We use compare positions of brighter objects from SDSS and 2MASS (JHK catalog), and between SDSS and NOMAD (astrometric catalog from USNO). For a few locations on the sky, it appears that there are some differences in position beyond the nominal spec of less than 100 mas accuracy per coordinate. Our way to show that differences is to calculate in arcsecond what we call the "Delta". The Delta is the distance between the positions of the stars in SDSS vs 2MASS. Matches for objects separated by more than 0.6 arcsec have been excluded from the averages. Beyond this, no outlier rejection has been performed.
We can also calculate the average contribution in Delta of RA and DEC separately.
Having chosen some part of the sky in coordinates (RA, DEC), we can get in each picture some information, like the average of Delta in this area of the sky ( AVR(Delta) ), the standard deviation of Delta (SIGMA(Delta)), or the average of the contribution of right ascension and declination (which are AVR(DeltaRA) and AVR(DeltaDEC)).
Each area compared is about 1 square degree in size.
We have not done a comprehensive comparison of the whole of the SDSS footprint vs. the other catalogs.
We conclude that there are a few areas on the sky where either the SDSS or one of the comparison catalogs is off by more than the standard quoted error of 100 mas per coordinate.
Here is an exemple for (Ra,Dec) = (180,10)
SELECT o.objId, o.ra,
o.dec, o.r, o.type,
t.objId, t.ra, t.dec, o.g, o.r, o.i, t.j_m, t.h_m
FROM
SDSS:PhotoPrimary o, TWOMASS:PhotoPrimary t
WHERE XMATCH(o, t) < 0.1 AND
Region('CIRCLE J2000 180 10 29.5') AND
o.type = 6
We can use with this query the following awk code to obtain Delta :
awk -F, ' NR>1 {print $2,",", sqrt(($2-$7)*($2-$7)*cos($3*3.14159/180)*cos($3*3.14159/180)+($3-$8)*($3-$8))*3600.0;}' data.csv > data_delta.csv
To obtain only the contribution of Ra, we can use :
awk -F, ' NR>1 {print $2,",", sqrt(($2-$7)*($2-$7)*cos($3*3.14159/180)*cos($3*3.14159/180))*3600.0;}' data.csv > data_deltaRA.csv
And the following to get the contribution of Dec :
awk -F, ' NR>1 {print $2,",", sqrt(($3-$8)*($3-$8))*3600.0;}' data.csv > data_deltaDEC.csv
Firstly, we cut the stars with Delta > 0.6 (because they are outliers and are not linked with our astrometry problem) :
awk -F, '$2 < 0.6 {print;}' data_delta.csv > clean_delta.csv
Then, we can use Excel to calculate both average and standart deviation.
We highlight those areas of overlap between SDSS and 2MASS where seems that when AVR(Delta) > 0.23 in red in the table.
STRIPE | RA | DEC | AVR(Delta) | SIGMA(Delta) | AVR(DeltaRA) | AVR(DeltaDEC) | AVR(Delta)-Nomad | AVR(DeltaRA) | AVR(DeltaDEC) | |||
(deg) | (deg) | (arcsec) | (arcsec) | (arcsec) | (arcsec) | (arcsec) | (arcsec) | (arcsec) | ||||
1188 | 225.9559042 | 56.98013139 | 0.194 | 0.126 | 0.042 | -0.033 | ||||||
1188 | 264.7419681 | 64.43645654 | 0.151 | 0.126 | 0.005 | 0.025 | ||||||
1188 | 287.3997103 | 63.26341684 | 0.236 | 0.127 | -0.100 | 0.098 | 0.224 | 0.102 | -0.086 | |||
1188 | 306.2138751 | 58.97310519 | 0.204 | 0.116 | 0.014 | 0.118 | ||||||
1188 | 320.0438116 | 52.62969748 | 0.207 | 0.122 | 0.023 | 0.082 | ||||||
1188 | 330.0970633 | 45.10372779 | 0.160 | 0.126 | -0.001 | -0.003 | ||||||
1260 | 18.2428094 | 18.78702952 | 0.215 | 0.137 | 0.024 | -0.064 | ||||||
1260 | 19.55515186 | 28.71491245 | 0.287 | 0.150 | 0.129 | -0.165 | 0.299 | 0.207 | -0.145 | |||
1260 | 21.14557734 | 38.62744541 | 0.177 | 0.127 | -0.015 | 0.003 | ||||||
1260 | 23.25128998 | 48.51146966 | 0.225 | 0.142 | -0.128 | -0.091 | ||||||
1260 | 26.38611505 | 58.33829694 | 0.334 | 0.100 | -0.148 | 0.262 | 0.222 | 0.103 | -0.002 | |||
1260 | 31.94278132 | 68.02907252 | 0.263 | 0.115 | -0.106 | 0.171 | 0.181 | 0.005 | 0.023 | |||
1260 | 45.27525281 | 77.26698364 | 0.246 | 0.121 | -0.043 | 0.167 | 0.159 | 0.017 | -0.017 | |||
1260 | 96.19431975 | 83.61873228 | 0.239 | 0.133 | -0.061 | 0.106 | 0.210 | 0.007 | -0.094 | |||
1300 | 59.98168844 | 51.62104928 | 0.217 | 0.126 | -0.065 | 0.104 | ||||||
1300 | 72.54559713 | 58.619552 | 0.290 | 0.133 | -0.142 | 0.134 | 0.232 | 0.086 | -0.099 | |||
1300 | 90.38627883 | 63.8437677 | 0.328 | 0.102 | -0.224 | 0.178 | 0.137 | 0.024 | -0.005 | |||
1300 | 113.566800 | 66.09084647 | 0.311 | 0.116 | -0.104 | 0.213 | 0.167 | 0.038 | -0.023 | |||
1220 | 292.183761 | 78.09942726 | 0.211 | 0.140 | -0.062 | -0.021 | ||||||
1220 | 326.3612711 | 72.62666558 | 0.173 | 0.125 | 0.044 | -0.057 | ||||||
1220 | 341.8681607 | 64.3233318 | 0.142 | 0.120 | 0.027 | -0.014 | ||||||
1220 | 349.9007666 | 55.15929513 | 0.172 | 0.134 | -0.057 | -0.011 | ||||||
1220 | 354.8711739 | 45.66743628 | 0.152 | 0.118 | -0.042 | 0.000 | ||||||
1220 | 358.3799125 | 36.02327954 | 0.190 | 0.122 | 0.092 | -0.014 | ||||||
14 | 180 | 10 | 0.236 | 0.138 | -0.129 | -0.012 | 0.200 | 0.074 | -0.122 | |||
32 | 230 | 45 | 0.192 | 0.126 | -0.056 | -0.004 | ||||||
37 | 135 | 55 | 0.178 | 0.126 | 0.046 | -0.030 |