Comparison of 1.3 and 1.0 sec Exposure Times for P565-C Calibration Scans on 970521 UT

R. Cutri - IPAC

24 May 1997

The plots below show the mean repeatability within 0.5 magnitude wide bins for photometry and astrometry of sources detected in the multiple scans of the P565-C Calibration Field on 970521 UT. Figures 1-3 show J, H and Ks data, respectively. The top panel in each figure shows the average magnitude dispersion (population sigma) for all sources extracted in at least 5 out of the 6 repeats of the field. The center and bottom panels show average dispersion in right ascension and declination (in arc-seconds) for each source seen in at least 5 of the 6 repeats. In each panel, the green curves represent the repeatability measured in Scans 052-057, which used a 1.3 second R2-R1 exposure time, and the red is the repeatability observed in Scans 058-062, which used a 1.0 second net R1-R1 exposure. Note that no photometric zero point or extinction corrections were applied to these data. This is a purely internal comparison.

Fig. 1 - Mean Photometric and Astrometric Repeatability for P565-C, J-Band

Fig. 2 - Mean Photometric and Astrometric Repeatability for P565-C, H-Band

Fig. 3 - Mean Photometric and Astrometric Repeatability for P565-C, Ks-Band

The photometric precision for bright sources is approximately the same for both exposure times, and satisfies the Level 1 Requirement for Point Source Photometric Precision of 4%. The origin of the unusually large dispersion in the bright star magnitudes for the 1.0sec Ks measurments is currently unknown. It is likely not related to the measurements, however. The dashed line in the magnitude dispersion panel indicates the SNR=10 level (0.109 mags). The table below summarizes the mean achieved 10-sigma photometric levels (in magnitudes) in each band for each of the exposures. For reference, the value of Level 1 Science Requirement for Point Source Photometric Accuracy is also listed. In each band, the 1.3 seconds exposures meet the Level 1 Requirements with a small margin. The 1.0 second exposures also meet the Requirements, but with very little margin. The degradation in the SNR=10 level for the 1.0 second exposures relative to 1.3 seconds is, on average, about 0.22 mags. This is larger than the 0.14 mags predicted solely from the ratio of exposure times which indicates that there are additional systematic affects impacting the photometric accuracy for the shorter exposures.

Band	Level 1 Req.	10-sigma (1.3 sec)	10-sigma (1.0 sec)
J	15.8	16.3	16.1
H	15.1	15.5	15.3
Ks	14.3	14.6	14.4

Note that these scans were obtained under conditions of moderately good seeing (scan averages ~2.4"), and moderate background levels (~490DN, ~1600DN, ~1900DN). The margins afforded by the 1.3 seconds exposures imply that we can operate successfully under slightly worse conditions, and still satisfy the Level 1 photometric accuracy requirements, and probably completeness and reliability requirements. With 1.0 seconds R2-R1 exposures, there is no margin, and any degradation in conditions will mean a failure to meet the Level 1 Requirements. In particular, the H background levels for this scan are approximately equal to the mean H background observed during the 1995 Prototype Camera run on Kitt Peak. The H-band sensitivity (in magnitudes) will decrease as 0.5 time the H-band background surface brightness ( 2MASS Airglow Page ). Thus, a ~20% increase in H background, which was not uncommon during the 1995 ProtoCam run, will result in a failure to meet the Level 1 Specifications for H-band point source photometric accuracy with 1.0 sec frame exposure times. A 1.3 second exposure can tolerate nearly double the background level before the sensitivity is degraded to that near the Level 1 Spec. The highest H-band backgrounds observed in 1995 were ~3000DN. Thus, adopting a 1.0 second frame exposure will likely result in an increased duration for the Survey, since a much smaller fraction of the nominally photometric time will be useful.

The repeatibility of position reconstruction for sources in these scans is relatively unaffected by the decrease in exposure time, except at the very faintest flux levels. The Level 1 Requirement for astrometric accuracy is 0.5" for SNR>20 sources. The high degree of internal accuracy of positions in these scans far exceeds the requirement, although external accuracy with respect to the Tycho Catalog remains to be seen.

There were also multiple repeats of 6 degree scans covering the region of the Hercules cluster of galaxies taken on this night. These longer scans will allow a more detailed study of the impact of varying exposure time on both point and extended source processing relative to the Level 1 Science Requirements for 2MASS.