Colors of Very Bright Stars Data Release Web Page Index

2MAPPS v3.0 Colors of Saturated Read_1 Stars


I. Very Bright Star Photometry

Preliminary 2MASS data processing included a rudimentary effort to estimate the brightness of stars that saturated the 0.51 ms 2MASS Read_1 exposures by photometering the first persistence artifacts from those objects. This proved to be ineffectual and for the incremental data releases "filler" or "placeholder" entries were included in the Point Source Catalog to indicate the presence of bright stars, but no brightness information was provided.

One of the key upgrades included in 2MAPPS v3.0 is an algorithm to estimate the brightness of saturated Read_1 stars by fitting the non-saturated wings of azimuthally-average radial profiles of these objects to a template. The algorithm and template development and calibration are described in a series of memos by Gene Kopan, Raymond Tam and Cong Xu.

The goal for photometric accuracy of the very bright stars is < 0.3 mag per band.

II. Bright Star Colors

The saturated star photometry has been implemented in FREXAS for 2MAPPS v3.0 and is running on the regression test baseline night set. One performance diagnostic of the new algorithm is the colors of the very bright stars relative to those expected for normal stars. The plots below show a series of color-color and color-magnitude diagrams for each processed version of the RTB nights. I suggest that these plots may be useful nightly diagnostics to be included in 2MAPPS v3.0 QA.

Caveat - In the runs before April 26, the identification procedure for saturated R1 stars incorrectly marked a small number of non-saturated stars as saturated and saturated stars as non-saturated. Runs after April 26 should not have this problem.



Table 1 - Color-color plots and CMD's for RTB nights

Date/hemis Proc_date J-H vs. H-Ks J vs J-H H vs H-Ks Ks vs J-Ks
971116n 010327 Fig 1a Fig 1b Fig 1c Fig 1d
010420 Fig 1e Fig 1f Fig 1g Fig 1h
010428 Fig 1i Fig 1j Fig 1k Fig 1l
010504 Fig 1m Fig 1n Fig 1o Fig 1p
980319s 010330 Fig 2a Fig 2b Fig 2c Fig 2d
010421 Fig 2e Fig 2f Fig 2g Fig 2h
010429 Fig 2i Fig 2j Fig 2k Fig 2l
010504 Fig 2m Fig 2n Fig 2o Fig 2p
980403n 010405 Fig 3a Fig 3b Fig 3c Fig 3d
010420 Fig 3e Fig 3f Fig 3g Fig 3h
010429 Fig 3i Fig 3j Fig 3k Fig 3l
010505 Fig 3m Fig 3n Fig 3o Fig 3p
980811s 010517 Fig 4a Fig 4b Fig 4c Fig 4d
981028n Fig 5a Fig 5b Fig 5c Fig 5d
990523n 010329 Fig 6a Fig 6b Fig 6c Fig 6d
010421 Fig 6e Fig 6f Fig 6g Fig 6h
010429 Fig 6i Fig 6j Fig 6k Fig 6l
010505 Fig 6i Fig 6j Fig 6k Fig 6l
990723s 010331 Fig 7a Fig 7b Fig 7c Fig 7d
010422 Fig 7e Fig 7f Fig 7g Fig 7h
010429 Fig 7i Fig 7j Fig 7k Fig 7l
010506 Fig 7m Fig 7n Fig 7o Fig 7p
990923s 010329 Fig 8a Fig 8b Fig 8c Fig 8d
010421 Fig 8e Fig 8f Fig 8g Fig 8h
010428 Fig 8i Fig 8j Fig 8k Fig 8l
010504 Fig 8m Fig 8n Fig 8o Fig 8p
000317n 010412 Fig 9a Fig 9b Fig 9c Fig 9d
010422 Fig 9e Fig 9f Fig 9g Fig 9h
010430 Fig 9i Fig 9j Fig 9k Fig 9l
010506 Fig 9m Fig 9n Fig 9o Fig 9p
000406s 010516 Fig 10a Fig 10b Fig 10c Fig 10d
001126n 010516 Fig 11a Fig 11b Fig 11c Fig 11d
010121s 010516 Fig 12a Fig 12b Fig 12c Fig 12d

A brief description of the plots listed in Table 1 follows:

III. Single-Frame Unsaturated Read 1 Sources

The standard procedure for deriving photometry for bright sources has been to quote the R1 magnitude if a source is unsaturated in two or more frames. If a source is unsaturated in only a single frame, then the pipeline currently passes the single frame R1 aperture magnitude, but encodes it with a magnitude uncertainty of 9.997. In doing the R1/R2 merging, MAPCOR does not pass along the single-frame aperture magnitude, but treats the source as saturated.

For final processing, we need to decide whether to use the single-frame unsaturated R1 aperture photometry for such sources, or the saturated star photometric template fitting from the <5 saturated frames. There are advantages and disadvantages to both.

Single-Frame Aperture Photometry

Saturated-Star Template Photometry

1. Expected Precision of Single-Unsaturated-Frame Read 1 Photometry

The relative precision of aperture photometry from a single unsaturated R1 frame can be estimated from the observed dispersion of the normal aperture photometry from multiple unsaturated frames. Gene Kopan has compiled statistics on the repeatability of the single frame photometry, and his report follows: 

I have checked the photometric repeatability for the single frame
r1 aperture photometry by computing the dispersion for sources found in
multiple frames for some scans from a northern and southern rtb night. The
results show ~0.045 scatter for the north in all 3 bands, a little higher in
the south, but H band in the south ran almost 0.10 mag in the scan I checked,
so I need to check more nights/scans in the south.

...I did a couple more nights/scans in the south, and they also show the H band
r1 single frame repeatablility to be worse in the south than in the
other bands and the north...

gene
---------------
	V3 r1 single frame (*_1.fex) photometric repeatability:

971116n dino:/b2/V3test       Apr 20 2001   no     no   online   82    ver3
971116n/s102-109:

    band     n      sigma mag
      j    1120      0.0449
      h    1447      0.0454
      k    1659      0.0448

980319s barney:/b1/V3test     Apr 21 2001   no     no   online   70    ver3
980319s/s039:

    band     n      sigma mag
      j    1532      0.0553
      h    1447      0.0964
      k    1659      0.0595

990923s wilma:/b1/V3test      Apr 28 2001   no     no   online   72    ver3
990923s/s013-019

    band     n      sigma mag
      j    1066      0.0471
      h    1807      0.0652
      k    1905      0.0578

/990723s/s025:

    band     n      sigma mag
      j    1735      0.0443
      h    4859      0.0834
      k    7685      0.0528

2. Observed Colors of Single-Unsaturated-Frame Read 1 Photometry

Figure 8 shows the JHKs color-color diagram for R1 sources in the 6 RTB nights. The small black points are sources with unsaturated R1 mags in all 3 bands. The larger color points show sources that have single-frame unsaturated R1 mags in one or more bands, as specified in the legend. For these points, the magnitudes in the other bands must be unsaturated R1 or R2, so there is no possible bias introduced by saturated R1 photometry.

Figure 8

There is an apparent bias in the single-unsaturated frame R1 photometry relative to multi-frame unsaturated R1 photometry. We now believe that this arises because single-unsaturated frame R1 star apparitions fall preferentially on edge or corner pixel boundaries where the light is distributed over 2 or 4 pixels. In addition, some light will be lost in the pixel gaps making these apparitions systematically fainter than those centered on pixels. This could account for the observed biases in the J- and H-only single-unsaturated frame colors, but not the Ks-only colors.

This is substantiated by the Intra-Pixel Response Variation analysis performed by Gene Kopan. Sources falling on the edges of edges of pixels can be systemmatically fainter by up to 0.1-0.2 mags than those falling on the centers of pixels. The magnitude of this effect is different for the 7 arrays used in the survey, being most pronounced for the southern H array and the new northern H-band array. The effect is most evident when the seeing is good, as poor seeing smooths the PSF over a larger area.

Figure 9 blinks between the color-color diagrams of single-unsaturated R1 frame sources from the northern and southern RTB nights. The unsaturated R1 points are not shown in this plot. The southern data has systematically better image quality, so if the bias seen in the single-frame photometry is related to the intrapixel losses, then the skew should be more pronounced in the southern data. Unfortunately, the number of single-frame R1 stars in the northern RTB nights is too small to draw any definitive conclusions.

Figure 9

Last Update - 18 May 2001
R. Cutri - IPAC