You are in: Observing Tool (OT) > Science Program > Elements > OT Components and Iterators > OT GMOS > tips

GMOS in the Observing Tool - Tips & Tricks

This page contains various tips and tricks for using the OT to configure GMOS observations. It also contains an overview over the required OT observations and components for programs of different types. Before submitting your Phase II program to Gemini, use the GMOS checklist and the general Phase II checklist to eliminate common mistakes.

Skeleton Phase II

The distributed skeleton (Phase II) science programs do not contain any "observes". You have to add a minimum of one "observe" to each of the defined observations. Without an "observe" no exposure will in fact be taken.

The distributed skeletons do not contain the correct GMOS configurations. All skeletons contain a GMOS component with GMOS in imaging mode and with no filter as a placeholder. This configuration is not a valid configuration. You will have to select the correct filter for imaging observations, while for spectroscopy observations you as a minimum need to select the disperser, the wavelength and the focal plane unit.

Required OT observations and components

For each type of program, there are different requirements to which OT observations and components the user has to define in the Phase II. The table below gives an overview over the requirements. "Required" means that a Phase II will not be accepted without this type of observation or component being defined. "As needed" means that the user can add this type of observation or component as needed by the science goals of the program. It is also recommended to use the GMOS OT library as a source of example observations.

CuAr arcs mixed with the science data, i.e. taken at night, and any special standard stars are charged towards the program's allocated time. Baseline standard stars, GCAL flats, twilight flats, baseline CuAr arcs, and mask images are not charged. The time for the acquisition observations is already included in the science observation overhead. No additional time is charged. However, the acquisition observations have to be defined as separate observations in the Phase II.

Each "Observe" command is given an Observe Class (see more details about Classes). This replaces the Calibration checkbox used in 2005A and allows for easier and more precise time accounting, data distribution, and queue planning. All new references to the class are in boldface in the descriptions below.

The details of each of the required observations and components are explained in the sections below.

Observation or Component	Program Type				Observe Class
Observation or Component	Imaging (details) Pre-imaging for MOS	MOS (details)	Longslit (details)	IFU (details)	Observe Class
Science Obs.	Required	Required	Required	Required	Science
Offset Comp.	As needed	As needed	As needed	As needed	N/A
GCAL flats	Do not add	Required Mix w/ science	Required Mix w/ science	Required Mix w/ science	Nighttime Partner Calibration
Twilight flats	Do not add	As needed Separate observation	As needed Separate observation	Required Separate observation	Daytime Calibration
CuAr arcs	N/A	Required Baseline: Separate observation	Required Baseline: Separate observation	Required Baseline: Separate observation	Daytime Calibration
CuAr arcs	N/A	Charged if mixed w/ science	Charged if mixed w/ science	Charged if mixed w/ science	Nighttime Program Calibration
Acquisition Obs.	N/A	Required Separate observation	Required Separate observation	Required Separate observation	Acquisition
Acquisition Obs.	N/A	If for baseline standard	If for baseline standard	If for baseline standard	Acquisition Calibration
Mask image	N/A	Required Separate observation	N/A	N/A	Daytime Calibration
Baseline standard stars	Do not add	Required Separate observation	Required Separate observation	Required Separate observation	Nighttime Partner Calibration
Nod & Shuffle Darks	N/A	Recommended for Nod & Shuffle only Separate observation	Recommended for Nod & Shuffle only Separate observation	Recommended for Nod & Shuffle only Separate observation	Daytime Calibration
Special standard stars Charged	As needed Separate imaging observation	As needed Separate longslit observation Follow requirements for longslit	As needed Separate longslit observation Follow requirements for longslit	As needed Separate IFU observation Follow requirements for IFU	Nighttime Program Calibration

Calculation of overheads

The current version of the OT assumes an acquisition time of 10 min for imaging and 20 min for all spectroscopic observations. Since 2006A, the overheads associated with each exposure (readout and Nod & Shuffle overheads) in the OT have been improved to more accurately reflect the true execution time. Detailed information about overheads for acquisitions as well as readout and configuration times can be found on the GMOS overheads page.

PIs can now use the calculated planned execution times in the OT as reasonable approximations of the actual time that will be required. The only additional overhead you will need to consider when you plan your observations is for reacquisition. As a rule of thumb you should consider another 20 min of acquisition time for every chunk of spectroscopy longer than about 3 hours. If you do not take this into consideration, it is likely that your Phase II will be overfull. Please note that Gemini queue observers will stop executing your program when the allocated time has been depleted, regardless of whether or not there are still unexecuted observations.

If your observation classes are set correctly, the OT will not add any planned time for acquisition observations, GCAL flats, twilight flats, baseline CuAr arcs or other daytime calibrations (eg mask images or Nod & Shuffle Darks). The support staff from your National Gemini Office and your Gemini Contact Scientist will work with you if you have questions about your OT calculated program execution time.

Grating choices

All of the gratings included in the OT are available for science use. Only three of the gratings can be mounted in GMOS simultaneously. Grating changes will not be done during the night.

Program Organization

It is recommended that groups be used as much as possible to keep the Phase II organized. Some recommended organizational practices are:

Group science observations with their associated acquisition observation(s).
Organize all the time for long observations (more than 2 hours) into a single observation.
If a standard observation must be taken before or after a science observation (e.g. a telluric standard) then place the observations for the standard in the same group as the science observations.
Place all daytime calibrations in a group called Calib.
Place time constraints (dates/times, temporal spacing of observations) within the new Scheduling Note.
Group names within a program should be unique.

Imaging observations

If full-frame readout is being used then an offset iterator should be used to define at least two offset positions separated by at least 5 arcseconds in p in order to fill in the chip gaps. The first offset position should always have p=0, q=0.

Twilight flats are baseline calibration and are handled by Gemini staff. Twilight or GCAL imaging flat observations should not be included in the Phase II.

Observations in the i'-band or redder are susceptible to fringing (GN fringing, GS fringing). Blank sky flats are taken every few months with GMOS-S and are baseline calibration but they are not taken for GMOS-N because the problem is less severe for the GMOS-N CCDs. If fringe frames cannot be derived from the science observations or the baseline calibration is not sufficient then blank sky observations need to be defined. The class should be 'Nighttime Program Calibration' and the time will be charged to the program. These observations should be like the science observations but the target component should be blank or removed. Gemini staff will add the appropriate blank sky field.

If iterating over filters some of which will have fringing and some of which will not, then separate the two types of filters into different GMOS sequences. See the GMOS OT libraries for examples. This makes it possible to use the On-line Data Processing System (OLDP) to process all the data in real time. This will speed the distribution of MOS pre-images.

MOS observations

Programs that contain MOS observations need pre-imaging for all fields. The pre-imaging will be obtained 3-4 weeks before the MOS observations. Thus, PIs with MOS programs are encouraged to meet the first deadline of the semester for submitting their programs.

MOS programs should contain the final Phase II information for the pre-imaging observations when it is first submitted. The MOS observations should also be included in the program. Small adjustments to the MOS observations are allowed when the mask design has been done. However, the pointing and PA of your target cannot be changed between the pre-imaging and the MOS observations. PIs with MOS programs will be contacted by Gemini when the pre-imaging data is available. Revised Phase II MOS programs should be submitted to your National Gemini Office support staff by the GMOS mask deadlines. Mask designs should be submitted directly to Gemini staff by the same deadlines, instructions will be sent to you when your pre-imaging is available.

PIs with pre-imaging from previous semesters should contact their Gemini Contact Scientist for information on how to submit their mask designs. Submission of Phase II should follow the normal procedure. The deadlines listed above apply, though no "first" submission of a Phase II will be accepted after normal deadline for Phase II in August 15, 2003.

The focal plane unit for the MOS observations should be specified as Custom Mask MDF, and the field for the name should contain your program ID and a running number for the mask within the program, e.g. for program GN-2003B-Q-27 the mask names should be GN2003BQ027-01 and GN2002BQ027-02 for the first and the 2nd mask, respectively. Note the leading zero on the program ID and mask numbers.

The total time used for both pre-imaging and MOS observations must not exceed the allocated time.

It is recommended that pre-imaging observations are dithered by 5 arcsec in both directions, e.g. 4 exposures in a square pattern with size 5 arcsec will work ok. (Examples are available in the GMOS OT library). Pre-imaging exposures should be taken in the broad band filter closest to the central wavelength coverage of the MOS observations.

GCAL flats should be mixed with your science MOS observations. For long observations, add one flat for every 1-2 hours of science exposure time. For observations shorter than one hour, add one flat. Make sure you get flats for all spectral configurations. GCAL flats within these guidelines have class 'Nighttime Partner Calibration' and are not charged to the program. Any additional GCAL flats should have class 'Nighttime Program Calibration' and will be charged to the program. Refer to the tables of recommended GCAL configurations and exposure times. Examples are available in the GMOS OT Library.

CuAr arcs taken during the day are baseline calibrations. These calibrations are not charged to the program, but the PI has to define them in the Phase II. Define these observations as separate from the science observations and set the class to 'Daytime Calibration'. Make sure the instrument configuration matches the science observation. If you copy the science observation in order to edit it for the CuAr arc, make sure to remove the guide star from the target component, remove any science exposures, and change the class. Arcs taken as part of of a science MOS observations are should have class 'Nighttime Program Calibration' and will be charged to the program. Refer to GCAL configurations for recommended GCAL configurations and exposure times (examples are available in the GMOS OT Library).

Mask images are taken of all MOS masks before they are used for science. These calibrations are not charged to the program, but the PI has to define them in the Phase II. Define these observations as separate from the science observations. The instrument configuration should be as follows:

   No guide star
   GMOS r-filter
   Detector: Bin 1x1, Fast read, Full frame readout
   Translation stage: 'Do Not Follow' or 'Follow in Z Only' (GMOS-S only)
   Class: Daytime Calibration
   Define a GCAL flat using the appropriate information from the following GCAL configurations.

For a complete example see the GMOS OT Library.

Acquisition observations need to be defined for each MOS mask. No extra time is charged for these observations, as the overhead for setting up is already included in the science observation. However, the acquisition observations should be defined as separate observations in the Phase II. An acquisition observation for a MOS mask should have the following instrument configuration:

   Same target, guide star and PA as for the science observation
   GMOS filter closest to the wavelength setting used 
   FPU should be Custom MDF mask and the same mask as for the science observation
   Exposure time: 30-90 sec depending on the brightness of the acquisition stars
   Detector: Bin 1x1, Fast read, Full frame readout
   Add an "observe", and edit it to show 4 exposures with class Acquisition

Long-pass filters cannot be used for MOS acquisitions. Narrow band filters should in general not be used for MOS acquisitions. See also the example in the GMOS OT Library.

Twilight flatfields are taken for each MOS mask if requested. No extra time is charged for these observations. However, the twilight flatfields should be defined as separate observations in the Phase II. Make sure the instrument configuration matches the science observation. If you copy the science observation in order to edit it for the twilight flatfield, make sure to remove the guide star from the target component, then edit the science exposures to have only one exposure with 30sec exposure time. The class should be 'Daytime Calibration.' If doing small wavelength dithers to fill in the chip gaps then only one wavelength setting will have a twilight flatfield taken, so you need to select the wavelength at which you want the twilight flatfield. If there are no twilight flats in the submitted Phase II for a MOS program, it will be assumed that they are not needed. See also the example in the GMOS OT Library.

Longslit observations

GCAL flats should be mixed with your science observations. For long observations, add one flat for every 1-2 hours of science exposure time. For observations shorter than one hour, add one flat. Make sure you get flats for all spectral configurations. GCAL flats within these guidelines have class 'Nighttime Partner Calibration' and are not charged to the program. Any additional GCAL flats should have class 'Nighttime Program Calibration' will be charged to the program. Refer to the GCAL configurations for recommended GCAL configurations and exposure times (examples are available in the GMOS OT Library).

CuAr arcs taken during the day as baseline calibrations. These calibrations are not charged to the program, but the PI has to define them in the Phase II. Define these observations as separate from the science observations and set the class to 'Daytime Calibration'. Make sure the instrument configuration matches the science observation. If you copy the science observation in order to edit it for the CuAr arc, make sure to remove the guide star from the target component, remove any science exposures, and change the class. Arcs taken as part of of a science longslit observations are should have class 'Nighttime Program Calibration' and will be charged to the program. Refer to the GCAL configurations for recommended GCAL configurations and exposure times (examples are available in the GMOS OT Library).

Acquisition observations need to be defined for each longslit target. No extra time is charged for these observations, as the overhead for setting up is already included in the science observation. However, the acquisition observations should be defined as separate observations in the Phase II. New for semester 2007A: to improve the observing efficiency, the longslit acquisition sequence has been modified. For Point Sources a ROI of Central Stamp (300x300 unbinned pixels) should be used to measure the slit center and to confirm if the science target is within the slit (steps 2 and 3). For Extended Objects, Double Sources and Off-axis Sources a ROI of Central Stamp should be used only to measure the slit center (step 2). An acquisition observation for a longslit target should have the following instrument configuration:

   Same target, guide star and PA as for the science observation
   GMOS filter closest to the wavelength setting used
   Detector: Fast read, low gain
   1) For Point Source:
      Add a GMOS sequence with 3 steps
      Step1 - FPU: None; Detector: bin2x2; ROI: CCD2; Exposure time: 2-5 min 
              depending of the brigthness of the target.
      Step2 - FPU: Same longslit as for science; Detector: bin1x1, ROI: Central Stamp;
              Exposure: 10 sec. This step is used to measure the slit center.
      Step3 - FPU: Same longslit as for science; Detector: bin1x1, ROI: Central Stamp;
              Exposure: 2-5 minutes depending of the brigthness of the target. 
   2) For Extended Source, Double source, off-axis source:
      Add a GMOS sequence with 3 steps
      Step1 - FPU: None; Detector: bin2x2; ROI: CCD2; Exposure time: 2-5 min 
              depending of the brigthness of the target.
      Step2 - FPU: Same longslit as for science; Detector: bin1x1, ROI: Central Stamp;
              Exposure: 10 sec. This step is used to measure the slit center.
      Step3 - FPU: Same longslit as for science; Detector: bin1x1, ROI: CCD2;
              Exposure: 2-5 minutes depending of the brigthness of the target. 
   
   Class: Acquisition

Long-pass filters cannot be used for longslit acquisitions. Narrow band filters should only be used if your target is an emission line source with no continuum. For a complete example see GMOS OT Library.

Twilight flatfields are only taken for longslit observations of extended targets. No extra time is charged for these observations. However, the twilight flatfields should be defined as separate observations in the Phase II. Make sure the instrument configuration matches the science observation. If you copy the science observation in order to edit it for the twilight flatfield, make sure to remove the guide star from the target component, then edit the science exposures to have only one exposure with 30sec exposure time and class 'Daytime Calibration'. If doing small wavelength dithers to fill in the chip gaps then only one wavelength setting will have a twilight flatfield taken, so you need to select the wavelength at which you want the twilight flatfield. See also the example in the GMOS OT Library. If there are no twilight flats in the submitted Phase II for a longslit programs, it will be assumed that they are not needed.

IFU observations

IFU observations can be done either in "two-slit mode" or "one-slit mode". For GMOS observations in "one-slit mode" the "IFU Right slit (red)" should be chosen The change between the two modes is a manual process that involves taking the IFU out of the instrument. This will not be done during the night. GMOS is used for blocks of time in either mode, and changes between the modes are done infrequently.

The OT visualization of the GMOS IFU shows the larger of the two IFU fields at the target position. Thus, if switching between fpu=none and the IFU, the user will see the OIWFS field patrol field move -- this is as expected, since the larger of the two IFU fields is approximately 30 arcsec from the center of the imaging field of view.

In one-slit mode the central wavelength you specify will be interpreted as the desired wavelength at the center of the detector array. Your Gemini Contact Scientist will make the required adjustment to your Phase II to accomplish this. In two-slit mode, the central wavelength you specify will be the wavelength at the location of the two pseudo-slits.

In two-slit mode you will have to use one of the color filters in order to avoid overlap between the spectra, see the GMOS IFU page for details. In 1-slit mode you may have to use a filter to avoid 2nd order contamination.

GCAL flats should be mixed with your science observations. For long observations, add one flat for every 1-2 hours of science exposure time. For observations shorter than one hour, add one flat. Make sure you get flats for all spectral configurations. GCAL flats within these guidelines have class 'Nighttime Partner Calibration' and are not charged to the program. Any additional GCAL flats should have class 'Nighttime Program Calibration' and will be charged to the program. Refer to the GCAL configurations for recommended GCAL configurations and exposure times (examples are available in the GMOS OT Library).

Acquisition observations need to be defined for each IFU target. No extra time is charged for these observations, as the overhead for setting up is already included in the science observation. However, the acquisition observations should be defined as separate observations in the Phase II. An acquisition observation for an IFU target should have the following instrument configuration:

   Same target, guide star and PA as for the science observation
   GMOS filter closest to the wavelength setting used 
   Exposure time: 2-5min depending on the brightness of the target
   Detector: Fast readout, low gain
   Add a GMOS sequence with 2 steps
       FPU: None, Detector: bin2x2, readout CCD2
       FPU: IFU,  Detector: bin1x1, readout full frame
   Class: Acquisition

Longpass filters cannot be used for IFU acquisitions. Narrow band filters should only be used if your target is an emission line source with no continuum. See also the example in the GMOS OT Library.

Twilight flatfields are taken for IFU observations. No extra time is charged for these observations. However, the twilight flatfields should be defined as separate observations in the Phase II. Make sure the instrument configuration matches the science observation. If you copy the science observation in order to edit it for the twilight flatfield, make sure to remove the guide star from the target component, then edit the science exposures to have only one exposure with 90sec exposure time with class 'Daytime Calibration'. For 2-slit mode in the blue where there is no fringing then twilights will be taken in only one wavelength if you are doing small (~5nm) spectral dithers. However, in 1-slit mode or when there is fringing twilight flats have to be taken at all wavelength settings for now. See also the example in the GMOS OT Library.

Standard Stars

Imaging standards sufficient to obtain flux calibration at the 5% level are base calibrations and are taken by Gemini staff. If better calibration is needed then observations for additional standards must be included in the Phase II. The class should be 'Nighttime Program Calibration' and the time will be charged to the program.

Spectroscopic flux standards sufficient to determine the spectral response function, not absolute flux calibration, are baseline calibration and are not charged to the program. Starting in 2005B the observations for the baseline flux standard need to be defined in the Phase II. All flux standards are taken in longslit mode using the same slit as for longslit science observation or the longslit closest in width to the width of the MOS slitlets. These standards should have the observing conditions set to 'Any' and should have the target component deleted and the PA set to 0. The observer will fill in the target and the PA when taking the standard. The observations needed are:

Longslit acquisition with class 'Acquisition Calibration.' For an IFU acquisition, the offsets from the center of the CCD2 to the IFU-1 and IFU-2 fields are well known by the telescope observers and do not need to be defined. For semester 2007A: the longslit acquisition sequence for all standard stars (flux standard, velocity standards, etc) has been modified. A ROI of Central Stamp (300x300 unbinned pixels) should be used to image the field, to measure the slit center and to confirm if the target is within the slit. For GMOS-N, the IFU acquistion sequence for standard stars also has been modified by including a ROI of Central Stamp to image the field. See the examples included in the GMOS OT Libraries.
Spectroscopic observation with class 'Nighttime Partner Calibration'. Please select an exposure time long enough to obtain good signal-to-noise (i.e., 120 seconds). For longslit and IFU programs the instrument configuration should match the science observations. For MOS programs the instrument configurations should also match the science observations. For the longslit and the MOS programs, the spectroscopic observations should have the ROI defined as cspec. If the grating is not the R150 then a GMOS sequence should be used to define three wavelength settings that will bracket the expected wavelengths of the MOS observations. The first, second, and third setting are suggested in the next table. GCAL flats at each wavelength setting should be included. Examples are included in the GMOS OT Libraries.

Grating First Setting Second Setting Third Setting

R400 central wavelength - 150 nm central wavelength central wavelength + 150 nm

B/R600 central wavelength - 100 nm central wavelength central wavelength + 100 nm

R831 central wavelength - 80 nm central wavelength central wavelength + 80 nm

B1200 central wavelength - 30 nm central wavelength central wavelength + 30 nm

Daytime arcs with class 'Daytime Calibration'. All wavelength settings used by the nighttime observation should be included. All other guidelines for daytime arcs apply.

Additional spectroscopic standards --- absolute flux standards, velocity standards, line-strength standards, telluric standards, etc.--- must also be defined. If absolute flux calibration is desired then the 5arcsec-wide longslit should be used and at least three wavelength settings should be used, as in the MOS spectral response calibration above, to cover the full wavelength ranges of most gratings. Recommended settings are the same as the ones given in the table above.

The guidelines are the same as for normal longslit observations except that the classes for the on-sky observes should be 'Nighttime Program Calibration'. The associated acquisition observations should have class 'Acquisition' and any GCAL flats are still 'Nighttime Partner Calibration'. The time will be charged to the program. Daytime arcs are always 'Daytime Calibration' and are not charged.

Observations that will be split over several nights

Some observations cannot be completed within one night and therefore will require multiple acquisitions. Exactly how such an observation will be split in several observations over several nights will normally be determined by the Gemini Staff at the time of the observations. The user should define such observations as one observation, and add a comment to explain which assumptions were made about the number of reacquisitions and the resulting overheads. Normally all observations taking more than about 4 hours to complete (including overheads and baseline calibrations) will be split in several observations. Shorter observations may also be split in order to accommodate them in the queue planning.

Observing conditions

The observing conditions are specified as percentiles in image quality, cloud cover, sky background (and water vapor). Refer to the observing conditions page for details about the meaning of these percentiles. If the percentiles do not give sufficient information to the queue observer about the observing conditions required for a given observation, the user should add a comment to the observation detailing the observing conditions, e.g. "Need fwhm better than 0.75 arcsec in r". Such comments cannot be used to request better observing conditions than approved by the time allocation process.

Baseline calibrations

GMOS Baseline calibrations that are not specifically mentioned above should not be included in the Phase II programs prepared by the users.

Special calibrations, e.g. special standard stars, should be defined in the Phase II program, and the user's observing time will be charged for these observations.

Last Version: December 22, 2006; Rodrigo Carrasco
Previous Version: December 10, 2005; Kathy Roth

Grating	First Setting	Second Setting	Third Setting
R400	central wavelength - 150 nm	central wavelength	central wavelength + 150 nm
B/R600	central wavelength - 100 nm	central wavelength	central wavelength + 100 nm
R831	central wavelength - 80 nm	central wavelength	central wavelength + 80 nm
B1200	central wavelength - 30 nm	central wavelength	central wavelength + 30 nm