<h2>DVO Documentation</h2>

<h3>Release Notes on 3pi.20120525</h3>

This database contains PS1 3pi observations through 2012/01/21.  The
ubercal zero points have been applied to all exposures included in
that analysis.  The remaining exposures have been tied to the ubercal
system via relative photometry.  The ubercal measurements are given
10x their poisson weight to ensure they drive the photometric
solution.

The database has been merged with the PS1 Synthetic photometry
database (based on 2MASS and USNO-B).  For objects brighter than the
saturation limit, the Synthetic photometry has been used for the mean
magnitudes (these can be identified by having bit 0x04 raised in the
mean filter flags).  Objects which used ubercal photometry have the
bit 0x08 raised in the mean filter flags.  

This database has also been merged with the 2MASS, WISE, and
SuperCOSMOS databases.  The following photcodes are used for those
external photometric sources:

<table class="page">
<tr><th> Source </th><th> Photcode </th></tr>
<tr><td> 2MASS_J </td><td> 2011 </td></tr>   
<tr><td> 2MASS_H </td><td> 2012 </td></tr>   
<tr><td> 2MASS_K </td><td> 2013 </td></tr>   
<tr><td> WISE_W1 </td><td> 2014 </td></tr>   
<tr><td> WISE_W2 </td><td> 2015 </td></tr>   
<tr><td> WISE_W3 </td><td> 2016 </td></tr>   
<tr><td> WISE_W4 </td><td> 2017 </td></tr>   
<tr><td> SYNTH.g </td><td> 3001 </td></tr>   
<tr><td> SYNTH.r </td><td> 3002 </td></tr>   
<tr><td> SYNTH.i </td><td> 3003 </td></tr>   
<tr><td> SYNTH.z </td><td> 3004 </td></tr>   
<tr><td> SYNTH.y </td><td> 3005 </td></tr>
<tr><td> SCOS.103a.E       </td><td>      1150  </td></tr>
<tr><td> SCOS.4414.OG590   </td><td>      1151  </td></tr>
<tr><td> SCOS.4415.OG590   </td><td>      1152  </td></tr>
<tr><td> SCOS.IIIaF.OG590  </td><td>      1153  </td></tr>
<tr><td> SCOS.IIIaF.RG610  </td><td>      1154  </td></tr>
<tr><td> SCOS.IIIaF.RG630  </td><td>      1155  </td></tr>
<tr><td> SCOS.IIIaJ.GG385  </td><td>      1156  </td></tr>
<tr><td> SCOS.IIIaJ.GG395  </td><td>      1157  </td></tr>
<tr><td> SCOS.IVN.RG715    </td><td>      1158  </td></tr>
<tr><td> SCOS.IVN.RG9      </td><td>      1159  </td></tr>
</table>

<img width=1000px src="relphot.20120528/sigma.i.ubercal.20120528.v0.png"><br>
Map of the mean per-star residual (inst mag brighter than -10) for
i-band (ubercal data only).
<br>

<img width=1000px src="relphot.20120528/sigma.g.ubercal.20120528.v0.png"><br>
Map of the mean per-star residual (inst mag brighter than -10) for
g-band (ubercal data only).
<br>

<h3>Comparison with 2MASS</h3>

The plots below compare the PS1 grizy photometry with 2MASS photometry
by examining the location of the color-color locus for early (&lt; K0)
main sequence stars.

For each pixel (0.5 degree on a side), I select all objects with PS1 r
&lt; 20 and detected JHK magnitudes, keeping only those with S/N in
the mean &gt; 50 and standard deviations less than 0.05 magnitudes.  I
then fit the stellar locus in g-r,r-i for the g-r range 0.2 to 1.0,
and keep only objects that fall within 0.05 of that (linear) stellar
locus fit.<br>

I then fit the stellar locus with a line in each of the color-color diagrams
g-i,i-J; g-i,i-H; g-i,i-K.  I measure the mean i-J color expected from
the fit for stars with g-i = 0.5.  The figures below show the maps of
the i-J, i-H, and i-K offsets as well as the J-H and H-K offsets at
g-i = 0.5.  The center of these maps is RA,DEC = 0,0, (east is left,
north is up).  The pixels are 0.5 degrees on a side.<br> 

These maps show three main features.  The Galactic Plane is quite
clear, and not surprising: this test is particularly sensitive to
reddening. <br>

Several large scale patches are also visible (eg, the black patch near
RA ~ 150 deg).  Disappointingly, these correspond to areas without
good ubercal ties -- in other words, these are regions in which the
data were obtained in non-photometric conditions.  This shows the
limitation of using relative photometry to tie across individual
exposures.  If the exposures have an extinction gradient, the
correction far from the photometrically tied region can drift from the
correct value.  The largest region at RA ~ 150 matches a section with
poor i-band photometry and has a depth of roughly 0.1 magnitudes in
i-J.  Assuming this is completely due to i-band errors, this
corresponds to about 0.06 magnitudes of drift to the center of this
region. <br>

The third pattern is at a lower level of intensity: N-S strips with
roughly E-W bands.  They are most noticable in K and H-K.  These
appear to correspond well to the 2MASS scanning pattern, and have an
amplitude corresponding to about 0.02 - 0.04 magnitudes of offset in
K.  <br>

<img width=1000px src="relphot.20120528/dJ.allsky.20120528.v3.png"><br>
Map of the i-J color at g-i = 0.5. 
<br>

<img width=1000px src="relphot.20120528/dH.allsky.20120528.v3.png"><br>
Map of the i-H color at g-i = 0.5. 
<br>

<img width=1000px src="relphot.20120528/dK.allsky.20120528.v3.png"><br>
Map of the i-K color at g-i = 0.5. 
<br>

Combinations of the above maps can highlight any problems in 2MASS
which are driven by the color rather than the mean magnitudes.

<img width=1000px src="relphot.20120528/dJH.allsky.20120528.v3.png"><br>
Map of the J-H color at g-i = 0.5. 
<br>

<img width=1000px src="relphot.20120528/dHK.allsky.20120528.v3.png"><br>
Map of the H-K color at g-i = 0.5. 
<br>
