5 Calibrations at Commissioning

 

Prior to and during commissioning, various calibrations must be performed before the instrument is ready to produce science data that can be readily interpreted. These calibrations concern the geometrical mappings between coordinate systems described in Chapter 2. They are:

Slitmask Astrometry, or mapping the sky (RSCS) onto the slitmask (SMCS);

TV Astrometry, or mapping the sky (RSCS) onto the TV system (TVCS); and

CCD alignment, or measuring the rotations and translations of the individual CCDs in the mosaic (CCDCSi) relative to the Image Coordinate System (ICS).

This does not describe all mappings. Certain mappings (FKCS to RSCS) are handled purely through modelling. Others will be handled during observing (SMCS to MFCS) or determined empirically during science calibrations (SM/MFCS to ICS; ICS to RSCS).

5.1 Background and Issues of Concern

(nothing major)

5.2 Database Products

The slitmask astrometry mapping will be maintained in the database.

The TV astrometry mapping will be maintained in the database.

The CCDCSi ICS mapping will be maintained in the database.

NOTES -- Delivery: 0=Commissioning, 1=Pipeline.

Required Hardware:

-

Mask form fiducial marks

-

Slitmask of astrometric field, based on expected projections.

-

Grid-of-holes (GOH) mask

-

Line-of-holes (LOH) mask

5.2.1 To create a slitmask astrometry map:

1.

Select astrometric field; prepare a slitmask using existing slitmask mapping (pure model initially). This slitmask will have small pinholes at the expected position of selected stars.

2.

Align mask with the selected field; take one or more images of the mask.

3.

Remove mask; take one or more images of the field; process all images through relative flux calibration.

4.

Measure the pixel positions of each selected star and its corresponding pinhole.

5.

Convert the differences between each coordinate pair into mm on the slitmask. (Detailed mapping not necessary in this conversion.) Fit the mm offsets as a function of slitmask location. This becomes the differential update to the slitmask astrometry model.

5.2.2 To determine SM/MFCS to TVCS mapping:

1.

Take TV exposure showing mask form fiducials and fixed mirror.

2.

Analyse location of maskform fiducials to solve for scale, rotation and translation of TVCS wrt MFCS.

3.

Use slitmask astrometry mapping to relate this to the refracted sky.

4.

Mapping can be verified against an astrometric field.

Note: the above assumes negligible distortion in the TV camera. If this is not valid, observations of astrometric fields must be used to map out the distortion. The level of distortion can be tested (pre-commissioning) in the lab, using the image of a grid-of-holes mask. The level of telescope distortion (also a potential factor) is ?? (negligible) over the FOV of the TV.

5.2.3 To determine relative alignment of CCDs:

1.

Take a direct image (no slitmask) of flat field. Measure the location of the mask form fiducials. Using the known location of the fiducials wrt the telescope/collimator optical axis, solve for the rotation, scale and approximate zeropoint offsets for the fiducial CCD.

2.

Take a flat field image through the grid-of-hole (GOH) mask. Measure the locations of the holes on each mask and solve for a rotation and translation relative to the fiducial CCD. Convert to ICS using the transformation above.

3.

Take a spectral flat-field exposure through the line-of-holes (LOH) mask. Measure the spectral traces across the CCDs. The change in slope characterizes the red vs blue CCD rotation. This rotation can be related to the fiducial CCD and thus the ICS.

4.

Take a spectral arc exposure through the line-of-holes (LOH) mask. Using the traces above, locate and measure the emission line (spots) in the spectrum from each hole. These may be used to determine a , across each CCD boundary. In conjunction with the angle, these deltas provide the translations for the red CCD relative to the blue CCDs, which in turn are known relative to the ICS.

5.3 Other Products

5.4 Outstanding Issues

We have not addressed the FCCSj yet.