|
|
AUSTRALIAN
NATIONAL UNIVERSITY System Design Note 4.04 Created: 6 April 2000 Last modified: 10 April 2000 |
NIFS CALIBRATION REQUIREMENTS
Ian Price
Research School of Astronomy
and Astrophysics
Institute of Advanced
Studies
Australian National
University
Revision History
|
Revision No. |
Author & Date |
Approval & Date |
Description |
|
Revision 1 |
Ian Price 25 October 1999 |
Peter J. McGregor 06 April 2000 |
Original document. |
|
Revision 2 |
Peter J. McGregor 06 April 2000 |
Jan van Harmelen 07 April 2000 |
Reformatted for Word 2000. Modified following CoDR. |
|
|
|
|
|
Contents
4.1.4 Wavelength
Calibration Frames
4.1.8 Atmospheric
Dispersion Effects
4.1.9 Non-Common
Path Phase Errors
1 Purpose
The purpose of this document is to describe the calibration procedures that will be required to obtain scientifically useful data from the Gemini Near-infrared Integral-Field Spectrograph (NIFS).
2 Applicable Documents
|
Document
ID |
Source |
Title |
|
RSAA |
NIFS Performance Model |
|
|
RSAA |
Atmospheric Dispersion Effects in NIFS |
|
|
|
|
|
|
|
|
|
3 Introduction
The extraction of scientifically useful data from raw data frames obtained with the science detector of the Gemini Near-infrared Integral-Field Spectrograph (NIFS) will require knowledge of the spatial and spectral mapping from the sky to the detector image plane, as well as the instrumental intensity transformation. The observations that will be required to determine these mappings are described in this document. The goal is to produce a fully rectified 3D data cube from the 2D spectral image recorded by the science detector, with uniform sampling in both the spatial and spectral directions.
4 Calibrations
NIFS data frames will require bias subtraction, dark frame subtraction, flatfielding, wavelength calibration, correction for geometrical distortion, flux calibration, and correction for absorption in the Earth’s atmosphere. A related, but different, calibration problem will be to determine the PSF appropriate to each science observation. NIFS will use continuum and emission-line lamps in the Gemini Calibration Unit (GCAL) for flatfield and wavelength calibration, respectively, using the near-infrared diffuser in GCAL. The lamp intensities have been specified for NIRI and GNIRS. The pixel scale and spectral resolution of NIFS are similar to modes of GNIRS, so the GCAL lamp intensities should also suit NIFS.
4.1.1 Bias Frames
Bias frames are needed to determine the DC electrical offsets for each detector pixel. NIFS will include a blanked off position in the filter wheel that will exclude external light from the remainder of the optical train. Bias frames will be recorded with this blank in place using the minimum possible exposure time (~ 5 s). Zero length exposures are not possible due to the readout architecture.
4.1.2 Dark Frames
Dark frames are used to remove the signal component due to spontaneously generated charge and background light from within the cryostat that does not scale in the same way as light entering from outside the cryostat. Dark frames will be recorded using the filter wheel blank as for bias frames.
4.1.3 Flatfield Frames
Throughput variations and pixel-to-pixel quantum efficiency and gain variations cause pixel-to-pixel signal variations in response to uniform illumination. NIFS will use continuum lamps in GCAL for flatfield calibration. The beam from GCAL is injected below ALTAIR, so flatfield frames obtained in this way do not allow for throughput variations within ALTAIR. This should not be a problem given the small field-of-view of NIFS. Spectroscopic flatfield frames are strictly needed only for the removal of pixel-to-pixel sensitivity variations. Large scale sensitivity variations in the spectral direction are calibrated using measurements of flux standard stars. Large scale sensitivity variations in the spatial direction can be calibrated using measurements of twilight sky spectra.
4.1.4 Wavelength Calibration Frames
Arc lamp spectra recorded during daylight will provide the primary transformation of NIFS spectra to known linear wavelength scales. NIFS will use the near-infrared arc lamps in GCAL as its wavelength calibration source. The required range of lamps and intensities will be similar to those needed for modes of GNIRS. The need to move M3 in the Instrument Support Structure to inject the GCAL beam could potentially compromise the positional repeatability of NIFS science observations. If this proves to be problematic, it will be possible to use sky spectra containing OH airglow emission to track wavelength zero point shifts during a night.
4.1.5 Geometric Distortion
NIFS will form a reformatted slit image that will be stepped because of the image slicing function of the IFU, and be distorted by spectral line curvature induced by the grating as well as other optical aberrations. Geometrical distortions in the spectral direction will be calibrated using arc lamp exposures as part of the wavelength calibration. Geometrical distortions in the spatial direction must be treated separately. It is not feasible to record stellar spectra at several positions along each slitlet in a manner analogous to long-slit spectroscopy because of the time involved and the need to perform this calibration during daytime. NIFS will use a multislit mask in the Focal Plane Mask Wheel in combination with the flatfield lamp in GCAL to produce multiple artificial star spectra that will be used to calibrate spatial distortions. The multislit mask will be oriented perpendicular to the IFU slitlets so it will illuminate an array of points along all of the IFU slitlets simultaneously. One central artificial star would be sufficient to trace the curvature (i.e., offset) of each IFU spectrum on the detector. Two artificial stars measure offset and linear stretch. More artificial stars allow higher order distortions to be measured, although we expect that offset and linear stretch will be sufficient. The NIFS field-of-view is ~ 2´2 mm at the Focal Plane Mask Wheel. NIFS will use a Ronchi grating with 200 mm pitch rulings which will produce 10 artificial star images of ~ 100 mm (~ 0.16² º 4 pixels) width along each IFU slitlet image.
4.1.6 Flux Calibration
Measurements of stellar flux standards will be used for absolute flux calibration of NIFS spectra. Flux standards should be measured close in time and position on the sky to science observations to minimize effects due to atmospheric transparency variations. Suitable flux standards will be drawn from lists of photometric standard stars.
4.1.7 Terrestrial Absorption
Near-infrared spectra suffer strong, time-variable, wavelength-dependent absorption in the Earth’s atmosphere. Systematic errors in correcting for this absorption can dominate statistical errors, and are likely to be particularly significant for the long (~ 1 hr) exposures required to overcome dark current and read-noise with the high resolution, short wavelength NIFS gratings. Smooth spectrum stars must be measured close in time and position on the sky to science observations to achieve accurate correction. Dwarf stars of F and G spectral type have the weakest spectral features and are the most common and widely distributed smooth spectrum stars over the sky. Intrinsic absorption in the smooth spectrum, solar-analog stars can be modeled in the manner described by Maiolino, Rieke, & Rieke (1996). Smooth spectrum stars are commonly drawn from the Bright Star Catalog. Most of these stars are brighter than the K ~ 6.4 mag limit for measurement with NIFS (NIFS Performance Model, SDN0004.01). It will therefore be necessary to define alternative lists of smooth spectrum, solar-analog stars for NIFS and, presumably, also for GNIRS.
4.1.8 Atmospheric Dispersion Effects
Atmospheric dispersion will produce a wavelength dependent position offset on the NIFS science detector. This effect has been considered in SDN0005.12 (Atmospheric Dispersion Effects in NIFS) and found to be small. If deemed important, the IFU slitlets can be oriented at the mean parallactic angle so that atmospheric dispersion smears images along slitlets. By modeling the effects of atmospheric dispersion, the spatial map for each slitlet in the detector plane can be adjusted as a function of wavelength so that the same region of sky is extracted at all wavelengths. Such observations should be performed away from the Zenith to minimize field rotation.
4.1.9 Non-Common Path Phase Errors
Phases errors generated in the non-common path between ALTAIR and NIFS will be determined so that ALTAIR can apply appropriate compensation to the corrected wavefront. The non-common path phase errors will be recorded during daylight by recording in-focus and out-of-focus images of an artificial star generated by ALTAIR. It is proposed that NIFS will not have a detector focus stage. ALTAIR is able to defocus the star image. Vignetting of the defocussed beams caused by the NIFS IFU requires further investigation.