F 
Name  Schema Table  Database  Description  Type  Length  Unit  Default Value  Unified Content Descriptor 
fastGuiderMode 
[nspid]Multiframe 
WSA NonSurvey 
Fast guider mode {image primary HDU keyword: FGMODE} 
varchar 
32 

NONE 

feiiAperMag3 
[nspid]Source 
WSA NonSurvey 
Default point source Feii aperture corrected mag (2.0 arcsec aperture diameter) If in doubt use this flux estimator 
real 
4 
mag 
0.9999995e9 
PHOT_MAG 
feiiAperMag3Err 
[nspid]Source 
WSA NonSurvey 
Error in default point source Feii mag (2.0 arcsec aperture diameter) 
real 
4 
mag 
0.9999995e9 
ERROR 
feiiAperMag4 
[nspid]Source 
WSA NonSurvey 
Point source Feii aperture corrected mag (2.8 arcsec aperture diameter) 
real 
4 
mag 
0.9999995e9 
PHOT_MAG 
feiiAperMag4Err 
[nspid]Source 
WSA NonSurvey 
Error in point source Feii mag (2.8 arcsec aperture diameter) 
real 
4 
mag 
0.9999995e9 
ERROR 
feiiAperMag6 
[nspid]Source 
WSA NonSurvey 
Point source Feii aperture corrected mag (5.7 arcsec aperture diameter) 
real 
4 
mag 
0.9999995e9 
PHOT_MAG 
feiiAperMag6Err 
[nspid]Source 
WSA NonSurvey 
Error in point source Feii mag (5.7 arcsec aperture diameter) 
real 
4 
mag 
0.9999995e9 
ERROR 
feiiaStratAst 
[nspid]VarFrameSetInfo 
WSA NonSurvey 
Parameter, c0 from FerreiraLopes & Cross 2017, Eq. 18, in fit to astrometric rms vs magnitude in Feii band. 
real 
4 

0.9999995e9 
stat.fit.param 
The best fit solution to the expected RMS position around the mean for all objects in the frameset. Objects were binned in ranges of magnitude and the median RMS (after clipping out variable objects using the medianabsolute deviation) was calculated. The Strateva function $\zeta(m)>=a+b\,10^{0.4m}+c\,10^{0.8m}$ was fit, where $\zeta(m)$ is the expected RMS as a function of magnitude. The chisquared and number of degrees of freedom are also calculated. 
feiiaStratPht 
[nspid]VarFrameSetInfo 
WSA NonSurvey 
Parameter, c0 from FerreiraLopes & Cross 2017, Eq. 18, in fit to photometric rms vs magnitude in Feii band. 
real 
4 

0.9999995e9 
stat.fit.param 
The best fit solution to the expected RMS brightness (in magnitudes) for all objects in the frameset. Objects were binned in ranges of magnitude and the median RMS (after clipping out variable objects using the medianabsolute deviation) was calculated. The Strateva function $\zeta(m)>=a+b\,10^{0.4m}+c\,10^{0.8m}$ was fit, where $\zeta(m)$ is the expected RMS as a function of magnitude. The chisquared and number of degrees of freedom are also calculated. This technique was used in Sesar et al. 2007, AJ, 134, 2236. 
feiibestAper 
[nspid]Variability 
WSA NonSurvey 
Best aperture (16) for photometric statistics in the Feii band 
int 
4 

9999 

Aperture magnitude (16) which gives the lowest RMS for the object. All apertures have the appropriate aperture correction. This can give better values in crowded regions than aperMag3 (see Irwin et al. 2007, MNRAS, 375, 1449) 
feiibStratAst 
[nspid]VarFrameSetInfo 
WSA NonSurvey 
Parameter, c1 from FerreiraLopes & Cross 2017, Eq. 18, in fit to astrometric rms vs magnitude in Feii band. 
real 
4 

0.9999995e9 
stat.fit.param 
The best fit solution to the expected RMS position around the mean for all objects in the frameset. Objects were binned in ranges of magnitude and the median RMS (after clipping out variable objects using the medianabsolute deviation) was calculated. The Strateva function $\zeta(m)>=a+b\,10^{0.4m}+c\,10^{0.8m}$ was fit, where $\zeta(m)$ is the expected RMS as a function of magnitude. The chisquared and number of degrees of freedom are also calculated. 
feiibStratPht 
[nspid]VarFrameSetInfo 
WSA NonSurvey 
Parameter, c1 from FerreiraLopes & Cross 2017, Eq. 18, in fit to photometric rms vs magnitude in Feii band. 
real 
4 

0.9999995e9 
stat.fit.param 
The best fit solution to the expected RMS brightness (in magnitudes) for all objects in the frameset. Objects were binned in ranges of magnitude and the median RMS (after clipping out variable objects using the medianabsolute deviation) was calculated. The Strateva function $\zeta(m)>=a+b\,10^{0.4m}+c\,10^{0.8m}$ was fit, where $\zeta(m)$ is the expected RMS as a function of magnitude. The chisquared and number of degrees of freedom are also calculated. This technique was used in Sesar et al. 2007, AJ, 134, 2236. 
feiichiSqAst 
[nspid]VarFrameSetInfo 
WSA NonSurvey 
Goodness of fit of Strateva function to astrometric data in Feii band 
real 
4 

0.9999995e9 
stat.fit.goodness 
The best fit solution to the expected RMS position around the mean for all objects in the frameset. Objects were binned in ranges of magnitude and the median RMS (after clipping out variable objects using the medianabsolute deviation) was calculated. The Strateva function $\zeta(m)>=a+b\,10^{0.4m}+c\,10^{0.8m}$ was fit, where $\zeta(m)$ is the expected RMS as a function of magnitude. The chisquared and number of degrees of freedom are also calculated. 
feiichiSqpd 
[nspid]Variability 
WSA NonSurvey 
Chi square (per degree of freedom) fit to data (mean and expected rms) 
real 
4 

0.9999995e9 

The photometry is calculated for good observations in the best aperture. The mean, rms, median, median absolute deviation, minMag and maxMag are quite standard. The skewness is calculated as in Sesar et al. 2007, AJ, 134, 2236. The number of good detections that are more than 3 standard deviations can indicate a distribution with many outliers. In each frameset, the mean and rms are used to derive a fit to the expected rms as a function of magnitude. The parameters for the fit are stored in VarFrameSetInfo and the value for the source is in expRms. This is subtracted from the rms in quadrature to get the intrinsic rms: the variability of the object beyond the noise in the system. The chisquared is calculated, assuming a nonvariable object which has the noise from the expectedrms and mean calculated as above. The probVar statistic assumes a chisquared distribution with the correct number of degrees of freedom. The varClass statistic is 1, if the probVar>0.9 and intrinsicRMS/expectedRMS>3. 
feiichiSqPht 
[nspid]VarFrameSetInfo 
WSA NonSurvey 
Goodness of fit of Strateva function to photometric data in Feii band 
real 
4 

0.9999995e9 

The best fit solution to the expected RMS brightness (in magnitudes) for all objects in the frameset. Objects were binned in ranges of magnitude and the median RMS (after clipping out variable objects using the medianabsolute deviation) was calculated. The Strateva function $\zeta(m)>=a+b\,10^{0.4m}+c\,10^{0.8m}$ was fit, where $\zeta(m)$ is the expected RMS as a function of magnitude. The chisquared and number of degrees of freedom are also calculated. This technique was used in Sesar et al. 2007, AJ, 134, 2236. 
feiiClass 
[nspid]Source 
WSA NonSurvey 
discrete image classification flag in Feii 
smallint 
2 

9999 
CLASS_MISC 
feiiClassStat 
[nspid]Source 
WSA NonSurvey 
N(0,1) stellarnessofprofile statistic in Feii 
real 
4 

0.9999995e9 
STAT_PROP 
feiicStratAst 
[nspid]VarFrameSetInfo 
WSA NonSurvey 
Parameter, c2 from FerreiraLopes & Cross 2017, Eq. 18, in fit to astrometric rms vs magnitude in Feii band. 
real 
4 

0.9999995e9 
stat.fit.param 
The best fit solution to the expected RMS position around the mean for all objects in the frameset. Objects were binned in ranges of magnitude and the median RMS (after clipping out variable objects using the medianabsolute deviation) was calculated. The Strateva function $\zeta(m)>=a+b\,10^{0.4m}+c\,10^{0.8m}$ was fit, where $\zeta(m)$ is the expected RMS as a function of magnitude. The chisquared and number of degrees of freedom are also calculated. 
feiicStratPht 
[nspid]VarFrameSetInfo 
WSA NonSurvey 
Parameter, c2 from FerreiraLopes & Cross 2017, Eq. 18, in fit to photometric rms vs magnitude in Feii band. 
real 
4 

0.9999995e9 
stat.fit.param 
The best fit solution to the expected RMS brightness (in magnitudes) for all objects in the frameset. Objects were binned in ranges of magnitude and the median RMS (after clipping out variable objects using the medianabsolute deviation) was calculated. The Strateva function $\zeta(m)>=a+b\,10^{0.4m}+c\,10^{0.8m}$ was fit, where $\zeta(m)$ is the expected RMS as a function of magnitude. The chisquared and number of degrees of freedom are also calculated. This technique was used in Sesar et al. 2007, AJ, 134, 2236. 
feiiDeblend 
[nspid]Source 
WSA NonSurvey 
placeholder flag indicating parent/child relation in Feii 
int 
4 

99999999 
CODE_MISC 
This CASU pipeline processing source extraction flag is a placeholder only, and is always set to zero in all passbands in the merged source lists. If you need to know when a particular image detection is a component of a deblend or not, test bit 4 of attribute ppErrBits (see corresponding glossary entry) which is set by WFAU's postprocessing software based on testing the areal profiles aprof28 (these are set by CASU to 1 for deblended components, or positive values for nondeblended detections). We encode this in an information bit of ppErrBits for convenience when querying the merged source tables. 
feiidStratAst 
[nspid]VarFrameSetInfo 
WSA NonSurvey 
Parameter, c3 from FerreiraLopes & Cross 2017, Eq. 18, in fit to astrometric rms vs magnitude in Feii band. 
real 
4 

0.9999995e9 
stat.fit.param 
The best fit solution to the expected RMS position around the mean for all objects in the frameset. Objects were binned in ranges of magnitude and the median RMS (after clipping out variable objects using the medianabsolute deviation) was calculated. The Strateva function $\zeta(m)>=a+b\,10^{0.4m}+c\,10^{0.8m}$ was fit, where $\zeta(m)$ is the expected RMS as a function of magnitude. The chisquared and number of degrees of freedom are also calculated. 
feiidStratPht 
[nspid]VarFrameSetInfo 
WSA NonSurvey 
Parameter, c0 from FerreiraLopes & Cross 2017, Eq. 18, in fit to photometric rms vs magnitude in Feii band. 
real 
4 

0.9999995e9 
stat.fit.param 
The best fit solution to the expected RMS brightness (in magnitudes) for all objects in the frameset. Objects were binned in ranges of magnitude and the median RMS (after clipping out variable objects using the medianabsolute deviation) was calculated. The Strateva function $\zeta(m)>=a+b\,10^{0.4m}+c\,10^{0.8m}$ was fit, where $\zeta(m)$ is the expected RMS as a function of magnitude. The chisquared and number of degrees of freedom are also calculated. This technique was used in Sesar et al. 2007, AJ, 134, 2236. 
feiiEll 
[nspid]Source 
WSA NonSurvey 
1b/a, where a/b=semimajor/minor axes in Feii 
real 
4 

0.9999995e9 
PHYS_ELLIPTICITY 
feiieNum 
[nspid]MergeLog 
WSA NonSurvey 
the extension number of this Feii frame 
tinyint 
1 


NUMBER 
feiiErrBits 
[nspid]Source 
WSA NonSurvey 
processing warning/error bitwise flags in Feii 
int 
4 

99999999 
CODE_MISC 
Apparently not actually an error bit flag, but a count of the number of zero confidence pixels in the default (2 arcsec diameter) aperture. 
feiiEta 
[nspid]Source 
WSA NonSurvey 
Offset of Feii detection from master position (+north/south) 
real 
4 
arcsec 
0.9999995e9 
POS_EQ_DEC_OFF 
When associating individual passband detections into merged sources, a generous (in terms of the positional uncertainties) pairing radius of 2.0 (UKIDSS LAS and GPS; UHS; also nonsurvey programmes) or 1.0 (UKIDSS GPS, DXS and UDS) arcseconds is used, the higher value enabling pairing of moving sources when epoch separations may be several years. Such a large association criterion can of course lead to spurious pairings in the merged sources lists (although note that between passband pairs, handshake pairing is done: both passbands must agree that the candidate pair is their nearest neighbour for the pair to propagate through into the merged source table). In order to help filter spurious pairings out, and assuming that large positional offsets between the different passband detections are not expected (e.g. because of source motion, or larger than usual positional uncertainties) then the attributes Xi and Eta can be used to filter any pairings with suspiciously large offsets in one or more bands. For example, for a clean sample of QSOs from the LAS, you might wish to insist that the offsets in the selected sample are all below 1 arcsecond: simply add WHERE clauses into the SQL sample selection script to exclude all Xi and Eta values larger than the threshold you want. NB: the master position is the position of the detection in the shortest passband in the set, rather than the ra/dec of the source as stored in source attributes of the same name. The former is used in the pairing process, while the latter is generally the optimally weighted mean position from an astrometric solution or other combinatorial process of all individual detection positions across the available passbands. 
feiiexpML 
[nspid]VarFrameSetInfo 
WSA NonSurvey 
Expected magnitude limit of frameSet in this in Feii band. 
real 
4 

0.9999995e9 

The expected magnitude limit of an intermediate stack, based on the total exposure time. expML=Filter.oneSecML+1.25*log10(totalExpTime). Since different intermediate stacks can have different exposure times, the totalExpTime is the minimum, as long as the number of stacks with this minimum make up 10% of the total. This is a more conservative treatment than just taking the mean or median total exposure time. 
feiiExpRms 
[nspid]Variability 
WSA NonSurvey 
Rms calculated from polynomial fit to modal RMS as a function of magnitude in Feii band 
real 
4 
mag 
0.9999995e9 

The photometry is calculated for good observations in the best aperture. The mean, rms, median, median absolute deviation, minMag and maxMag are quite standard. The skewness is calculated as in Sesar et al. 2007, AJ, 134, 2236. The number of good detections that are more than 3 standard deviations can indicate a distribution with many outliers. In each frameset, the mean and rms are used to derive a fit to the expected rms as a function of magnitude. The parameters for the fit are stored in VarFrameSetInfo and the value for the source is in expRms. This is subtracted from the rms in quadrature to get the intrinsic rms: the variability of the object beyond the noise in the system. The chisquared is calculated, assuming a nonvariable object which has the noise from the expectedrms and mean calculated as above. The probVar statistic assumes a chisquared distribution with the correct number of degrees of freedom. The varClass statistic is 1, if the probVar>0.9 and intrinsicRMS/expectedRMS>3. 
feiiGausig 
[nspid]Source 
WSA NonSurvey 
RMS of axes of ellipse fit in Feii 
real 
4 
pixels 
0.9999995e9 
MORPH_PARAM 
feiiHallMag 
[nspid]Source 
WSA NonSurvey 
Total point source Feii mag 
real 
4 
mag 
0.9999995e9 
PHOT_MAG 
feiiHallMagErr 
[nspid]Source 
WSA NonSurvey 
Error in total point source Feii mag 
real 
4 
mag 
0.9999995e9 
ERROR 
feiiIntRms 
[nspid]Variability 
WSA NonSurvey 
Intrinsic rms in Feiiband 
real 
4 
mag 
0.9999995e9 

The photometry is calculated for good observations in the best aperture. The mean, rms, median, median absolute deviation, minMag and maxMag are quite standard. The skewness is calculated as in Sesar et al. 2007, AJ, 134, 2236. The number of good detections that are more than 3 standard deviations can indicate a distribution with many outliers. In each frameset, the mean and rms are used to derive a fit to the expected rms as a function of magnitude. The parameters for the fit are stored in VarFrameSetInfo and the value for the source is in expRms. This is subtracted from the rms in quadrature to get the intrinsic rms: the variability of the object beyond the noise in the system. The chisquared is calculated, assuming a nonvariable object which has the noise from the expectedrms and mean calculated as above. The probVar statistic assumes a chisquared distribution with the correct number of degrees of freedom. The varClass statistic is 1, if the probVar>0.9 and intrinsicRMS/expectedRMS>3. 
feiiisDefAst 
[nspid]VarFrameSetInfo 
WSA NonSurvey 
Use a default model for the astrometric noise in Feii band. 
tinyint 
1 

0 
meta.code 
feiiisDefPht 
[nspid]VarFrameSetInfo 
WSA NonSurvey 
Use a default model for the photometric noise in Feii band. 
tinyint 
1 

0 

feiiMagMAD 
[nspid]Variability 
WSA NonSurvey 
Median Absolute Deviation of Feii magnitude 
real 
4 
mag 
0.9999995e9 

The photometry is calculated for good observations in the best aperture. The mean, rms, median, median absolute deviation, minMag and maxMag are quite standard. The skewness is calculated as in Sesar et al. 2007, AJ, 134, 2236. The number of good detections that are more than 3 standard deviations can indicate a distribution with many outliers. In each frameset, the mean and rms are used to derive a fit to the expected rms as a function of magnitude. The parameters for the fit are stored in VarFrameSetInfo and the value for the source is in expRms. This is subtracted from the rms in quadrature to get the intrinsic rms: the variability of the object beyond the noise in the system. The chisquared is calculated, assuming a nonvariable object which has the noise from the expectedrms and mean calculated as above. The probVar statistic assumes a chisquared distribution with the correct number of degrees of freedom. The varClass statistic is 1, if the probVar>0.9 and intrinsicRMS/expectedRMS>3. 
feiiMagRms 
[nspid]Variability 
WSA NonSurvey 
rms of Feii magnitude 
real 
4 
mag 
0.9999995e9 

The photometry is calculated for good observations in the best aperture. The mean, rms, median, median absolute deviation, minMag and maxMag are quite standard. The skewness is calculated as in Sesar et al. 2007, AJ, 134, 2236. The number of good detections that are more than 3 standard deviations can indicate a distribution with many outliers. In each frameset, the mean and rms are used to derive a fit to the expected rms as a function of magnitude. The parameters for the fit are stored in VarFrameSetInfo and the value for the source is in expRms. This is subtracted from the rms in quadrature to get the intrinsic rms: the variability of the object beyond the noise in the system. The chisquared is calculated, assuming a nonvariable object which has the noise from the expectedrms and mean calculated as above. The probVar statistic assumes a chisquared distribution with the correct number of degrees of freedom. The varClass statistic is 1, if the probVar>0.9 and intrinsicRMS/expectedRMS>3. 
feiimaxCadence 
[nspid]Variability 
WSA NonSurvey 
maximum gap between observations 
real 
4 
days 
0.9999995e9 

The observations are classified as good, flagged or missing. Flagged observations are ones where the object has a ppErrBit flag. Missing observations are observations of the part of the sky that include the position of the object, but had no detection. All the statistics are calculated from good observations. The cadence parameters give the minimum, median and maximum time between observations, which is useful to know if the data could be used to find a particular type of variable. 
feiiMaxMag 
[nspid]Variability 
WSA NonSurvey 
Maximum magnitude in Feii band, of good detections 
real 
4 

0.9999995e9 

The photometry is calculated for good observations in the best aperture. The mean, rms, median, median absolute deviation, minMag and maxMag are quite standard. The skewness is calculated as in Sesar et al. 2007, AJ, 134, 2236. The number of good detections that are more than 3 standard deviations can indicate a distribution with many outliers. In each frameset, the mean and rms are used to derive a fit to the expected rms as a function of magnitude. The parameters for the fit are stored in VarFrameSetInfo and the value for the source is in expRms. This is subtracted from the rms in quadrature to get the intrinsic rms: the variability of the object beyond the noise in the system. The chisquared is calculated, assuming a nonvariable object which has the noise from the expectedrms and mean calculated as above. The probVar statistic assumes a chisquared distribution with the correct number of degrees of freedom. The varClass statistic is 1, if the probVar>0.9 and intrinsicRMS/expectedRMS>3. 
feiimeanMag 
[nspid]Variability 
WSA NonSurvey 
Mean Feii magnitude 
real 
4 
mag 
0.9999995e9 

The photometry is calculated for good observations in the best aperture. The mean, rms, median, median absolute deviation, minMag and maxMag are quite standard. The skewness is calculated as in Sesar et al. 2007, AJ, 134, 2236. The number of good detections that are more than 3 standard deviations can indicate a distribution with many outliers. In each frameset, the mean and rms are used to derive a fit to the expected rms as a function of magnitude. The parameters for the fit are stored in VarFrameSetInfo and the value for the source is in expRms. This is subtracted from the rms in quadrature to get the intrinsic rms: the variability of the object beyond the noise in the system. The chisquared is calculated, assuming a nonvariable object which has the noise from the expectedrms and mean calculated as above. The probVar statistic assumes a chisquared distribution with the correct number of degrees of freedom. The varClass statistic is 1, if the probVar>0.9 and intrinsicRMS/expectedRMS>3. 
feiimedCadence 
[nspid]Variability 
WSA NonSurvey 
median gap between observations 
real 
4 
days 
0.9999995e9 

The observations are classified as good, flagged or missing. Flagged observations are ones where the object has a ppErrBit flag. Missing observations are observations of the part of the sky that include the position of the object, but had no detection. All the statistics are calculated from good observations. The cadence parameters give the minimum, median and maximum time between observations, which is useful to know if the data could be used to find a particular type of variable. 
feiimedianMag 
[nspid]Variability 
WSA NonSurvey 
Median Feii magnitude 
real 
4 
mag 
0.9999995e9 

The photometry is calculated for good observations in the best aperture. The mean, rms, median, median absolute deviation, minMag and maxMag are quite standard. The skewness is calculated as in Sesar et al. 2007, AJ, 134, 2236. The number of good detections that are more than 3 standard deviations can indicate a distribution with many outliers. In each frameset, the mean and rms are used to derive a fit to the expected rms as a function of magnitude. The parameters for the fit are stored in VarFrameSetInfo and the value for the source is in expRms. This is subtracted from the rms in quadrature to get the intrinsic rms: the variability of the object beyond the noise in the system. The chisquared is calculated, assuming a nonvariable object which has the noise from the expectedrms and mean calculated as above. The probVar statistic assumes a chisquared distribution with the correct number of degrees of freedom. The varClass statistic is 1, if the probVar>0.9 and intrinsicRMS/expectedRMS>3. 
feiimfID 
[nspid]MergeLog 
WSA NonSurvey 
the UID of the relevant Feii multiframe 
bigint 
8 


ID_FRAME 
feiiminCadence 
[nspid]Variability 
WSA NonSurvey 
minimum gap between observations 
real 
4 
days 
0.9999995e9 

The observations are classified as good, flagged or missing. Flagged observations are ones where the object has a ppErrBit flag. Missing observations are observations of the part of the sky that include the position of the object, but had no detection. All the statistics are calculated from good observations. The cadence parameters give the minimum, median and maximum time between observations, which is useful to know if the data could be used to find a particular type of variable. 
feiiMinMag 
[nspid]Variability 
WSA NonSurvey 
Minimum magnitude in Feii band, of good detections 
real 
4 

0.9999995e9 

The photometry is calculated for good observations in the best aperture. The mean, rms, median, median absolute deviation, minMag and maxMag are quite standard. The skewness is calculated as in Sesar et al. 2007, AJ, 134, 2236. The number of good detections that are more than 3 standard deviations can indicate a distribution with many outliers. In each frameset, the mean and rms are used to derive a fit to the expected rms as a function of magnitude. The parameters for the fit are stored in VarFrameSetInfo and the value for the source is in expRms. This is subtracted from the rms in quadrature to get the intrinsic rms: the variability of the object beyond the noise in the system. The chisquared is calculated, assuming a nonvariable object which has the noise from the expectedrms and mean calculated as above. The probVar statistic assumes a chisquared distribution with the correct number of degrees of freedom. The varClass statistic is 1, if the probVar>0.9 and intrinsicRMS/expectedRMS>3. 
feiindof 
[nspid]Variability 
WSA NonSurvey 
Number of degrees of freedom for chisquare 
smallint 
2 

9999 

The photometry is calculated for good observations in the best aperture. The mean, rms, median, median absolute deviation, minMag and maxMag are quite standard. The skewness is calculated as in Sesar et al. 2007, AJ, 134, 2236. The number of good detections that are more than 3 standard deviations can indicate a distribution with many outliers. In each frameset, the mean and rms are used to derive a fit to the expected rms as a function of magnitude. The parameters for the fit are stored in VarFrameSetInfo and the value for the source is in expRms. This is subtracted from the rms in quadrature to get the intrinsic rms: the variability of the object beyond the noise in the system. The chisquared is calculated, assuming a nonvariable object which has the noise from the expectedrms and mean calculated as above. The probVar statistic assumes a chisquared distribution with the correct number of degrees of freedom. The varClass statistic is 1, if the probVar>0.9 and intrinsicRMS/expectedRMS>3. 
feiinDofAst 
[nspid]VarFrameSetInfo 
WSA NonSurvey 
Number of degrees of freedom of astrometric fit in Feii band. 
smallint 
2 

9999 
stat.fit.dof;stat.param 
The best fit solution to the expected RMS position around the mean for all objects in the frameset. Objects were binned in ranges of magnitude and the median RMS (after clipping out variable objects using the medianabsolute deviation) was calculated. The Strateva function $\zeta(m)>=a+b\,10^{0.4m}+c\,10^{0.8m}$ was fit, where $\zeta(m)$ is the expected RMS as a function of magnitude. The chisquared and number of degrees of freedom are also calculated. 
feiinDofPht 
[nspid]VarFrameSetInfo 
WSA NonSurvey 
Number of degrees of freedom of photometric fit in Feii band. 
smallint 
2 

9999 

The best fit solution to the expected RMS brightness (in magnitudes) for all objects in the frameset. Objects were binned in ranges of magnitude and the median RMS (after clipping out variable objects using the medianabsolute deviation) was calculated. The Strateva function $\zeta(m)>=a+b\,10^{0.4m}+c\,10^{0.8m}$ was fit, where $\zeta(m)$ is the expected RMS as a function of magnitude. The chisquared and number of degrees of freedom are also calculated. This technique was used in Sesar et al. 2007, AJ, 134, 2236. 
feiinFlaggedObs 
[nspid]Variability 
WSA NonSurvey 
Number of detections in Feii band flagged as potentially spurious by u12bkasi1Detection.ppErrBits 
int 
4 

0 

The observations are classified as good, flagged or missing. Flagged observations are ones where the object has a ppErrBit flag. Missing observations are observations of the part of the sky that include the position of the object, but had no detection. All the statistics are calculated from good observations. The cadence parameters give the minimum, median and maximum time between observations, which is useful to know if the data could be used to find a particular type of variable. 
feiinFlaggedObs 
[nspid]Variability 
WSA NonSurvey 
Number of detections in Feii band flagged as potentially spurious by u13akasi1Detection.ppErrBits 
int 
4 

0 

The observations are classified as good, flagged or missing. Flagged observations are ones where the object has a ppErrBit flag. Missing observations are observations of the part of the sky that include the position of the object, but had no detection. All the statistics are calculated from good observations. The cadence parameters give the minimum, median and maximum time between observations, which is useful to know if the data could be used to find a particular type of variable. 
feiinFlaggedObs 
[nspid]Variability 
WSA NonSurvey 
Number of detections in Feii band flagged as potentially spurious by u13akasi2Detection.ppErrBits 
int 
4 

0 

The observations are classified as good, flagged or missing. Flagged observations are ones where the object has a ppErrBit flag. Missing observations are observations of the part of the sky that include the position of the object, but had no detection. All the statistics are calculated from good observations. The cadence parameters give the minimum, median and maximum time between observations, which is useful to know if the data could be used to find a particular type of variable. 
feiinFlaggedObs 
[nspid]Variability 
WSA NonSurvey 
Number of detections in Feii band flagged as potentially spurious by u13akasi3Detection.ppErrBits 
int 
4 

0 

The observations are classified as good, flagged or missing. Flagged observations are ones where the object has a ppErrBit flag. Missing observations are observations of the part of the sky that include the position of the object, but had no detection. All the statistics are calculated from good observations. The cadence parameters give the minimum, median and maximum time between observations, which is useful to know if the data could be used to find a particular type of variable. 
feiinGoodObs 
[nspid]Variability 
WSA NonSurvey 
Number of good detections in Feii band 
int 
4 

0 

The observations are classified as good, flagged or missing. Flagged observations are ones where the object has a ppErrBit flag. Missing observations are observations of the part of the sky that include the position of the object, but had no detection. All the statistics are calculated from good observations. The cadence parameters give the minimum, median and maximum time between observations, which is useful to know if the data could be used to find a particular type of variable. 
feiiNgt3sig 
[nspid]Variability 
WSA NonSurvey 
Number of good detections in Feiiband that are more than 3 sigma deviations (feiiAperMagN < (feiiMeanMag3*feiiMagRms) 
smallint 
2 

9999 

The photometry is calculated for good observations in the best aperture. The mean, rms, median, median absolute deviation, minMag and maxMag are quite standard. The skewness is calculated as in Sesar et al. 2007, AJ, 134, 2236. The number of good detections that are more than 3 standard deviations can indicate a distribution with many outliers. In each frameset, the mean and rms are used to derive a fit to the expected rms as a function of magnitude. The parameters for the fit are stored in VarFrameSetInfo and the value for the source is in expRms. This is subtracted from the rms in quadrature to get the intrinsic rms: the variability of the object beyond the noise in the system. The chisquared is calculated, assuming a nonvariable object which has the noise from the expectedrms and mean calculated as above. The probVar statistic assumes a chisquared distribution with the correct number of degrees of freedom. The varClass statistic is 1, if the probVar>0.9 and intrinsicRMS/expectedRMS>3. 
feiinMissingObs 
[nspid]Variability 
WSA NonSurvey 
Number of Feii band frames that this object should have been detected on and was not 
int 
4 

0 

The observations are classified as good, flagged or missing. Flagged observations are ones where the object has a ppErrBit flag. Missing observations are observations of the part of the sky that include the position of the object, but had no detection. All the statistics are calculated from good observations. The cadence parameters give the minimum, median and maximum time between observations, which is useful to know if the data could be used to find a particular type of variable. 
feiiPA 
[nspid]Source 
WSA NonSurvey 
ellipse fit celestial orientation in Feii 
real 
4 
Degrees 
0.9999995e9 
POS_POSANG 
feiiPetroMag 
[nspid]Source 
WSA NonSurvey 
Extended source Feii mag (Petrosian) 
real 
4 
mag 
0.9999995e9 
PHOT_MAG 
feiiPetroMagErr 
[nspid]Source 
WSA NonSurvey 
Error in extended source Feii mag (Petrosian) 
real 
4 
mag 
0.9999995e9 
ERROR 
feiippErrBits 
[nspid]Source 
WSA NonSurvey 
additional WFAU postprocessing error bits in Feii 
int 
4 

0 
CODE_MISC 
Postprocessing error quality bit flags assigned (NB: from UKIDSS DR2 release onwards) in the WSA curation procedure for survey data. From least to most significant byte in the 4byte integer attribute byte 0 (bits 0 to 7) corresponds to information on generally innocuous conditions that are nonetheless potentially significant as regards the integrity of that detection; byte 1 (bits 8 to 15) corresponds to warnings; byte 2 (bits 16 to 23) corresponds to important warnings; and finally byte 3 (bits 24 to 31) corresponds to severe warnings: Byte  Bit  Detection quality issue  Threshold or bit mask  Applies to     Decimal  Hexadecimal   0  4  Deblended  16  0x00000010  All VDFS catalogues  0  6  Bad pixel(s) in default aperture  64  0x00000040  All VDFS catalogues  1  15  Source in poor flat field region  32768  0x00008000  All but mosaics  2  16  Close to saturated  65536  0x00010000  All VDFS catalogues (though deeps excluded prior to DR8)  2  17  Photometric calibration probably subject to systematic error  131072  0x00020000  GPS only  2  19  Possible crosstalk artefact/contamination  524288  0x00080000  All but GPS  2  22  Lies within a dither offset of the stacked frame boundary  4194304  0x00400000  All but mosaics  In this way, the higher the error quality bit flag value, the more likely it is that the detection is spurious. The decimal threshold (column 4) gives the minimum value of the quality flag for a detection having the given condition (since other bits in the flag may be set also; the corresponding hexadecimal value, where each digit corresponds to 4 bits in the flag, can be easier to compute when writing SQL queries to test for a given condition). For example, to exclude all K band sources in the LAS having any error quality condition other than informational ones, include a predicate ... AND kppErrBits ≤ 255. See the SQL Cookbook and other online pages for further information. 
feiiprobVar 
[nspid]Variability 
WSA NonSurvey 
Probability of variable from chisquare (and other data) 
real 
4 

0.9999995e9 

The photometry is calculated for good observations in the best aperture. The mean, rms, median, median absolute deviation, minMag and maxMag are quite standard. The skewness is calculated as in Sesar et al. 2007, AJ, 134, 2236. The number of good detections that are more than 3 standard deviations can indicate a distribution with many outliers. In each frameset, the mean and rms are used to derive a fit to the expected rms as a function of magnitude. The parameters for the fit are stored in VarFrameSetInfo and the value for the source is in expRms. This is subtracted from the rms in quadrature to get the intrinsic rms: the variability of the object beyond the noise in the system. The chisquared is calculated, assuming a nonvariable object which has the noise from the expectedrms and mean calculated as above. The probVar statistic assumes a chisquared distribution with the correct number of degrees of freedom. The varClass statistic is 1, if the probVar>0.9 and intrinsicRMS/expectedRMS>3. 
feiiPsfMag 
[nspid]Source 
WSA NonSurvey 
Point source profilefitted Feii mag 
real 
4 
mag 
0.9999995e9 
PHOT_MAG 
feiiPsfMagErr 
[nspid]Source 
WSA NonSurvey 
Error in point source profilefitted Feii mag 
real 
4 
mag 
0.9999995e9 
ERROR 
feiiSeqNum 
[nspid]Source 
WSA NonSurvey 
the running number of the Feii detection 
int 
4 

99999999 
ID_NUMBER 
feiiSerMag2D 
[nspid]Source 
WSA NonSurvey 
Extended source Feii mag (profilefitted) 
real 
4 
mag 
0.9999995e9 
PHOT_MAG 
feiiSerMag2DErr 
[nspid]Source 
WSA NonSurvey 
Error in extended source Feii mag (profilefitted) 
real 
4 
mag 
0.9999995e9 
ERROR 
feiiskewness 
[nspid]Variability 
WSA NonSurvey 
Skewness in Feii band (see Sesar et al. 2007) 
real 
4 

0.9999995e9 

The photometry is calculated for good observations in the best aperture. The mean, rms, median, median absolute deviation, minMag and maxMag are quite standard. The skewness is calculated as in Sesar et al. 2007, AJ, 134, 2236. The number of good detections that are more than 3 standard deviations can indicate a distribution with many outliers. In each frameset, the mean and rms are used to derive a fit to the expected rms as a function of magnitude. The parameters for the fit are stored in VarFrameSetInfo and the value for the source is in expRms. This is subtracted from the rms in quadrature to get the intrinsic rms: the variability of the object beyond the noise in the system. The chisquared is calculated, assuming a nonvariable object which has the noise from the expectedrms and mean calculated as above. The probVar statistic assumes a chisquared distribution with the correct number of degrees of freedom. The varClass statistic is 1, if the probVar>0.9 and intrinsicRMS/expectedRMS>3. 
feiitotalPeriod 
[nspid]Variability 
WSA NonSurvey 
total period of observations (last obsfirst obs) 
real 
4 
days 
0.9999995e9 

The observations are classified as good, flagged or missing. Flagged observations are ones where the object has a ppErrBit flag. Missing observations are observations of the part of the sky that include the position of the object, but had no detection. All the statistics are calculated from good observations. The cadence parameters give the minimum, median and maximum time between observations, which is useful to know if the data could be used to find a particular type of variable. 
feiiVarClass 
[nspid]Variability 
WSA NonSurvey 
Classification of variability in this band 
smallint 
2 

9999 

The photometry is calculated for good observations in the best aperture. The mean, rms, median, median absolute deviation, minMag and maxMag are quite standard. The skewness is calculated as in Sesar et al. 2007, AJ, 134, 2236. The number of good detections that are more than 3 standard deviations can indicate a distribution with many outliers. In each frameset, the mean and rms are used to derive a fit to the expected rms as a function of magnitude. The parameters for the fit are stored in VarFrameSetInfo and the value for the source is in expRms. This is subtracted from the rms in quadrature to get the intrinsic rms: the variability of the object beyond the noise in the system. The chisquared is calculated, assuming a nonvariable object which has the noise from the expectedrms and mean calculated as above. The probVar statistic assumes a chisquared distribution with the correct number of degrees of freedom. The varClass statistic is 1, if the probVar>0.9 and intrinsicRMS/expectedRMS>3. 
feiiXi 
[nspid]Source 
WSA NonSurvey 
Offset of Feii detection from master position (+east/west) 
real 
4 
arcsec 
0.9999995e9 
POS_EQ_RA_OFF 
When associating individual passband detections into merged sources, a generous (in terms of the positional uncertainties) pairing radius of 2.0 (UKIDSS LAS and GPS; UHS; also nonsurvey programmes) or 1.0 (UKIDSS GPS, DXS and UDS) arcseconds is used, the higher value enabling pairing of moving sources when epoch separations may be several years. Such a large association criterion can of course lead to spurious pairings in the merged sources lists (although note that between passband pairs, handshake pairing is done: both passbands must agree that the candidate pair is their nearest neighbour for the pair to propagate through into the merged source table). In order to help filter spurious pairings out, and assuming that large positional offsets between the different passband detections are not expected (e.g. because of source motion, or larger than usual positional uncertainties) then the attributes Xi and Eta can be used to filter any pairings with suspiciously large offsets in one or more bands. For example, for a clean sample of QSOs from the LAS, you might wish to insist that the offsets in the selected sample are all below 1 arcsecond: simply add WHERE clauses into the SQL sample selection script to exclude all Xi and Eta values larger than the threshold you want. NB: the master position is the position of the detection in the shortest passband in the set, rather than the ra/dec of the source as stored in source attributes of the same name. The former is used in the pairing process, while the latter is generally the optimally weighted mean position from an astrometric solution or other combinatorial process of all individual detection positions across the available passbands. 
fieldID 
[nspid]ProgrammeFrame 
WSA NonSurvey 
UID of position on sky, set just after ProgrammeBuilder runs 
int 
4 

99999999 
meta.bib 
fieldID 
[nspid]RequiredMosaic, [nspid]RequiredStack 
WSA NonSurvey 
UID of position on sky. 
int 
4 

99999999 
?? 
fieldID 
[nspid]StdFieldInfo 
WSA NonSurvey 
The fieldID is a UID identifying each of the 43 standard fields that are observed as part of the calibration observations 
int 
4 


ID_FIELD 
fieldName 
[nspid]StdFieldInfo 
WSA NonSurvey 
reference name of field 
varchar 
16 

NONE 
???? 
fileName 
[nspid]Multiframe 
WSA NonSurvey 
the filename for the multiframe, eg. server:/path/filename.fit 
varchar 
256 


meta.id;meta.file 
fileTimeStamp 
[nspid]Multiframe 
WSA NonSurvey 
Time stamp digits (from the original CASU directory name and file time stamp) for enforcing uniqueness 
bigint 
8 


?? 
filter1 
[nspid]RequiredDiffImage 
WSA NonSurvey 
UID of WFCAM narrow band (primary) filter 
tinyint 
1 


INST_FILTER_CODE 
filter2 
[nspid]RequiredDiffImage 
WSA NonSurvey 
UID of WFCAM broad band (reference) filter to be subtracted 
tinyint 
1 


INST_FILTER_CODE 
filterID 
[nspid]Detection, [nspid]Filter, [nspid]ListRemeasurement, [nspid]RequiredFilters, [nspid]RequiredMosaic, [nspid]RequiredStack, [nspid]SatelliteDetection, [nspid]UKIDSSDetection 
WSA NonSurvey 
UID of combined filter (assigned in WSA: 1=Z,2=Y,3=J,4=H,5=K,6=H2,7=Br,8=blank) 
tinyint 
1 


INST_FILTER_CODE 
filterID 
[nspid]Multiframe 
WSA NonSurvey 
UID of combined filter (assigned in WSA: 1=Z,2=Y,3=J,4=H,5=K,6=H2,7=Br,8=blank,9=1.205nbJ,10=1.619nbH,11=1.644FeII) 
tinyint 
1 


meta.code;instr.filter 
filterID 
[nspid]MultiframeDetector 
WSA NonSurvey 
UID of combined filter (assigned in WSA: 1=Z,2=Y,3=J,4=H,5=K,6=H2,7=Br,8=blank,9=1.205nbJ,10=1.619nbH,11=1.644FeII) {image primary HDU keyword: FILTER} 
tinyint 
1 


meta.code;instr.filter 
filterID 
[nspid]Orphan 
WSA NonSurvey 
UID of combined filter 
tinyint 
1 


INST_FILTER_CODE 
filterID 
[nspid]RequiredListDrivenProduct 
WSA NonSurvey 
filterID of the data that the list is driven from. If 0, then take the source list. 
tinyint 
1 



filterName 
[nspid]Multiframe 
WSA NonSurvey 
WFCAM combined filter name {image primary HDU keyword: FILTER} 
varchar 
8 


?? 
firstDerMM 
[nspid]SatelliteOrbits 
WSA NonSurvey 
First time derviative of the Mean Motion divided by two 
float 
8 



flag 
[nspid]SourceXDetectionBestMatch, [nspid]SourceXSynopticSourceBestMatch 
WSA NonSurvey 
Flag for potential matching problems 
tinyint 
1 

0 

flag=1 if the same intermediate stack detection is linked to two different unique sources. This can happen in images where the seeing was poorer than average or if a source has moved over time and overlaps with another source. flag=2 no intermediate stack detection, but the expected location is in 1 dither offset of the edge of the stack. 
flatID 
[nspid]Multiframe 
WSA NonSurvey 
UID of library calibration flatfield frame {image extension keyword: FLATCOR} 
bigint 
8 

99999999 
obs.field 
flux 
[nspid]SatelliteDetection 
WSA NonSurvey 
Instrumental isophotal flux counts 
real 
4 
ADU 

PHOT_INTENSITY_ADU 
fluxErr 
[nspid]SatelliteDetection 
WSA NonSurvey 
Error in instrumental isophotal flux counts 
real 
4 
ADU 

ERROR 
focusFiltOff 
[nspid]Multiframe 
WSA NonSurvey 
Focus filter offset {image primary HDU keyword: FOC_FOFF} 
real 
4 
millimetres 
0.9999995e9 
instr.param 
focusInstFiltOff 
[nspid]Multiframe 
WSA NonSurvey 
focus offset for inst. filter {image primary HDU keyword: TEL_FOFF} 
real 
4 
millimetres 
0.9999995e9 
instr.param 
focusNominOff 
[nspid]Multiframe 
WSA NonSurvey 
Offset from nominal focus position {image primary HDU keyword: FOC_OFF} 
real 
4 
millimetres 
0.9999995e9 
instr.param 
focusOffset 
[nspid]Multiframe 
WSA NonSurvey 
Focus offset {image primary HDU keyword: FOC_OFFS} 
real 
4 
millimetres 
0.9999995e9 
instr.param 
focusPos 
[nspid]Multiframe 
WSA NonSurvey 
Focus position {image primary HDU keyword: FOC_POSN} 
real 
4 
millimetres 
0.9999995e9 
instr.param 
focusSerial 
[nspid]Multiframe 
WSA NonSurvey 
Serial number in focus scan {image primary HDU keyword: FOC_I} 
int 
4 

99999999 
?? 
focusZero 
[nspid]Multiframe 
WSA NonSurvey 
Focus zeropoint position {image primary HDU keyword: FOC_ZERO} 
real 
4 
millimetres 
0.9999995e9 
instr.param 
frameSetID 
[nspid]DxsSource, [nspid]ExtendedSource, [nspid]GcsPointSource, [nspid]GpsPointSource, [nspid]JHKsource, [nspid]JKsource, [nspid]LasPointSource, [nspid]PointSource, [nspid]UdsSource, [nspid]YJHKsource, [nspid]ZYJHKsource 
WSA NonSurvey 
UID of the set of frames that this merged source comes from 
bigint 
8 


REFER_CODE 
frameSetID 
[nspid]FrameSets, [nspid]JHKmergeLog, [nspid]JKmergeLog, [nspid]YJHKmergeLog, [nspid]ZYJHKmergeLog 
WSA NonSurvey 
frame set ID, unique over the whole WSA via programme ID prefix, assigned by merging procedure 
bigint 
8 


ID_FIELD 
frameSetID 
[nspid]MergeLog, [nspid]VarFrameSetInfo, [nspid]Variability 
WSA NonSurvey 
frame set ID, unique over the whole WSA via programme ID prefix, assigned by merging procedure 
bigint 
8 


ID_FIELD 
Each merged source in the merged source tables come from a set of individual passband frames (with different filters and/or different epochs of observation). In the WSA, a frame is generally the image provided by one detector (ditherstacked and interlaced as appropriate); hence a frame set comprises a set of individual detector frames in different passbands and/or at different observation epochs. Each frame set is uniquely identified by the attribute frameSetID, and this references a row in the corresponding merge log for the source table (for example, lasSource.frameSetID references lasMergeLog.frameSetID. The merge log in turn references the full set of image descriptive data held in the tables MultiframeDetector and ultimately Multiframe (these two tables map directly onto the multiextension FITS file hierarchy of extension FITS headers beneath a single primary HDU FITS header  primary HDU FITS keys will be found in Multiframe, while the corresponding extension FITS keys for each primary set will be found in table MultiframeDetector). In this way, you can trace the provenance of a merged source record right back to the individual image frames from which it is derived. 
frameSetID 
[nspid]Source 
WSA NonSurvey 
UID of the set of frames that this merged source comes from 
bigint 
8 


REFER_CODE 
Each merged source in the merged source tables come from a set of individual passband frames (with different filters and/or different epochs of observation). In the WSA, a frame is generally the image provided by one detector (ditherstacked and interlaced as appropriate); hence a frame set comprises a set of individual detector frames in different passbands and/or at different observation epochs. Each frame set is uniquely identified by the attribute frameSetID, and this references a row in the corresponding merge log for the source table (for example, lasSource.frameSetID references lasMergeLog.frameSetID. The merge log in turn references the full set of image descriptive data held in the tables MultiframeDetector and ultimately Multiframe (these two tables map directly onto the multiextension FITS file hierarchy of extension FITS headers beneath a single primary HDU FITS header  primary HDU FITS keys will be found in Multiframe, while the corresponding extension FITS keys for each primary set will be found in table MultiframeDetector). In this way, you can trace the provenance of a merged source record right back to the individual image frames from which it is derived. 
frameSetID 
[nspid]SourceRemeasurement 
WSA NonSurvey 
UID of the set of frames that this remeasured source comes from 
bigint 
8 


REFER_CODE 
frameSetTolerance 
[nspid]Programme 
WSA NonSurvey 
The match tolerance for different passband frames 
real 
4 
Degrees 

?? 
frameType 
[nspid]Multiframe 
WSA NonSurvey 
The type of multiframe (eg. stacktilemosaicdifferencecalibrationinterleaved etc). A multiframe can have a combination of different types. 
varchar 
64 

normal 
meta.code.class 
The frame types and their abbreviations are: confidence = "conf"  dark = "dark"  deep = "deep"  difference = "diff"  filtered = "filt"  flat = "flat"  interleaved = "leav"  mosaic = "mosaic"  sky = "sky"  stack = "stack"  default value = "normal"  
frinID 
[nspid]Multiframe 
WSA NonSurvey 
UID of library calibration fringe frame 
bigint 
8 

99999999 
obs.field 
fsTemp 
[nspid]Multiframe 
WSA NonSurvey 
CCC 1st stage temperature {image primary HDU keyword: FS_TEMP} 
real 
4 
degrees_Kelvin 
0.9999995e9 
