filters.
I note that Arne Henden has observed a number of fields on "many nights" in Selected Areas at -15 Dec anchored by pairs (usually) of Landolt stars published in his 1983 "equipment stability" standards paper (usually overlooked):
https://ui.adsabs.harvard.edu/abs/1983AJ.....88..853L/abstract
I'm guessing he used those (at USNO-Flagstaff) to get extinction data quickly, but nothing is published about them.
I have poked around the literature and in the Peter Stetson files to identify a few such fields as candidates, and hope to try them out in the coming months. (I'll be pleased to share the list.) I wonder if folks see some pitfalls in doing this, or if others have tried the scheme. I also have access to a 1-m telescope on an alt-az mount. Since in this case there is no hour-angle issue pointing anywhere in the sky, one could play the same trick by having standard fields at about +80 Dec, so that you could observe them at lower culmination at sec z ~2.35 or so, supposedly the optimum airmass.
\Brian
Hi Brian,
While at USNO-Flagstaff Station, I used a multitude of techniques for extinction, depending on what I needed. Some nights were gong to be short, due to personal commitments or weather; some nights I wanted to follow specific fields for second-order extinction determination. I tried to use standard fields scattered over the sky in order to randomize azimuth-dependent extinction.
The problem for CCD observers has always been to find multiple standard stars in the field of view. This is why the Landolt fields were so convenient, while the southern E-region standards are less useful. I typically used a few Landolt fields around -15 Dec (such as Markarian A), as you mention, which transit at X=1.5, but observed them an hour or two off of the meridian. You want a bunch of calibrated stars in each field, and stars should be fairly bright so observing standards takes less of the night hours. Meridian is great for fork or alt-az telescopes, less so for GEMs. I don't remember why I didn't look for a few -30 fields.
You can use poorer-quality stars for extinction, if you calculate extinction separately from the transformation coefficients, as extinction has more scatter in general due to changing sky conditions. You can also do extinction within a single CCD field of view, as you can measure significant difference from one edge to another with enough stars.
You might post any fields that you find!
Arne
Thanks to Arne for his useful comments. I will attach to this note the working list I've built of some candidate fields for this experiment...
....looks as though one can no longer attach files, or I don't see how to do it. Will copy to:
https://ftp.lowell.edu/pub/bas/starcats/extin-flds.txt
The groups are of varying quality, ranging from utterly reliable to somewhat tentative. The effort has been to find fields with data reliably close to the standard system, and not necessarily with high internal precision. Many of the fields have additional candidate stars to the ones with good data; the idea is that after getting 10 or more nights (say) with consistent results, those extra stars become new standards that can be used more generally.
I have been following the precepts of Peter Stetson in thinking about this. Four of his papers are "required reading" in this regard, specifically:
https://ui.adsabs.harvard.edu/abs/2000PASP..112..925S/abstract
https://ui.adsabs.harvard.edu/abs/2005PASP..117..563S/abstract (the NGC 2419 'mea culpa' paper)
https://ui.adsabs.harvard.edu/abs/2009IAUS..258..197S/abstract (notice Figure 2 in re the limits to which one can match the Landolt system)
https://ui.adsabs.harvard.edu/abs/2019MNRAS.485.3042S/abstract (appendix B only)
...the papers might well be read in reverse order by date.
My opinion is that there is no reason for the standard fields to be in blank sky, but can be just as randomly located near something of astrophysical interest, such as an open or globular cluster, or a galaxy field. Thus things that get looked at often by many observers can have calibration in the same images. (The same reasoning could apply to Arne's many, many variable-star calibration fields.) My field centers are rigged for the 20'x20' field of the Lowell 1.1-m telescope + CCD and the similar field of a robotic 1-m PlaneWave we are using. I generally think of using standards between V mag 9 and 13, not wanting to spend any more time than necessary getting the standard-star data (Arne has made the same comment), consistent with shutter accuracy.
The attached lists are flat text in a somewhat kludged format, awaiting a decision about just what the file needs to look like for the reduction software. Each line can be quite long, over 200 spaces wide; rejigger as you see fit. The lists include some star-name, RA/Dec for J2000 (usually UCAC2), the proper motion (mainly from GAIA EDR3) in milliarsec/year, then the photometry as V magnitudes and color-indices, but also Sloan z as available. One could add separate U,B,V,R,I magnitudes, but almost all the science is in the color-indices, not the separate magnitudes. (The reductions will have to be done filter-by-filter, however.) I give a spectral type if available, either cribbed from my large file (item B/mk at VizieR) or (in the north) from my own determinations from LAMOST spectra. The end of each line shows the photometry data source, errors as available (mostly mean errors, though Arne's data have rms errors) and other comments. One problem is that most of the fields have a rather limited range in color, but often the range is not so different from a typical equatorial Landolt field at intermediate/high galactic latitude.
Specific fields:
SA 140 two Landolt stars of solar color that are good UBV standards, but low-weight for VRI; additional stars could provide some color extension.
NGC 253 this relies mainly on photometry by Alcaino & Liller; normally one would treat stuff from Alcaino with caution, but most of the data are from a follow-up paper that includes Bill Liller, and things seem to be OK when compared to other sources. Color range limited here.
NGC 1097 the first of several fields observed by Mario Hamuy and Jose' Maza in what appears to be a first-rate paper with sequences around southern galaxies having active nuclei. It is unfortunate that the results were given to only two decimals; based on the comparison in the Mrk 509 field (below) and elsewhere, the data are dead-on the Landolt system probably with only ~0.005 mag uncertainties in the mean.
NGC 1904 high-weight data mainly from Stetson, located on the NW periphery of the cluster.
MCG -05-23-16 another active galaxy field from Hamuy & Maza with some additional candidate stars to be calibrated using the others.
IC 4329A also from Hamuy & Maza, but a few stars overlap with Stetson. This is the galaxy in the Centaurus cluster where one of the first LIGO gravitational-wave events happened that got a lot of attention a few years ago (an actual supernova-like event appeared at visible wavelengths). The Stetson B,V,I data have low weight, coming from only two nights of data. Good color range here.
SA 132 a low-latitude field near the galactic center that I was using to help calibrate photometry of Titan. Besides the two Landolt stars, Arne has observed the field on a bunch of nights with results consistent with Landolt. The red star SA 132-801 is slightly variable in V, but the color-indices are surely stable at the couple-percent level.
NGC 6401 a second galactic-center field adopted for Titan observations. The field includes the globular cluster, but the interest for our purposes is Stetson data for the brighter field stars (not too crowded) with a good range in colors.
GSPC S596 another field that I have been using to get rough extinction measurements in recent years for Titan. Normally one would not take the GSPC data at face value, but this one has some external confirmation from SAAO observers, and both APASS and ASAS-3 are consistent with these. Probably OK at the couple-percent level. The brighter star toward the SE side should provide a good red extension once it is properly calibrated. (Other GSPC fields at -25 and -30 Dec might be suitable for this business, but I haven't looked too far into it --- fields too sparse?)
Mrk 509 this is an extension of the Landolt 'Mark A' field --- the four stars Arlo observed are simply too faint to be convenient. However, we have not only Stetson data of high weight in this much-observed field, but also Hamuy & Maza, Arne's results, and the 2013 Clem & Landolt CCD paper, all of which are quite consistent. I show the available data for each star so that intercomparisons can be made. I would adopt specific values roughly in the order: Stetson, Landolt, CL2013, other. The field, by the way, is best centered not far from the active galaxy Mrk 509, somewhat off from the very hot O-type subdwarf that Arlo concentrated on. (It is remarkable that the sdO star wasn't picked up in the many surveys for uv-bright objects.)
All for now.
\Brian
Hi Brian and Arne,
How do you work out extinction coefficients from high airmass standards only? Cheers,
Mark
Mark Blackford asked (quite reasonably): "How do you work out extinction coefficients from high airmass standards only
You don't! Implicit in this discussion was that the low-altitude (high-airmass) fields would be the complement to fields that are overhead. Since in the north we now have the Landolt standards fields at +45 Dec --- similar to the southern E-regions --- it is now straightforward to observe one near the meridian at about sec z ~1.02, and then go directly to one of the fields at about -30 Dec (sec z ~2.3 to 2.5 from Arizona) in the list I prepared, and thus have an optimal baseline for an extinction measurement from the two sets of data in fairly short order. One might add one or two equatorial fields (sec z ~1.25), for instance, if only to fill-in the range of star colors.
I have added two fields to the list mentioned previously, and have posted a new copy to the Lowell site here:
https://ftp.lowell.edu/pub/bas/starcats/extin-flds.txt
(yes, what used to be ftp is now https)
\Brian