Table of contents:
In this Tech Note, I examine the properties of the "Good" subset of the entire TASS database. This subset contains stars which were detected on at least 10 occasions, and which have had their R-band magnitudes corrected for a color term. The catalog can be found in the TASS archive as the cor_cat table. It contains 80,737 stars in total. Figure 1 shows the distribution of these stars on the sky.
I will compare stars in this subset against stars in other catalogs, to see how well the TASS photometry and astrometry agrees with that from other sources.
The LONEOS Project has collected a large amount of photometry on relatively faint stars all over the sky. Brian Skiff has been kind enough to share this information with others. You can read more about the catalog elsewhere on the TASS WWW site. I downloaded a copy of the June 13, 1998 version of the catalog from LONEOS, and refer to it below.
All stars in the LONEOS catalog have magnitudes in the B and V passbands, and some have R-band and I-band measurements, too. I selected stars from the LONEOS catalog near the equator, finding 380 stars between Declination -5 and +5 degrees. Comparing these stars to those in the "Good" subset of the TASS archive, I discovered only 11 matches; most of the LONEOS stars are slightly farther north than the TASS stars. Here they are:
# comparison of stars in LONEOS catalog of faint, low-precision BVRI # standards (provided by Brian Skiff, 13 June 1998 version) # vs. the TASS "cor_cat" subset of stars detected at least 10 times. # # the dRA column lists (TASS RA) - (LONEOS RA) in arcsec # the dDec column lists (TASS Dec) - (LONEOS Dec) in arcsec # the DV column lists (TASS V) - (LONEOS V) in mag # # from LONEOS from LONEOS arcsec #star RA Dec B V dRA dDec dV dR dI G11.1_4 0.6758 -3.3911 10.82 9.74 -0.04 0.80 0.11 - - F10.1_7 7.2583 -4.0653 11.84 11.16 0.49 1.12 0.12 - - F10.2_7 8.9050 -3.1194 12.05 11.43 4.91 2.45 -0.08 - - F10.2_3 8.9229 -3.1231 14.81 13.63 -3.10 1.05 -0.19 - - F13.1_5 19.8554 -3.7786 11.58 11.10 3.55 1.35 0.03 - - Hyp_2M 319.0467 -1.0647 12.07 11.46 -2.07 0.10 0.05 - - Hyp_2F 319.1208 -0.8131 13.50 12.88 -2.43 -4.22 -0.02 - - Hyp_2B 319.1237 -0.9033 10.98 10.42 -0.55 -0.09 0.01 - - Hyp_1B 319.1704 -1.0325 11.16 10.70 -0.21 -1.57 0.00 - - Hyp_3M 319.2392 -1.0556 12.84 11.99 2.40 -1.14 0.07 - - Hyp_3B 319.3071 -0.9028 10.93 10.45 -0.28 -0.16 0.02 - -Note that none of the eleven stars have measured R-band or I-band magnitudes in the LONEOS catalog. Rats.
Eleven stars is a small overlapping sample, but we can nonetheless compute the mean and standard deviation of the differences between their properties in the two catalogs:
In each case, the two catalogs are consistent within the errors. This is good -- but remember, there are only 11 stars in the overlapping sample. We really need more stars in order to make strong statements about the accuracy of TASS measurements.
The US Naval Observatory has a special interest in astrometry: it can help sailors to find their location on a map. One of their many projects is the USNO A1.0 catalog, which is based on scans of large Schmidt photographic plates. The USNO WWW pages on this project seem to have disappeared, but you can select information from this catalog on the Web at the ESO/ST ECF USNO A1.0 Catalog Server.
The USNO A1.0 catalog contains precise measurements of position, which I will use herein. It also lists photometry in two passbands: b-band is based on blue-senstive plates, and r-band on red-sensitive plates. Neither is precisely equivalent to its Johnson-Cousins counterpart, so I will not compare the USNO A1.0 photometry against that from TASS.
Arne Henden expressed an interest in finding out if the quality of TASS measurements decreases in crowded fields. I decided to use the USNO A1.0 catalog as a source of precise position, and compare them to the mean positions of "Good" TASS stars. I picked two fields, one far from the galactic plane (and hence uncrowded), the other near the plane (and hence crowded). Each field is roughly 1 degree in size: the USNO server picks stars within a 30-arcmin radius, whereas the TASS server picks stars in a square with sides of length 60 arcmin.
J2000.0 galactic # USNO # TASS
Field RA Dec latitude stars stars # match
----------------------------------------------------------------------------
A 15 -2 -65 348 94 63
B 295 -3 -12 4449 267 149
The column labelled "# match" contains the number of TASS stars which
matched up with a single USNO star, using a maximum matching radius
of 15.0 arcsec.
Many TASS stars matched up with 2 or 3 USNO stars, since the USNO
catalog goes much deeper than TASS.
I ignore all stars with multiple matches.
How well do the positions agree? I divided the stars into 3 catagories, based on their TASS V-band magnitudes: bright (V < 10), medium (10 < V < 13), and faint (V > 13). For each star, I calculated its difference in RA, in Dec, and the total distance between its USNO and TASS positions. The Table below shows the mean of the differences within each catagory, and the standard deviation of the mean in parantheses. All differences are in the sense (USNO - TASS).
Field A (uncrowded)
N delta RA delta Dec delta Pos
-------------------------------------------------------
V < 10 1 -0.2 0.3 0.3
10 < V < 13 27 -0.9 (1.7) 0.1 (1.1) 1.9 (1.2)
13 < V 35 -0.8 (4.0) -0.1 (2.2) 3.6 (2.8)
Field B (crowded)
N delta RA delta Dec delta Pos
-------------------------------------------------------
V < 10 6 0.5 (0.9) -0.4 (0.6) 1.0 (0.7)
10 < V < 13 78 -0.1 (2.7) 0.3 (2.2) 2.5 (2.3)
13 < V 65 -0.0 (3.7) 0.2 (3.4) 4.0 (2.9)
The table shows
It might be interesting to examine the cases in which a single TASS detection fell within 15 arcsec of multiple stars in the USNO A1.0. I urge interested readers to try their hand at that project. This Technical Note provides the URLs at which you can grab lists of stars in selected areas; match them up and analyze them at home.