Calibration techniques for nextgeneration astronomical systems

1
Calibration techniques for next-generation
astronomical systems

• S. Slatera, C.W. Stubbsa, J.L. Tonryb, J.R.
  Masierob, R.C. Smithc, Team ESSENCE supernova,
  Project LSST
• aDept. of Physics, Harvard University, 17 Oxford
  St., Cambridge, MA, USA 02138
• bInstitute for Astronomy, University
  of Hawaii, 2680 Woodlawn Dr., Honolulu, HI USA
  96822
• cCerro Tololo Inter-American Observatory, Casilla
  603, La Serena, Chile

2
Problems with Conventional Data Reduction
Techniques

• Source spectra can be very unlike stellar
  spectra used for calibration
• Redshift may vary widely between sources under
  comparison
• Systematic error between data sets from
  different telescopes can be difficult to eliminate

Colina, L., Bohlin, R. and Castelli, F. 1996,
Instrument Science Report CAL/SCS-008
3
How Conventional Techniques Fail
System Response function varies with wavelength,
so flux from calibration sources (usually stars)
with different spectra than science sources can
only approximately determine actual science
flux. This problem is made worse when sources of
different redshifts are compared, since the
correction will vary between sources. We propose
an improved method, using a direct, in-situ
measurement of the spectrally-resolved system
response function.
4
K-Corrections are only a Partial Solution

• K-corrections attempt to correct for these errors
  using an estimate of the system response function
• The estimate does not take into account
  distortion by the telescope optics
• To incorporate this effect a measurement of total
  system response function is necessary

5
Laser Photometric Calibration System Design
6
Error Determination

• Single-source flux at pixel i
• Synthetic photometry constructed from measured
  response function
• Illumination-corrected dome flat
• Synthetic dome flat constructed from measured
  response function

7
Determination of Error from Response Function
Measurements

• Scatter in ratio of illumination-corrected to
  synthetic dome flats is a measure of error due to
  non-uniform illumination of the flatfield screen
• Error gt 1, improvements to illumination design
  are necessary
• We have assumed illumination corrections are
  perfect

8
Determination of Error from Conventional Data
Reduction

• Scatter in synthetic photometry
• normalized with synthetic dome flat
• is a measure of error due to
• conventional data reduction
• Error is sub-percent
• Synthetic photometry was performed
• on a Nugent template of a 20-day-old
• SNIa and an atmospheric model,
• scatter did not vary significantly with
• redshift or airmass.

Nugent, Kim and Perlmutter, K-Corrections and
Extinction Corrections for Type Ia Supernovae,
PASP (2002), 114, 803.
9
Determination of Systematic Error due to Optical
Distortion

• K-corrections do not incorporate
• optical distortion of the filter function
• Synthetic photometry performed
• with and without distortion show little
• difference
• Integrated flux varies by 0.06,
• a negligible source of error

10
Conclusions

• Current design provides insufficient uniformity
  in illumination of flat-field screen
• For the system and source in question,
  K-corrections eliminate all but 0.06 of error
• Similar systems will be more useful for
  comparison of data between telescopes, where
  larger systematic errors are likely to occur