Coma Galaxy: 194.915329+27.953882
T. Jarrett, IPAC
(970522)
The Coma galaxy 194.915+27.9539 is located
near the giant elliptical NGC 4881. Its photometric
repeatibility is very poor, with variations as
large as 30%. The following images show some of the
detailed GALWORKS examination of this galaxy, including
the flux growth (adaptive aperture) results.
The bottom line is that the presence of the large
elliptical nearby (within 90 arc sec) has rendered the
adaptive aperture photometry unreliable.
The following are postage stamp images of the galaxy as
seen in the eight scans. There is very little difference
between them.
Two scans are examined in detail: 050 and 054, both
from the night of the 23rd. The adaptive ellipical
aperture mag for scan 050 is 11.429 compared to
12.642 for scan 054. The difference is due to the
elliptical aperture radius: 35 vs. 9 pixels,
respectively.
Before the images are flux integrated, they are first
masked of nearby stars and objects previously
processed. The following images show subimage sections
centered on the galaxy of interest (194.915+27.9539)
before star masking, after star masking, and finally
after isophotal substitution.
The subimages show the large elliptical to the west of
galaxy 194.915+27.9539 masked from the image (because it
was processed previously before galaxy 194.915+27.9539).
The two scans look much the same -- there is no
indication of any problem at this juncture.
Flux Growth Aperture Photometry
The detailed elliptical adaptive aperture photometry results
are given below. The tables show the
K band integrated flux
at each aperture radius, as well as the surface
brightness change in the outer annulus. Column (1) is the
semi-major axis of the aperture, col (2) the
integrated flux in mag units, col (3) the change in
int. flux in mag units, and col (4) the change in
surface brightness (ratio of the change in surf brightness
to the surf brightness in the outer annulus). The other
columns are not important to this discussion.
The basic method is better explained in
Adaptive Aperture Photometry
Scan 050
7.000 12.781 -.---- -.--- -.--- petv= 1.000
8.000 12.699 0.0819 4.905 2.578 petv= 3.013
9.000 12.625 0.0740 0.364 0.236 petv= 3.158
10.000 12.562 0.0637 0.481 0.455 petv= 3.814
11.000 12.518 0.0436 0.188 0.216 petv= 4.057
12.000 12.467 0.0513 0.007 0.008 petv= 3.579
13.000 12.420 0.0465 0.130 0.177 petv= 3.711
14.000 12.374 0.0461 -0.001 -0.001 petv= 3.333
15.000 12.328 0.0463 -0.061 -0.077 petv= 2.832
17.000 12.210 0.1178 -0.061 -0.071 petv= 2.253
19.000 12.095 0.1154 -0.041 -0.046 petv= 1.927
21.000 11.988 0.1072 0.107 0.136 petv= 1.968
24.000 11.847 0.1404 0.089 0.127 petv= 1.933
27.000 11.720 0.1275 0.055 0.084 petv= 1.869
30.000 11.626 0.0932 0.189 0.415 petv= 2.329
35.000 11.422 0.2042 -0.155 -0.253 petv= 1.544
The convergence criteria are: (1) 0.005 for flux convergence
and (2) -0.1155 for the surface brighteness measure.
Notice that the curve of growth did not actually have
flux convergence, but instead met the surface brightness
criteria at the point between R=30 and 35 aperture.
The sort
of behavior is typical of galaxies contaminated by either
nearby stars or by background gradient (in this case, the
giant elliptical galaxy nearby). In general, galaxies
will first meet the flux convergence criteria before
the surface brightness criterion.
Now compare these results with those of scan 054, below.
Scan 054
6.000 12.901 -.---- -.--- -.--- petv= 1.000
7.000 12.776 0.1248 6.731 2.699 petv= 2.860
8.000 12.705 0.0703 0.959 0.624 petv= 3.782
9.000 12.647 0.0585 -0.135 -0.081 petv= 3.076
Again, this galaxy does not complete flux convergence,
but instead has an immediate reversal in the surface
brightness at the R=9 aperture, tripping the
surface brighteness criterion (which is -0.119 in this
case). Note however that the aperture is much smaller
for scan 054 than for scan 050. We already know that the
nearby elliptical galaxy is contaminating the results,
but since we arrive at two very different radii, this indicates
that our convergence criteria is not robust to this
sort of situation (e.g., the surf brightness criterion
is a function of the coadd sky noise normalized to an
aperture size of radius = 10 pixels -- a parameter not
well optimized to large (but still local) flux gradients).
The surf brightness criterion should probably be
tighter in this case, to reflect the actual noise
"local" to the galaxy (not just the coadd noise).
Inspection of the scan 050 table indicates that
a tighter threshold would force termination of the
aperture growth around R = 15 or so, corresponding
to a flux mag of about 12.4, which is only 20%
different from the 12.6 mag value for scan 054.
The circular adaptive photometry for both scans
never converges or trips the surf brightness crit.
Instead the aperture grows until the maximum radius
is achieved, about 70 pixels. The results are given
below:
Scan 050, circular adaptive aperture photom (K band)
8.000 12.596 0.0793 4.120 3.251 petv= 3.459
9.000 12.536 0.0604 0.386 0.438 petv= 4.093
10.000 12.471 0.0650 0.001 0.001 petv= 3.472
11.000 12.419 0.0519 0.090 0.113 petv= 3.409
12.000 12.363 0.0560 -0.049 -0.059 petv= 2.888
13.000 12.289 0.0743 -0.022 -0.025 petv= 2.514
14.000 12.230 0.0586 0.106 0.140 petv= 2.610
15.000 12.166 0.0639 -0.008 -0.010 petv= 2.368
17.000 12.040 0.1260 -0.058 -0.071 petv= 1.944
19.000 11.900 0.1400 -0.047 -0.054 petv= 1.669
21.000 11.816 0.0844 0.327 0.602 petv= 2.377
24.000 11.684 0.1319 -0.002 -0.004 petv= 2.047
27.000 11.522 0.1618 -0.103 -0.159 petv= 1.565
30.000 11.380 0.1422 0.038 0.061 petv= 1.537
35.000 11.145 0.2346 -0.008 -0.013 petv= 1.378
40.000 10.945 0.1997 0.053 0.093 petv= 1.388
45.000 10.734 0.2117 -0.070 -0.110 petv= 1.186
50.000 10.557 0.1765 0.065 0.114 petv= 1.261
55.000 10.422 0.1357 0.110 0.239 petv= 1.466
60.000 10.292 0.1290 0.007 0.016 petv= 1.412
65.000 10.211 0.0815 0.121 0.364 petv= 1.805
70.000 10.112 0.0987 -0.111 -0.250 petv= 1.314
Scan 054, circular adaptive aperture photom (K band)
7.000 12.712 0.0741 5.011 3.512 petv= 3.664
8.000 12.636 0.0761 0.249 0.211 petv= 3.600
9.000 12.551 0.0851 -0.038 -0.031 petv= 2.938
10.000 12.465 0.0866 0.043 0.037 petv= 2.633
11.000 12.400 0.0646 0.173 0.174 petv= 2.757
12.000 12.334 0.0661 -0.019 -0.019 petv= 2.457
13.000 12.247 0.0873 -0.032 -0.030 petv= 2.153
14.000 12.178 0.0690 0.116 0.124 petv= 2.226
15.000 12.113 0.0650 0.116 0.141 petv= 2.332
17.000 11.977 0.1356 -0.119 -0.127 petv= 1.815
19.000 11.823 0.1545 -0.091 -0.088 petv= 1.522
21.000 11.703 0.1198 0.183 0.217 petv= 1.701
24.000 11.542 0.1610 0.094 0.125 petv= 1.699
27.000 11.401 0.1411 0.111 0.173 petv= 1.778
30.000 11.305 0.0957 0.189 0.418 petv= 2.235
35.000 11.137 0.1679 -0.010 -0.021 petv= 1.869
40.000 11.020 0.1174 0.138 0.426 petv= 2.276
45.000 10.865 0.1549 -0.101 -0.238 petv= 1.580
50.000 10.733 0.1321 0.003 0.008 petv= 1.491
55.000 10.632 0.1010 0.055 0.149 petv= 1.612
60.000 10.594 0.0374 0.226 1.613 petv= 3.798
65.000 10.501 0.0937 -0.194 -0.580 petv= 1.519
Isophotal Photometry
The isophotal photometry comparison is just the opposite
of the adaptive aperture photometry comparison, scan 054
finds a larger isophotal radius (20 mag per sq. arcsec) than
that of scan 050. The results are given below:
scan 050:
elliptical: RK20e = 9.2, K20e = 12.608+-0.062
circular: RK20c = 9.0, K20e = 12.533+-0.068
scan 054:
elliptical: RK20e = 24.0, K20e = 11.826 +-0.078
circular: RK20c = 17.0, K20e = 11.973+-0.077
On further thought, this result is consistent with the adaptive aperture
results. The surface brightness is apparently higher or simply
the noise local to the galaxy is higher, for
scan 054 than scan 050 (the fixed circ aper photom,
see below, suggests that it is 10% higher), thus the surface brightness criterion
is more quickly achieved in scan 054 -- terminating the adaptive
aperture growth before the corresponding termination for scan 050.
This effect suggest one possible solution to help out the adaptive
aperture photometry. Force the adaptive aperture to be at least
as large as the 20 mag per sq. arcsec isophot radius. For scan 054, the
adaptive aperture would then be set to a radius of 24 (equal to the
isophot radius, since the adaptive aperture terminated at a radius of
9 pixels) and for scan 050 there would be no change. The adaptive
aperture photometry for the repeats would then be closer in line to each other.
On the other hand, having a larger adaptive aperture to capture the
total flux might in fact be overestimating this flux due to contamination
from nearby objects (galaxies included).
Fixed Circular Aperture Photometry
Scan 050; note col (1) and col (3)
5 640.4 12.921 0.054 79. 0
10 962.9 12.478 0.072 314. 0
15 1280.9 12.168 0.081 706. 2
20 1716.4 11.850 0.080 1256. 2
25 2098.1 11.632 0.082 1963. 2
30 2653.4 11.377 0.078 2827. 2
40 3953.1 10.945 0.070 5026. 2
50 5653.9 10.556 0.061 7853. 3
60 7203.7 10.293 0.057 11272. 3
70 8507.7 10.112 0.055 14634. 3
Scan 054; note col (1) and col (3)
5 651.0 12.907 0.054 79. 0
10 972.8 12.471 0.072 314. 0
15 1347.7 12.117 0.078 706. 0
20 1877.9 11.757 0.075 1256. 2
25 2420.9 11.481 0.072 1963. 2
30 2845.7 11.306 0.074 2827. 2
40 3705.3 11.019 0.076 5026. 2
50 4823.8 10.733 0.072 7677. 3
60 5488.7 10.592 0.073 10293. 3
70 6429.5 10.421 0.071 13285. 3
Final Remarks
The results presented here demonstrate the problems with adaptive
aperture photomety when the conditions are less than ideal, either
from a high stellar density (in which star contamination is rife) or
from galaxy crowding (the Coma core, e.g.). The latter condition will
of course be rare -- there are only so many nearby dense galaxy
clusters. In most cases -- most of the sky -- the adaptive aperture
photometry should perform satisfactorily (as tests with the protocam
data have shown).
