SSD Technical/Training Series - Photometry III

Brad Whitmore
10-10:30am, Monday
December 7, 1998
Auditorium

SSD Technical/Training Series - Photometry I- III Outline

November 5 - Photometry I - Brad Whitmore

1. MOTIVATION (Big Science from small measurements)
   • H-R diagram
   • Age of the universe from Cepheid variables
2. PRACTICAL MEASUREMENTS
   • Using APPHOT (walk through the WFPC2 Tutorial)
   • Homework (due Nov. 11)
3. BASICS
   • Types of photometry
   • Magnitude systems
   • Zeropoints

November 12 - Photometry II - Brad Whitmore

1. MORE BASICS
   • Second part of the homework (due Nov. 18)
   • Zeropoints
   • Aperture corrections
   • Optimal aperture sizes
   • Contamination
   • Error estimates
2. THE NITTY GRITTY
   • Color terms
   • spatial PSF variations
   • Charge Transfer Efficiency (CTE)
   • Long vs. short problem

December 7 - Photometry III - Brad Whitmore

• Homework #2 Results
• Homework Assignment #3
• Charge Transfer Efficiency (CTE)
• "Long-vs-short" Anomaly
• Common Gotchas
• Other Things I Never Metioned

December 10 - Photometry IV -

1. STIS PHOTOMETRY - Harry Ferguson
2. NICMOS PHOTOMETRY - Mark Dickinson

PHOTOMETRY III

HOMEWORK # 2 RESULTS -

• The efficiency of transferring the charge from pixel to pixel during the readout of an image is not 1.0000000, but more like 0.99995 ("four nines five"). Hence a star at the top of the chip looses more of its charge during the readout, and appears to be FAINTER by 800 X (1-0.99995) = 4 %.
• There is a similar, but smaller effect along the X axis, presumably due to the shift register.
• The amount of CTE also depends on:
  • target brightness,
  • background, and
  • aperture size.
• The CTE loss is relatively small for bright stars, but can be 20 - 30 % for very faint stars on very faint backgrounds.
• See ISR 97-08 for details, and for a set of equations to correct for CTE to about the 5 % level or better.
• CTE seems to be increasing with time, primarily for faint objects. See TIR 98-01 for details.

• Short exposures appear to give fainter magnitudes than long exposures. However, this turns out to be a simple non-linearity which depends soley on TOTAL COUNTS in a stellar image, and is actually independent of exposure time. Hence "Long-vs-Short" is a misnomer.
• The effect can be as large as 40 % for very faint stars.
• A simple formula can be used to correct for the effect at roughly the 0.05 magnitude level.
• See ISR 98-02.

• gain = 7 or 15
• ST-MAG (see section 8.7 in WFPC2 Instrument Handbook)
• red leaks (see section 3.8 in WFPC2 Instrument Handbook)

• Dependence on how well centered on the pixel. Holtzman-1 found < 0.01 effect.
• The PSF varies over the field of view of the chip. For 1 pixel radius this can be roughly a 0.05 magnitude difference near the corners.
• The quantized nature of the pixel values add noise to the signal and can also result in subtle systematic errors (e.g., if you use medians). See p. 279 of the 1995 Calibration Workshop.
• etc. etc. etc.