This comparison involves matching northern and southern scans of seven 2MASS tiles.
Sources were matched using the (RA, Dec) positions, with a tolerance set at
+/-0.001 degrees. I haven't looked quantitatively at the positional agreement, but
a cursory scan over the paired file indicates that the offset is typically
smaller than this tolerance. The second page of the statistics file for
each tile gives the number of matches as a function of J magnitude (from the North
measurement). NB - the columns are mislabelled: No match actually means 1 match;
1 match, 2 matches etc. The overwhelming majority of sources are matched only
once, and since completeness is not an issue for this comparison, I have not
attempted to trim duplicate pairings from the dataset. The first
3 columns list raw number counts in J, H and K before pairing.
The photometric comparison is
limited to PSF magnitudes with Delta(mag) = mag(S) - mag(N) < 1.5 magnitudes;
I also imposed a cutoff at J < 18.5 and sigma(J) < 0.5 (i.e. rms uncertainty
in catalogued PSF magnitude).
Page 1 of the postscript statistics file lists the mean offset, rms dispersion and
number of sources used to define those two parameters; magnitude is again from
the Northern field.
revised Jan 31, 2000 - all sources with bandfil magnitudes eliminated
This plot compares default magnitudes for all sources with J, H, K < 10th magnitude
The above postscript file compares the mean Delta-mag for all
seven fields. In each case, there is a systematic offset, in the
sense that the Southern calibration gives fainter magnitudes
(delta-mag = south-north: positive offset in each case).
At J, the offset is evident for J>9, rising from 0.04 mag to 0.08
At H, the offset is ~0.02 to H~14; rises to ~0.08 at H=16
At K, the offset is ~0.01 to K~12, rising to ~0.07 at K=15.5