Change page style: 

Baseline Calibrations

For all queue observations, a set of standard calibrations (the "baseline calibrations") will be taken by Gemini Staff to ensure the long-term utility of data in the archive. The baseline calibration set varies from instrument to instrument and from mode to mode. The table below describes the GSAOI baseline calibrations, as well as provides information on how and when to request additional frames.

All GSAOI baseline calibrations are taken from the GS-CALYYYYMMDD (shared) program, and not from the individual science programs. This means the users need to search for the calibrations separatelly when downloading their science data from the GSA. Instructions/tutorials on the procedures can be found in the "GSA Search and Download Instructions" page.

GSAOI Baseline Calibrations

Measurement Type Notes
Bad pixel mask Derived each semester from GCAL H band flats and dark images. The GSAOI detectors have less than optimal cosmetic quality, due to the presence of "dead" pixels, light-emitting pixels, "killed" spots and a significant number of hot pixels. All  of these are found to be quite stable, but the number of hot pixels detected (for a given cut level above the mean value in the dark frame), is dependent on the exposure time of the dark frames used. The standard BPM included in the Gemini GSAOI IRAF package is defined using 150sec darks. If desired, the applicants can require to take one set of 10-15 frames (darks + flats) per semester using darks with exposure times/filter tailored to their program in order to construct a BPM. Flats can be either GCAL or domeflats. A document describing the process followed to derive the standard BPM is now available.
Flat field

The baseline flat fields for GSAOI are taken as dome flats for all broad band and narrow band filters. The current calibration plan is to obtain two sets of 10 flats per filter per run (at the start and at the end of the laser block). Pending analysis of the flat stability, this can be reduced to one set per run (the data so far indicate that flats are quite stable on periods of months). OFF flats are taken for the H and K band filters.


GCAL flats can be obtained by request in all filters but H, K, Ks, Kp and CO 2.36mu - GCAL flats in these five filters are rendered useless due to structure  originated in the GCAL ND filter.  Users are reminded that GCAL flats do not correspond to the same optical path as the science frames (the GCAL light does not pass through the GeMS AO bench).

Twilight flats in the broad band filters only (Z, J, H, K, Ks, Kp) will be taken as often as possible (we aim at a complete set per laser run). These will be evening twilights. We do not plan to take twilights in the narrow band filters as part of routine baseline calibrations, as these do not show any significant difference against the domeflats.

All flats are taken using the Bright Object read mode. 

Fringe Frames Not applicable
Dark / Bias

The dark current for GSAOI is of the order of 0.01 e-/s/pixel, implying on a total dark charge of ~12e- for the longest integration time allowed (20min). This very low dark current implies that daily darks taken with the same exposure time as the science frames (for dark subtraction as part of the standard reduction process) are NOT necessary, and in fact, subtracting a dark frame from the science and sky frames prior to subtracting the sky from the science proper will result in increasing noise. Both the dark current and the cross-shaped "warm" pixels are accounted for when subtracting the sky.

Darks (or rather "flats-off") are not really required for Dome or Twilight Flats - this was tested during commissioning and attempting to subtract darks or flats-off from twilight and dome frames, respectively, have not produced any measurable improvement in the quality of the resulting flat-field. By request of users, however, domeflats OFF are taken for the filters in the H and K bands. GCAL flats-off can be obtained as part of the GCAL flat sequence, if such is required (but users are reminded that GCAL flats are sub-optimal for data reduction).

If the applicant deems that darks are required for the data reduction, a daytime sequence can be added to the PhaseII and will be taken on a best effort basis within a few days of the observation execution (not charged to the programme).

Given the low dark current and the temporal behaviour of the dark frames, current plans are to create a library of dark frames with different exposure times, which can be downloaded from the GSA and used for data reduction if necessary.

Wavelength calibration Not applicable
Telluric standard star Not applicable
Flux standard stars

Selected from the Persson (1988) list. Observations of standard stars are only obtained in broadband filters J, H, K, Ks, Kp. A more limited set of standards include the Z filter. Our goal is to obtain at least 2-3 standards at different airmasses each observing night. Approximate zero points are also available, and an study of the color effects originated from the differences in QE between the four arrays is ongoing. Photometric accuracy is limited to 5-10% due to uncertainties in the atmospheric extinction over Cerro Pachon, flat-fielding and sky subtraction, and the precision in the gain determination for each of the arrays. If the user is aiming at a higher photometric precision, it is necessary to self-calibrate the images, detector by detector.

Users should be aware that the standard star observations are NOT performed with the GeMS loops closed (NO AO correction - natural seeing only).  If the applicants would like to obtain the flux calibration stars with the AO loops closed, these should be provided as separate targets in PhaseI and included as Program Standards in PhaseII, therefore incurring in charged time to the program. The same procedure should be followed if the program requires a specific calibrator to be observed (e.g., for narrow band filters). Users should be careful to select standards which will NOT saturate in the fully corrected images, and check if the guide stars can be mantained for all offset positions. Please remember that the ODGW guide stars cannot be moved from one detector to another within the same sequence.

Atmospheric extinction Not included as part of the baseline calibration set.
Point Spread Function star Not included as part of the baseline calibration set. Applicants are reminded that the PSF in an MCAO-corrected image is dependent on the position in the field AND on the geometry of the guide star constelation (plus higher order terms such as laser power and return, relative brightness of the guide stars, etc). There is probably little scientific gain in attempting to derive a PSF from a separate star field.
World coordinate system Automatically included for each GSAOI image. Relative accuracy is currently ~0.3arcsec on average over the field. Absolute accuracy is around 5arcsec (limited by the precision in the telescope pointing). 
Focal plane mask image Not applicable.