i. Aperture Photometry
Aperture photometry is performed for each non-saturated
source detection made on
the Atlas Images to provide a reference to the absolute photometric scale
for the profile-fit photometry, statistics on the detectability of an object,
and as a back-up source of brightness information when profile-fitting
fails to converge to a valid measurement.
As with the profile fitting,
the aperture photometric measurement is made at the position of a
detection on the individual frames. On each frame, the brightness is
measured in a series of apertures ranging in radius from 3´´ to
14´´, in 1´´ steps, by summing pixels entirely
within the aperture and interpolating
pixels partially within the aperture. The sky background for each object
is computed in an annulus with an inner radius of 14.0´´ and
an outer radius of 20.0´´. Pixels in the sky annulus are entirely
included or excluded based on the distance of their centers from the source.
The sky value is estimated by first excluding saturated, masked, or
unreasonably low pixels. A -trimmed median
of the surviving sky pixels is then used as the sky brightness estimate.
The aperture measurements from each of the six input frames are combined
using an unweighted average which takes into account non-detections
in frames to avoid flux-overestimation.
Aperture measurements are usually possible on all six
(and sometimes seven) frames that sample the position of each detection.
However, if one of more of the frames contains a masked or saturated pixel
within 4" of the source centroid, or if the source centroid
falls within <4'' of a frame edge, that frame is excluded from
the the measurement. The aperture measurements from
each of the remaining available frames are combined
using an unweighted average.
The statistics of the aperture measurements are compiled in
the ndet
parameter included in the PSC record. Ndet is a six-digit flag,
with two digits per band
(NJMJNHMHNKMK)
that tabulates for each band the number of frames on which a source
was detected with >3 in aperture the photometry,
Nb, and
the number of frames which were available for measurement, Mb.
Thus, the frame detection thresholds are also referred to
as the "N-out-of-M" criteria.
For example,
an ndet value of "665634" indicates that a source had
six >3 detections
on all six of the available frames in the J-band, five detections out of
six possible frames in the H-band, and three detection on four possible
frames in the Ks-band. For brighter sources, the
ndet parameter can be used as a reliability indicator
and is used in source selection for
the final PSC.
ii. Curve-of-Growth Correction
The "standard aperture" used for 2MASS aperture photometry had
a 4'' radius (or ~2 camera pixels).
However, the point-spread-function of the 2MASS optical system
is broad enough that light from the wings of star images
is not completely captured within this aperture.
Depending on the atmospheric seeing conditions, between 2% and 15%
of the total flux from a point source will be missed.
To correct for the loss of light in the standard aperture, a
curve-of-growth correction is applied to the
measurements, and it is this corrected photometry that is
listed in the standard aperture magnitudes
(j_m_stdap,
h_m_stdap,
k_m_stdap)
in the PSC source records.
The curve-of-growth correction is a constant factor
that when added to the magnitudes measured in the 4'' radius aperture
makes them equivalent to an "infinite" size aperture.
Correcting small-aperture measurements rather than
using much larger apertures avoids the degradation in signal to noise ratio
due to increased sky photon noise and possible confusion with nearby sources
that would result from simply using larger aperture measurements.
However, the correction means that the aperture photometry quoted
for 2MASS sources assumes that the sources are unresolved, and
have a profile consistent with the point-spread-function of the
optical system plus seeing. Resolved or multiple sources
may not be accurately characterized by the standard aperture measurements.
The curve-of-growth corrections applied to standard aperture photometry
during final 2MASS data processing were drawn from look-up tables
indexed by atmospheric seeing. Photometry for all sources in a scan
observed within specified ranges of measured seeing conditions were
corrected using the same factor, per band.
The correction-table values
were derived empirically using the multi-aperture
photometry of large ensembles of non-saturated sources
(rd_flg="2")
observed in relatively low source density scans
from each observatory under a wide variety of seeing conditions.
For each observatory, time period and seeing value, the differences
between magnitudes in successive aperture were compiled, and
the point at which the differences converge to zero was determined.
The correction factor, in magnitudes, is the median difference
between the 4´´ aperture magnitude and the magnitude in the
aperture at which the magnitude differentials become indistinguishable from
zero. The mean correction values for each band and
each seeing range were internally consistent; the median and the mean
values agree to <0.01 magnitudes.
J-band curves-of-growth evaluated during periods of good (2.5´´
FWHM) and bad (3.4´´ FWHM) seeing in scans from the Second
Incremental Data Release are shown in Figure
1 and Figure 2, respectively. The
x-axes show aperture radius in pixels (2´´/pixel), and the
y-axes show the magnitude differences measured in successive aperture pairs
vs.
the outer radius of the pair. Indicated on the figures are the radius
of the "standard aperture" (open circle), and the radius at which the
curve-of-growth was determined to converge (open star). The good seeing
case shown in Figure 1, converges
at a smaller radius, 2.5 pix (5´´) than the poor seeing case,
4.0 pix (8´´).
The curve-of-growth correction determined in these examples were
-0.016±0.005 mags and -0.121±0.015 mags, respectively.
Note that the net corrections in each case are the sum of the differentials
for all pairs between 4'' and the convergence radii.
Figure 1 | Figure 2 |
iii. Photometric Normalization
Profile-fit Photometry
Curve-of-growth-corrected aperture photometry of high SNR, unconfused,
non-saturated sources on the 1.3 s
exposures best defines the true photometric scale for 2MASS.
In the absence of confusion, such measurements best capture the total
flux of points sources and are least influenced by effects such as focal
plane distortion and differences between the seeing-influenced instantaneous
point-spread-function and model point-spread-functions used in
profile-fitting photometry.
Therefore, both the profile-fitting magnitudes, and those derived from
aperture photometry on the 51 ms "Read_1" frames are normalized to
the curve-of-growth-corrected aperture photometry of bright,
non-saturated sources measured on 1.3 s "Read_2" frames.
Normalization of the profile-fitting photometry was usually
done by deriving empirical offsets between the profile-fit and
curve-of-growth-corrected aperture magnitudes for high SNR sources
in each scan. Since the normalization changes with seeing,
different offsets were derived for sources in a scan having
different measured characteristic seeing values.
For a given band and seeing range, the normalization constants were
derived by measuring the median difference between profile-fit and
curve-of-growth-corrected aperture photometry for all sources in a scan
within that seeing range
using iterative 3 rejection.
The -rejection excluded confused sources
which can have contaminated aperture or profile-fit photometry, or both.
The resulting offset correction is added to the profile-fit photometry,
and this corrected value is the default magnitude
listed in the PSC record for sources with
rd_flg="2".
Figure 3 shows the relationship between
profile-fit and curve-of-growth-corrected aperture magnitudes for a scan
in the All-Sky Data Release,
after the normalization correction has been applied to the profile-fit
photometry.
In scans of Tiles with source densities too low to provide enough
sources for the normalization derivation, or with source densities
so high that confusion corrupts the aperture photometry
(>41,000 deg-2), the photometric
normalization constants could not be reliably determined empirically.
The profile-fit photometry in these scans were
selected from a look-up table indexed by the seeing.
The correction tables were derived using the empirical corrections calculated
in the preliminary processing, using all survey scans with fewer than 2000
sources per Atlas Image (17´ in length).
Bright Star Aperture Photometry
Normalization of all of the aperture photometry of bright stars
measured on the 51 ms "Read 1" exposures was made using the
corrections taken from a seeing-indexed look-up table.
As with the curve-of-growth corrections, these normalization factors
were derived before processing began using photometry of large ensembles of
stars from both observatories taken under a wide range of seeing conditions
that were bright enough to be detected in the 51 ms exposures,
but not bright enough to be saturated in the 1.3 s "Read 2" exposures.
There was typically 2-3 magnitudes of brightness overlap in which
sources would satisfy this requirement.
The mean offsets between curve-of-growth-corrected aperture magnitudes
and the aperture magnitudes from the 51 ms exposures were derived
for all stars within various seeing ranges.
The results for each seeing range were internally consistent;
the median and the mean values agree to better than 0.001 magnitude, and
the standard deviation of the mean offset values were <0.01 mag.
The 51 ms exposure aperture photometry corrections ranged in amplitude
from -0.02 mag to -0.12 mag for seeing values of ~2´´ to
3.5´´.
These normalization factors were added to the 51 ms exposure aperture
photometry, and the corrected magnitudes are listed in the
default magnitude fields for
the PSC sources with
rd_flg="1").
Figure 4 shows the normalized
51 ms exposure aperture magnitudes plotted as a function of their
corrected profile-fit magnitudes for a full night's data in the
All-Sky Data Release. This figure also illustrates the overlap in the
magnitude range over which both 51 ms and 1.3 s exposures provide
useful measurements for determining the normalizations.
Figure 3 | Figure 4 |
vi. Curve-of-Growth and Photometric Normalization Errors
After the final 2MASS data processing was completed, it was discovered
that out-of-date correction look-up-tables were used for
the normalization of profil-fit photometry in the very high and low
source density cases. The differences between the correct
values and those used resulted in a systematic overestimate of
source brightnesses after the correction, the amplitude of which is a
function of seeing FWHM and the photometric band. The overestimation
ranges from <0.01 mag in good seeing, to a maximum of 0.02-0.03 mag
for the worst seeing in the Survey. Because the amplitude of the bias
can be different in each band, this error can produce residual color-biases
on ~17´ scales along a Tile, and ~8.5´ across Tile boundaries.
This bias is believed to cause at least part of the discrete color
jumps correlated with Atlas Image boundaries seen in spatially-averaged
color maps of high source density regions. An example of this
is given in Figure 5 which
shows a map of mean H-Ks source color made using the
normalized profile-fit photometry of PSC sources in a
3°x6° field near the galactic center.
However, the amplitude of these jumps can be as large as 0.05 mag, as seen in
Figure 6 which shows the
H-Ks colors of individual stars along one of the
scans in Figure 5.
Therefore, the observed discrete color biases cannot be fully
accounted for by the PSF-normalization errors. We do not have a
complete understanding of the origin of these color jumps, but they
are known to be correlated to changes in the PSF used in profile-fit
photometry. The biases are not seen in colors derived
from aperture photometry.
Figure 5 | Figure 6 |
[Last Updated: 2003 March 11; by R. Cutri, T. Evans, J. Carpenter]