Gindex

SYSTEM DESIGN NOTE

SDN00005 - Baseline Observing Scenario

Prepared by	Date	Approved by	Date	Rev.	Rev Date
Jay Elias	3/18/99	N.Gaughan	3/19/99	a	4/7/99

1. Introduction

Observing scenarios are discussed in greater detail in the Operational Concept Definition Document. The purpose of this design note is to outline a baseline scenario as it affects the instrument design.

2. Preparation

Some calibrations of the instrument will be carried out prior to the start of observations. There are likely to be relatively few of these – typically observation of a flat-field lamp and a spectral lamp in one or more instrument configurations. Exposures are likely to be short, although individual images may be co-added to improve signal to noise; total exposure times are still likely to be a few minutes at most.

Dark exposures may also be done. These can be lengthy, since they may be as long as the longest exposures on faint objects, possibly up to an hour. Multiple darks would be done to obtain a low noise “mean dark”, so several hours could be expended in this way. Note that darks are done blocking the light internally, since the external environmental cover emits at longer wavelengths.

3. Observation

Although actual observing programs will vary from one another, a representative situation is one in which an object and a standard are each acquired and observed during the course of an hour. The instrument will thus need to switch between acquisition and spectroscopy a total of four times during the hour, but the spectroscopic configuration (grating selection and tilt, cross-dispersion, and camera selection) will not change. Note that it is assumed that some kind of acquisition can always be done through the camera required for spectroscopy.

3a. Acquisition

Acquisition consists of positioning of the OIWFS gimbal mirror to a pre-selected location, insertion of the acquisition (also known as “flip-in,” “bypass” or “slide-in”) mirror, inserting a wide slit and (and possibly a wider decker), and possibly changing the filter. This can happen while the telescope is moving to the new position. Once the telescope and instrument are correctly configured and positioned, the guide star for the OIWFS is acquired and the OIWFS starts guiding and sets telescope focus. Then, an exposure (usually fairly short) is taken to see where in the acquisition field the object is located¹. Normally the object will not be exactly centered, and it will be necessary to adjust its location by adjusting the gimbal mirror tilt and moving the telescope. If the slit and acquisition mirror are highly repeatable, the position corresponding to the slit used for spectroscopy will be accurately known, and one could in principle move the object to the correct position and reconfigure to spectroscopic mode without further exposures. If repeatability is less precise, or the observer wants further assurance that things are really optimized, he/she will take one or two more exposures as the telescope and OIWFS are moved or the slit is re-positioned.

The main difference between the fast acquisition scenario and the slow acquisition scenario is that it the first case a single acquisition image is taken and the instrument is then fully reconfigured for spectroscopy; the mechanisms (and telescope) are moved in parallel. In the slow scenario, the mechanisms are moved serially and a new image is taken for each step. It may be necessary to tweak the OIWFS gimbal mirror, but apart from this the number of mechanism motions is the same as for the fast scenario. However, the time spent is significantly longer, because the mechanisms are moved serially and because multiple images must be taken, processed, and examined (to determine offsets).

For faint objects, whose observation requires more than one hour of total observation, “acquisition” is essentially recentering or checking centering. In this case, the OIWFS will already have acquired the guide star and will be guiding on it. (One would probably not observe the standard star every hour in this case.)

Although the OIWFS filter and focus may change, neither is likely to change often – the same filter is likely to serve most programs, and focus will vary at worst on a time scale of several hours, if not longer.

3b. Observation

Once the object has been acquired, the instrument is reconfigured to spectroscopic mode. For several reasons, the object will be observed at more than one location on the spectrograph slit. One reason is that this allows one to compensate for the occasional bad pixel on the detector. It also averages out flat fielding uncertainties and provides better sky subtraction. The fact that multiple exposures are taken also assists in removing cosmic rays. Normally, the standard star is observed in the same way as the object for consistency, though with much shorter exposures.

The number of different positions depends on the configuration and the faintness of the object. Ideally, one would like a moderately large number – 5 or so – but this may not be necessary if the object is bright, and it may not be possible if the slit length is short compared with the object or image size.

In the case of large objects (usually in the case of the IFU), it may not be possible to observe the object at more than one position. In these cases, the spectrograph will be offset from the object to a nominally empty sky position nearby. For lengthy observations the observations will alternate between object and sky, and a different sky position will be used each time. Only experience with the instrument on the telescope will show whether it is necessary to offset the OIWFS when the telescope moves to a sky position, then offsetting back when it returns to the object, or whether it is enough to turn off the OIWFS when the telescope moves to the sky position, and simply reacquire the object with the OIWFS when the telescope returns to the object.

3c. Major Reconfigurations

The frequency of major reconfigurations – changes in the spectroscopic configuration – depends very much on the scientific program. At one extreme, there will be programs where the same configuration is used throughout the night. This might be the case where one is getting cross-dispersed, low-resolution spectra of a set of objects. At the other extreme, one might be doing high-resolution spectra of a particular emission line for a set of objects at a range of redshifts. In this case, the grating, filters, and sometimes the camera would change from one set of observations to the next.

4. Lifetime Usage

Approximate usage of the instrument can be calculated as follows:

The GNIRS is defined as having a ten-year lifetime. The instrument will not be used every night during the ten years, though. If one argues that it will usually be resident on the telescope, and that it occupies one of three science ports, then one could argue that it would plausibly be used 1/3 of the time. Gemini will have more than three instruments per telescope, so perhaps usage will be slightly less.

The average night is taken to be 11 hours. Not all nights are usable; if one assumes the usable fraction of time is 80% for IR spectroscopy (probably rather optimistic, though this is the fraction of time during which observations are attempted; bad data still exercise the instrument’s functions).

The total number of hours used is then:

3650 x 1/3 x 11 x 0.8 = 10707 hours.

The scenarios above indicate the mechanism fall into the following ranges for “lifetime use”:

>10⁵ motions/lifetime

OIWFS gimbal mirror

10⁴ – 10⁵ motions/lifetime

slit mechanism
decker mechanism (at low end)
filter wheels
flip-in (slide-in) mirror

<10⁴ motions/lifetime

environmental cover
OIWFS filter wheel
OIWFS focus (assume refocus every few hours)
grating turret
prism turret
camera turret
detector focus

¹ It is assumed that it is not necessary to take a second exposure at an offset position in order to subtract sky – either because the object is bright enough that this is not required, or because an archival image is available that is good enough for this purpose. For the faintest objects, this may not be the case and one may need to deliberately offset the telescope and take a second exposure.