Home

GNIRS Users Manual - National Optical Astronomy Observatory

1. Grating Order Atmospheric Transmission Telescope Transmission 1 CM 80 93 2 CL 90 93 3 CK 93 92 4 H 93 91 CI 90 88 6 x 90 84 A 1 3 Instrument Throughput The instrument throughput is calculated using for the most part measured transmissions or reflections for the optical components including gratings and filters The grating efficiency varies across the band somewhat so an average value is given for the wavelength region defined by the order sorting filters The detector quantum efficiency is also included Table A 4 GNIRS Throughput Calculations for Long Cameras Grating Order Grating 10 mm 32 I mm 110 I mm 1 CM 35 23 36 2 L 52 28 38 3 CK 45 28 40 4 H 49 29 43 5J 50 30 41 6 x 41 27 36 The tabulated values are all for the long cameras 0 05 arcsec pixel scale The calculated values for the short cameras are all within 1 of the long camera values The lower values for the 32 1 mm grating are due to the reduced efficiency of this ruling The integral field unit is expected to have about 80 throughput Sensitivity for an object observed in a single slice will be 0 1 0 2 mag brighter compared with the same slit width 0 15 arcsec depending on whether the primary noise source is background or read noise and dark current The reflection losses in the prisms are 10 or less in addition the SF57 g
2. GMIRS CNOAO GEMINI NEAR INFRA RED SPECTROGRAPH USERS MANUAL GNIRS Users Manual Table of Contents 1 TRE OC UCEROR sscciss sce seiesustse cl cstceescvesaisseaeenstineeshenccasnsaslebesciveedonsatacuseteen 1 2 Instrument Overview sccsesescsssccsenecesssedscssdescsscdsiadscecesacecsnadecsnecdense sens 2 2 1 Instrument Description sssssseesseessesessesreserssesseesresrrssressessrsseessesse 2 2 2 Science Channel Performance x ciad ccisiacesscedacsdesecendssheceatessdeaves coatnnas 16 2 3 Observing Mode TAGES 9 feusness a Laed nieabedovcdedacensestadoctiiagetaanlatdontuaaestee 26 ZA WES Performan esineen ana a r eu 28 3 Observing with GNIRS e ssoesssesssesssesssoossoossoossssosssosessesesoossosossossssse 29 3 1 Preparation for Observing ss ssesessssesseserssreseessessrssressessessresseeseese 29 3 2 Engineernng Interface siare ordinii aa aa a aa 33 3 35 User Interfaco rene Setani ae el eater alia 38 zA Calibrations necne eee e A E EA 39 3 5 Preliminary Data Reduction sssssesseeseeseesseesreseesseesreserssesseseessesse 41 4 Set Up and Operation ssesssesssesssocssoossoosssssssseessocssoossoosssossssesssoee 43 4 1 Software Sta rt Upseccsnn s aR a E E RA etd 43 4 2 Initialize Spectrograph MechanismS eccescceseceteereeereeeeeeeeeeaees 43 4 3 Initialize Detectoren aiina onai aia 44 WA Imt hze QIWF S See cacita cscs cot aetdetocdune tata o a a 44 4 5 Sensor CSC KS ies es Seach c
3. The second major section describes how to observe with the instrument The information contained therein allows a user to prepare an observing program and to carry it out at the telescope For observers assigned queue time not all parts of this section are relevant Calibration data and initial data reduction procedures are also discussed in this section Portions of this section may be relevant when writing a proposal if calibrations or observing strategies are a concern The remaining sections are primarily relevant for people responsible for supporting the instrument that is the instrument scientist and other observatory staff more than the visiting classical observer Procedures for setting up and shutting down the instrument are described as well as basic trouble shooting procedures In general visiting astronomers will not find themselves carrying out procedures described here and should certainly embark on them with caution Additional procedures related to servicing and calibrating the instrument are found in the Service and Calibration Manual GNIRS Users Manual 1 2 Instrument Overview This section provides an overview of GNIRS including a functional description 2 1 on telescope performance of the science channel 2 2 including a discussion of observing modes 2 3 and performance of the available wavefront sensors 2 4 2 1 Instrument Description GNIRS is a cryogenic 1 5 um spectrograph with an on instrument wav
4. 3 1 92 2 54 um 4 open order 4 1 47 1 80 um 5 open order 5 1 17 1 37 um 6 open order 6 1 03 1 17 um 7 dark cross disp 0 9 2 5 um 8 left mask open 9 right mask open Note that the sorting filters for orders 3 4 and 5 are roughly equivalent to broadband K H and J filters respectively The long wavelength cut off of sorter 3 is significantly longer than for standard K filters as a result the sensitivity for target acquisition will be worse due to increased background see section 3 1 4 for more on acquisition procedures The filters are slightly tilted 2 7 degrees to reduce ghost images this is also why the filters precede the slit Because the filters are located in a converging beam they all have the same optical thickness in order to avoid refocusing the telescope each time a filter is changed This also ensures that filter changes keep the object centered on the slit Any user supplied filters must have the same thickness equivalent to 3 mm of BK7 in order to operate properly Users considering supplying such filters must check with Gemini beforehand as installation of such filters requires considerable prior planning 2 1 1 2 Spectrograph The spectrograph section consists of the following elements GNIRS Users Manual 5 Slit Decker IFU 2 mechanisms Collimator Acquisition mirror Prism turret Grating turret Camera turret Focus stage detector mount The spectrograph entrance slit is def
5. 83 10 80 90 19 11 18 98 18 33 17 57 13 45 10 36 1700 0 15 ave 21 93 21 76 20 52 19 59 15 32 12 18 90 20 54 21 36 20 08 19 26 14 94 11 74 5900 0 15 ave 21 32 21 16 20 04 19 07 14 90 11 77 90 20 98 20 81 19 61 18 74 14 42 11 33 5 sigma in 1 hour 50 percentile IQ Resolution Background Order Scale 6 x 15 CJN 4 H 3 CKY 2 CL 1 CM 1700 0 05 ave 20 68 20 60 19 54 18 61 14 45 11 21 90 20 41 20 31 19 13 18 29 14 07 10 77 5100 0 05 ave 19 28 19 15 18 42 17 65 13 52 10 39 90 19 17 19 03 18 12 17 38 13 15 9 95 17800 0 05 ave 18 35 18 22 17 72 16 95 13 00 9 97 90 18 29 18 17 17 51 16 75 12 62 9 54 1700 0 15 ave 21 26 21 10 19 85 18 91 14 64 11 50 90 20 86 20 69 19 41 18 58 14 26 11 06 5900 0 15 ave 20 66 20 50 19 37 18 39 14 12 11 09 90 20 31 20 14 18 94 18 07 13 74 10 65 GNIRS Users Manual 21 For observations with the Wollaston prism there will be two spectra plus some reflection losses and vignetting in the prism so the magnitude required to get 5 sigma in each spectrum is roughly 0 5 mag brighter than the tabulated values for background limited observations and roughly 0 9 mag brighter than the tabulated values for detector limited observations For observations in cross dispersed mode there are some reflection losses in the prism as well as absorption losses in the K window mainly at the red end The
6. Timing Array Readout ROI Header Comment Image Title Setup Command Screens Data Image Cntrl Array setup cntrl Figure 3 4 Observing Set Up Used to set up array control parameters GNIRS Users Manual 36 The observing set up screen provides more functionality than is normally needed for observations Don t alter parameters that are best left alone The following items mainly on the left side of the screen are relevant e Activation This starts up defaulting to inactive so you must activate manually when starting up the instrument Once the detector is activated you can leave this alone Integration time This should be set as needed Digital averages and low noise reads Digital averages should be set to 16 except for L and M band acquisition set to 1 in these cases Low noise reads should be set to 1 for exposures under 1 minute and to 64 for exposures gt 5 minutes For exposure times between these values consult the ITC Co adds This is the number of frames to be added together Co addition is useful if you have many short exposures since you save on time writing to the DHS and the number of images you have to manipulate is less However if any one frame is bad the co added image will be bad so co addition is not recommended if you are taking longer exposures Number of images This is the number of separate images to be taken Note that if you are also co adding the number of exposures is the produc
7. activation 4 3 It is a good idea but not necessary to re datum the mechanisms 4 2 4 4 If you are doing afternoon calibrations do the datum before calibrating 4 9 Shut Down This section outlines the steps required to prepare the instrument for removal from the telescope This includes all steps up to but not including powering down electronics This is not required except at the end of the run 4 9 1 Mechanism Positioning The mechanisms should all be positioned to their park positions 4 9 2 Detector Shut Down The detector should be de activated from the observing control screen Don t forget to use the set button to implement the deactivation GNIRS Users Manual 47 4 10 Nightly Shut Down It is only necessary to de activate the detector see 4 9 2 above and close the window cover closed or park The other mechanisms can be left as is GNIRS Users Manual 48 5 Basic Trouble Shooting This section is not intended to be provide comprehensive diagnostics of all possible problems with the instrument Instead it provides preliminary diagnosis of the most common problems with the instrument This section is expected to be revised as experience using the instrument on Gemini accumulates Problems diagnosis and solutions No signal at detector This may be due to a number of causes Detector not activated Activate the detector If it does not activate verify that the detector temperatu
8. amount typically around 40 The net gain in signal to noise is about 15 GNIRS Users Manual 51 equivalent to 0 15 mag or to a factor of 1 3 in integration time for the same object This may be worthwhile if the associated loss in spectral resolution is not a concern The length of the slit used to actually extract the spectrum is also relevant It is not necessarily optimal to use a wide window and get nearly all the signal since the increased total background and total dark current and read noise may provide a larger increase in noise Table A 2 tabulates the percentage of transmitted flux extracted as a function of slit window length Thus the total signal extracted is the product of the factors in this table and the preceding table For example a 4 pixel window and a 2 pixel slit with the short cameras and 50 percentile image quality will extract 44 of the flux incident on the slit 52 x 84 Note that still better results can be obtained by means of optimal extraction using the actual light profile along the slit but the values in Table 2 11 will be close to those results For reference the relative amount of noise from background including dark and read noise is also indicated it s simply proportional to the square root of the window size Table A 2 Flux Extracted vs Slit Window Size Slit Window 20 Percentile 50 Percentile 85 Percentile Relative Length 0 30
9. and masks Table 2 except for filters in FW 1 FW 2 any 2 1 3 On Instrument Wavefront Sensor The OIWFS was built for GNIRS by the Institute for Astronomy Hawaii The light path is also shown in Figure 2 2 Light from the guide field anything in a 3 arcminute diameter field that misses the pick off mirror enters the OIWFS field lens and then passes to a collimator doublet and then a 2 axis gimbal mirror The gimbal mirror can be tilted precisely to direct light from any part of the guide field to the rest of the OIWFS optics The usable field is roughly 10 arcsec diameter which is sufficient to acquire individual guide stars Light from the selected patch of sky is reflected off the gimbal mirror through a second doublet and the guide star is re imaged on the filter wheel which contains JHK filters and apertures From the filter wheel light enters the Shack Hartmann optics which are mounted on a snout on the detector mount The Shack Hartmann optics form a pupil image at a shallow four facet prism then reimage the star on the detector Because the light has passed through the S H prism four images of the star are actually formed corresponding to 1 4 of the pupil each Only a small portion of the array is read out allowing rapid data rates which permit the OIWFS to provide high speed tip tilt and focus correction in GNIRS Users Manual 12 addition to slower flexure and tracking correction F
10. color of the object determines whether the limiting sensitivity is set at the short or long wavelength end of the spectral region covered with the prism the decrease in sensitivity ranges from 0 1 mag at the shorter wavelengths to 0 3 mag at the longest wavelengths Sensitivity with the IFU is more complicated because one wants to look at sensitivity for each spatial sample IFU observations will generally be done under very good seeing conditions A rough approximation is to use the ong camera sensitivities even though the actual configuration uses the short camera 2 2 1 6 Long vs Short Under the same image quality conditions except for very best conditions i e 10th percentile the short cameras will provide better performance than the long cameras The difference is most significant at R 5100 5900 where the short camera configuration uses amore efficient grating At R 1700 the short camera grating is less efficient and performance is more nearly comparable The short camera configurations use fewer pixels along the slit so there is a reduced contribution from read noise and dark current which is significant at short wavelengths mainly xJH The long camera is therefore mainly recommended in two cases e R 17800 is needed e High spatial resolution along the slit is needed 2 2 2 Wavelength Calibration The night sky will always have enough well defined emission features to provide a good wavelength calibration In addition
11. own engineering screen accessed from the top level screen 4 5 Sensor Checks The health of the temperature and pressure sensor values is indicated on the top level CC screen If both health indicators are green then all values are within range and there is no need to check further If the health indicators are yellow warning or red bad one or more readings is out of range Open the relevant sensor screen and examine the faulty value or values Consult the Service and Calibration Manual for details on the individual sensor values If the health is yellow due to a single slightly out of range sensor it may be reasonable to continue observing However in this case one should check the relevant screen periodically to confirm that the situation has not gotten worse 4 6 Configuration Checks This section describes simple functionality checks that can be carried out without need to access the low level code Additional checks are described in the service and calibration manual GNIRS Users Manual 44 4 6 1 Basic Configurations The instrument s ability to correctly configure can be checked by setting up a few configurations taking an image and comparing the results with expectations For these tests GNIRS should be looking at GCAL with the arc lamp Ar on Exposure times of 1 second TBD are sufficient Make sure the window cover is open Try the following Configure in acquisition mode mirror in acquisition slit and d
12. performance This is best determined using multiple frames use IRAF imcombine The noise should be measured using minimum Fowler sampling 16 digital average and 1 low noise read and maximum sampling 16 digital averages and 64 low noise reads Use a bias of 0 3V The respective median noise values measured should be 40 and 7 electrons in the sigma images It is also useful to visually inspect the difference frames for evidence of noise pick up which would appear in the form of patterns superposed on the random noise Some pick up will probably be present in the minimally sampled frames but it should be averaged out fairly well with multiple samples 4 7 Night Time Tests The following tests should be carried out at the start of the night The weather does not need to be photometric Pupil viewing test Configure for pupil viewing through the K sorter 3 filter and acquisition mirror Use the long blue camera for best results Verify that the telescope secondary is centered on the internal cold stop The images Fig 4 1 below show a properly centered and somewhat decentered image Decentering at the 1 level is not critical Figure 4 1 Pupil viewing images of secondary well centered on cold stop left and slightly decentered right GNIRS Users Manual 46 Background test Go to the configuration with the 32 I mm grating short blue camera and long slit mode with a 2 pixel slit and sorter 3 Set the grating to 2 2 um an
13. repeatability will then be several pixels as opposed to 1 pixel or less for slit motions only GNIRS Users Manual 32 3 2 Engineering Interface The engineering interface is not intended as the primary user interface to GNIRS while observing but can be used for that purpose At present there is no user interface so there is no alternative The engineering interface consists of a set of EPICS windows which can be used to monitor and control the instrument configuration data taking and some operations of the OIWEFS For routine observations the OIWFS is controlled by the telescope acquisition and guiding system and the user needs to interact only with the mechanisms components controller and the array array controller In addition users may wish to open windows to run IRAF or some other data reduction analysis package The descriptions below cover only those screens required for routine operation of the instrument Additional windows used for diagnostics are covered in the Service and Calibration Manual and in the Software Maintenance Manual Detailed data reduction procedures are outside the scope of the manual but see section 3 5 for an overview These descriptions also assumed that the instrument and its associated software are running see section 4 for start up procedures The windows for GNIRS are EPICS dm screens which can be used to monitor and modify EPICS variables The diagnostic windows are an interface to the C
14. the on telescope calibration unit GCAL provides the ability to observe arc lamps for set up when the dome is closed One should not rely on the instrument control software for more than rough calibration good to about 10 pixels for center wavelength The accuracy is limited by the grating turret repeatability and the approximations used in modeling the wavelength as a function of grating tilt and array position GNIRS Users Manual 22 2 2 3 Slit Tilt and Curvature The instrument is configured so that the spectrograph slit is aligned with the array columns when in long slit mode This alignment is not exact however There is some residual error in the alignment which leads to a net rotation of lt 1 pixel along a slit running the full width of the array 1024 pixels e As with any long slit spectrograph there is also curvature along the slit The size of the effect depends on the configuration The displacement of the center position relative to the ends varies from less than 0 1 pixel to almost 3 pixels The effect is larger for higher resolution and for the short cameras There can be some change in the dispersion as well in that the curvature varies slightly as a function of wavelength on the array The cross dispersed modes introduce a tilt of the slit image which is a function of position on the array For the short slits used in these modes the tilt is modest but still significant typically around 0 5 pixel end to en
15. the two linear polarization components of the light and can be used through the L band Because there is substantial internal polarization in the spectrograph itself the Wollaston prism configurations must be used with GPOL on the telescope s up looking port From the prism turret light goes to the grating turret The grating turret contains three gratings Table 2 5 GNIRS Grating Resolutions Grating Long Camera Resolution Short Camera Resolution 10 44 I mm 1700 570 31 7 mm 5100 1700 110 5 I mm 17800 5900 All three gratings are blazed for 6 8 um first order Littrow which provides an effective first order blaze wavelength of 6 6 um in the configuration actually used scattering angle of 27 degrees The different orders of the gratings then correspond fairly well to the atmospheric windows at 5 3 5 2 2 1 6 and 1 2 um for orders 1 through 5 respectively the sorting filters specified in Table 2 1 cover the free spectral range of the individual orders with some allowance for filter roll off A filter for order 6 is also supplied the orders above 5 don t match the atmosphere particularly well The resolutions provided by the gratings are tabulated above Table 2 5 The values given are with the gratings operated at the blaze peak Tilts to longer wavelength provide GNIRS Users Manual 8 somewhat higher resolution while tilts to shorter wavelengths provide lower resolution The resolution i
16. 2 1 2 Spectrograph Configurations With multiple filters slits prisms gratings and cameras there are in principle a very large number of possible configurations Although the instrument can be configured to any of these in practice only a much smaller number are of interest These are tabulated below for reference Table 2 8 Acquisition Configurations ID Name Filter Slit Decker Acq Prism Grating Camera Mirror 1 SBAcq sorter 3 7 acquisition acquisition in N A N A short blue 2 SRAcq sorter 1 3 short red 3 LBAcq sorter 3 7 long blue 4 LRAcq sorter 1 3 long red Notes Ability to read array with sorter 1 uncertain not a requirement Table 2 9 Spectroscopic Configurations ID Name Filter Slit Decker Acq Prism Grating Camera Mirror 5 SB10 sorter 3 0 3 100 out mirror 10 44 short 6 SB31 7 arcsec arcsec 31 7 blue 7 SB111 110 5 8 SBXD broad 6 arcsec short 31 7 band prism 9 SB10P sorter 3 14 4 Wollaston 10 44 10 SB31P 7 arcsec 31 7 11 SB111P 110 5 12 SB10IFU IFU 6 arcsec mirror 10 44 13 SB3 LIFU 31 7 14 SB111IFU 110 5 15 SR10 sorter 1 0 3 100 10 44 short red 16 SR31 amp 2 arcsec arcsec 31 7 17 SR111 110 5 18 SR10P 14 4 Wollaston 10 4
17. 4 19 SR31P arcsec 31 7 20 SR111P 110 5 21 SR10IFU IFU 6 arcsec mirror 10 44 22 SR3 LIFU 31 7 23 SR111IFU 110 5 24 LB10 sorters 3 0 1 50 arcsec 10 44 long blue 25 LB31 7 arcsec 31 7 26 LB111 110 5 27 LB10XD broad 3 arcsec long prism 10 44 band 28 LB10P sorters 3 14 4 Wollaston 10 44 29 LB31P 7 arcsec 31 7 30 LB111P 110 5 31 LR10 sorters 1 50 arcsec mirror 10 44 long red GNIRS Users Manual 11 32 LR31 amp 2 31 7 33 LR111 110 5 34 LR10P 14 4 Wollaston 10 44 35 LR31P arcsec 31 7 36 LR111P 110 5 Notes PSlit listed is width optimally sampled at detector other widths can be used Spectrum does not fill entire detector width Grating listed provides complete 0 9 2 5 um spectrum other gratings can be used In principle any configuration can be used for diagnostic purposes However the two main optical diagnostics contemplated for the instrument are pupil viewing and focus tests Note that unlike the configurations listed in the two preceding tables there are more variables in entries below Table 2 10 Regular Diagnostic Configurations ID Name Filter Slit Decker Acq Prism Grating Camera Mirror 37 Pupil View FW 1 pupil pupil pupil view in N A N A long blue view lens 1 view or long FW 2 lens 2 red sorters 2 4 38 Focus test FW 1 pupil Focus tests may be done on any configuration listed in Table 1
18. 9 3 4 2 Wavelength Calibration The easiest way to determine the wavelength calibration is to use the night sky lines which are present in every spectrum taken with the instrument Suitable lines are identified in the spectra in Appendix B GCAL also provides arc lamp spectra which can be observed to determine wavelength settings during the day or when the dome is closed by weather The GCAL lamps can be useful for diagnostic purposes or for verifying exact wavelength coverage but are otherwise not needed Consult the GCAL manual for operation of the arc lamps It is not absolutely necessary to subtract darks or flat field these GCAL frames 3 4 3 Standard Stars Selection of standard stars is discussed above section 3 1 3 Standard star observations are generally fairly short so it is probably possibly to carry out the observations without setting up the OIWFS which can save a certain amount of time It is still advisable to measure the star at several slit positions This is especially true for telluric standards used with the long slits where one should really measure at each position on the slit that is used for objects If a large number of positions are involved or the exposures are longer than a few seconds it is probably advisable to take the trouble to use the OIWFS rather than letting the star drift off the slit 3 4 4 Polarization The issue here is that one would like to calibrate the re ative instrument response i
19. AD CAR APPLY records and also show the current status of the system Several display buttons will display a menu when right clicking with the mouse Some of these menus are dynamically loaded so if the system is rebooted or the menu is reloaded after the dm screen is already displayed the right click menu won t be updated This is easily fixed by exiting the affected window and re displaying it from its parent window Wp nirsseqtopd rr Gemini Near InfraRed Spectrograph Top Level Screen S H B gt Instrument Sequencer o o gt Components Controller E E gt Detector Controller gt Wavefront Sensor S State H Health B Busy JERE Exit Figure 3 1 GNIRS Top Level Engineering Window The window shown above Figure 3 1 is the top level engineering window for GNIRS GNIRS Users Manual 33 On this window there are four buttons and status indicators that correspond to the four major components of the system see the Service Manual for a complete description This window must not be closed though it can be minimized The buttons on the left bring up additional screens including the screens used for observation The State Health Busy indicators provide a top level indication of the instrument status Normally all indicators in the two left columns should be green the busy indicators may be busy Note operation of the motors can induce pickup in the temperature sensors so it is possible for the components
20. CUS amp FOCUS STAGE Figure 2 1 Optical schematic for GNIRS showing light flow through the science channel The instrument opto mechanical layout is shown in Figure 2 2 GNIRS Users Manual 2 eua OPTICAL BENCH WINDOW LIGHT FROM PICK OFF TELESCOPE MIRROR OIWFS C FIELD LENS Ls Wir FOLD MIRROR Bw if 5 cou stop mm ese P Borie SHACK HARTMANN FILTER ae TeS OFFNER RELAY gag OIWFS DETECTOR LONG un D STATIC FOLD a MIRROR 2 Ld ew AAN COUSIN MIsROR f S WOSHOWN lt at SCIENCE 7 ae amene oe CONCTRED CAMERA DETECTOR LES sui BLUE CAMERANE Tg REDICAMERA SCIENCE CHANNEL RAY RED OIWFS CHANNEL RAY BLUE OFF AXIS COLLIMATOR Figure 2 2 Instrument internal structure showing light paths The light path in red is the path through the spectrograph starting at the entrance window which is not shown The light path in blue is the path through the OIWFS starting at the pick off mirror The actual internal structure of the instrument differs in some details from this figure Both the science channel and the on instrument wavefront sensor are mounted within a cold structure contained within a vacuum vessel The instrument electronics are mounted externally The cold structure is operated at a temperature of 60K in order to minimize excess background on the detector 2 1 1 Spectrograph Description The spectrograph design is a fairly conventional one for infrared
21. Data Taking Two screens are required to set up the array controller and take an observation The first screen Figure 3 3 is the GNAAC main control screen from that you can access the Observing Set Up screen Figure 3 4 which is used to set up the observation GNIRS Users Manual 35 gnaacEngDisp dl GNAAC Engineering Display Apply Cad Test Screen Additional Screens Overall Control Image Setup a Raw Housekeeping Setup Commands Data Path E02 Image Filename M T is ARNING Data Label MN Ste es rire nlth Headers NORMAL Por sup xlth Observation Setup EELE Observation State eee sage Ene IDLE nt 0 285 HK Process State RUNNING Controller State Installed array size aen RUNNING i 1024 Rows by 1024 cols Low Noise Rds Current Readout Size Digital Avgs Rows to he read 1024 i e oome Pictures Cols to be Read 1024 uCode Waveform Generator Program A bo obs setup ome ucode Path fhome gemini gnirs ucode NAAC F cmd Path fhome gemini gnirs ucode NAAC Figure 3 3 Array control screen Used to start an observation and control some parameters GNAAC Observation Setup Display Minimum gt 0 253 Activation Digital Avgs e i a Low Noise nds EEH Co adds of Pictures Proc Mode Ef CHOP TES Array Readout ROI ri aii j Area Frea Collecting Data Data HK Process Statef Area not Area not Collecting A Header Detail Frames per Exp 2 Header
22. alter grating and prism tilts This facilitates prolonged observing sequences on faint objects since the dispersive settings remain stable even while target positions are checked The position of the acquisition mirror is shown in Figure 2 2 where the return beam from the collimator crosses the beam into the camera When inserted it diverts the light at the point where the two beams cross If the acquisition mirror is out of the beam light goes from the collimator to the prism turret The prism turret has four possible positions GNIRS Users Manual Dy Table 2 4 GNIRS Prisms Position Application Mirror Long slit mode for all 4 cameras Long camera cross dispersion Used with long blue camera 10 I mm grating for 0 9 2 5 um coverage Short camera cross dispersion Used with short blue camera 32 1 mm grating for 0 9 2 5 um coverage Wollaston prism Polarization mode for all 4 cameras The mirror is used for work beyond 3 um or when one wants to work with a long slit at shorter wavelengths The two cross dispersion prisms provide a cross dispersed low resolution spectrum over the approximate range 0 9 2 4 um where the two prisms are matched to the two pixel scales produced by the cameras A complete spectrum is produced at a resolution of 1700 2 pixels use of higher spectral resolution results in more or less parallel portions of multiple orders but not a complete spectrum The Wollaston prism separates
23. amera configuration 4 6 3 Background and Throughput The instrument internal background should be measured in two configurations Filter wheel 1 in dark position filter 2 in sorter 1 position 180 micron slit short camera long decker short red camera prism turret in mirror position 110 line grating set to 5 0 um 1 order Use an exposure time of 300 seconds The observed count levels includes dark should correspond to 0 1 electrons sec pixel Gain is 13 e ADU don t forget to divide by number of low noise reads GNIRS Users Manual 45 Same as above except that grating is now set to 2 2 um 3 order filter wheel 2 in sorter 3 position and short blue camera is used The observed count should still correspond to 0 1 electrons second pixel Optionally the same configuration can be repeated with the short camera cross disperser in place instead of the mirror the measured counts should be the same In general the best measurements of system throughput are done on the sky However an initial check can be done with GCAL and the continuum flat field source For the second configuration above the count level at the center of the array should be approximately TBD ADU sec Needs more detail The measured value should be within 10 TBD of this Note this assumes that the GCAL configuration is fairly stable 4 6 4 Detector Noise In conjunction with the background tests one can also look at the detector noise
24. anation of these coefficients also confirm that you are using the most current values for the coefficients To remove dark and flat field subtract the equivalent dark frame and then divide by the appropriate flat field Bad pixels or unexposed areas should be masked off in the operation If no exact match exists for the dark frame either interpolate using exposure times that are close or use the closest dark frame If unmasked bad pixels are noisy in the flat fielded frames it is desirable to limit their maximum and minimum values to reasonable numbers this ensures that IRAF s statistical processes are more reliable 3 5 2 Sky Reference Generation and Subtraction The details of the sky reference generation depend on the observing procedure If the object was measured at many slit positions the sky reference can be generated by calculating the scaled median of all images If a simply on off or A B procedure was GNIRS Users Manual 4 used then the sky reference is the off image or the mean of the adjacent off images For a modest number of positions one can construct a reference for each position by determining the average of all the other positions This is the preferred procedure The reference should be scaled to the object position flux values and subtracted 3 5 3 Extraction If multiple positions on the slit were observed it is generally preferable to extract individual spectra which can then be wavelength calibra
25. arcsec 0 50 arcsec 0 75 arcsec Background arcsec Noise 0 30 52 36 0 71 0 45 52 0 87 0 60 98 64 1 00 0 75 100 76 1 12 0 90 100 96 84 1 22 The optimum window for a given image quality is highlighted in yellow If performance is within 2 multiple values are highlighted Generally a window slightly larger than the image FWHM is optimum Note that the window to be used can be determined after data are taken Sampling along the slit with the IFU is slightly different than the short camera about 0 12 arcsec pixel so a window of 0 6 arcsec corresponds to 5 pixels rather than 4 Since the purpose of IFU observations is spatial reconstruction the observer will probably extract spectra row by row rather than defining a wider window A 1 2 Telescope and Atmosphere The remaining contributors to throughput are the instrument itself the telescope and the atmosphere The estimated transmission of the telescope and the atmosphere are tabulated below in Table A 3 The atmospheric transmission values are for regions free of strong absorption lines GNIRS is assumed to be on a side port so the ISS fold mirror is included as well as the obscuration by the secondary mirror Telescope reflections are GNIRS Users Manual 52 assumed to be clean aluminum Values are tabulated for the 6 orders for which the instrument currently has blocking filters Table A 3 Atmospheric and Telescope Throughput
26. be provided by Gemini will provide a more complete interface to the instrument and relevant telescope systems It provides the capability to program a series of rastered dithered observations of a single object including OIWFS control It will also be possible to program a sequence of observing configurations for example a sequences of dithered observations with multiple filters The interface despite these added capabilities will not involve a profusion of windows or don t change these parameters This section will contain a complete description of the interface once it is available or in test GNIRS Users Manual 38 3 4 Calibrations Several types of calibration frames must be obtained for proper data reduction These fall into four categories Darks and flats Wavelength calibration Standard stars Polarization calibration 3 4 1 Darks and Flats Dark and flat field frames are required to remove the detector signature Dark frames are produced by putting dark filter in Filter Wheel in position and taking appropriate exposures In general it is a good idea to take dark frames that match the regular data frames in terms of bias exposure time and numbers of digital samples and low noise reads Some interpolation on the times is permissible but in general you will get better results carrying out your program with a limited number of different integration times Flat fields are produced using the on telescope calibratio
27. bservation of a reference star to provide an absorption template should be carried out with the reference star close to object and with the spectrograph in the same orientation slit angle Flexure in acquisition mode is larger as much as 1 pixel hour with the long cameras See 3 1 4 for a discussion of object acquisition 2 2 6 Repeatability The mechanisms in GNIRS are not perfectly repeatable That is once a mechanism is moved a return to the same nominal position will be close but not exact The nominal performance specification is 10 pixels repeatability with the long cameras lt 1 pixel with the slit From the point of view of exact wavelength calibration this means that each configuration of the instrument is a new configuration requiring its own calibration even it is nominally the same as a previous calibration There are a couple of exceptions to this rule Filter changes do not change the configuration after the slit at all so one can cycle through orders at a fixed grating orientation and maintain calibration Also the acquisition mode inserts a mirror in the beam but does not move the prism grating or camera so checking centering will not alter the spectroscopic configuration Note that the acquisition mode itself is only repeatable to a few pixels TBD so one should check centering relative to the image of the slit and not relative to absolute pixel coordinates These considerations would affect observations whe
28. conditions are relevant and also which were used for the baseline calculations provided here There are five main observing constraints Image quality seeing Cloud cover Water vapor Sky background Air mass zenith distance Image quality values relevant to GNIRS are tabulated below Note that these are the values used in computations in this manual if Gemini changes the specifications both this table and the computations must be changed Table 2 11 Gemini Image Quality Constraints Image FWHM arcsec Wavelength 20 percentile 50 percentile 85 percentile any 0 9 um 0 40 0 75 1 05 1 70 1 2 um 0 35 0 55 0 80 1 55 2 2 um 0 30 0 50 0 75 1 40 3 4 um 0 30 0 45 0 70 1 25 GNIRS Users Manual 19 Image quality at wavelengths beyond 3 4 um will be similar to that at 3 4 um Results are provided below for both 20 and 50 percentile conditions Trade offs involved in selecting better or worse image quality are discussed in 2 2 2 Cloud Cover All calculations assumed conditions were photometric which corresponds to 50 percentile cloud cover The presence of cloud is undesirable for two reasons signal is attenuated and background will be more variable complicating sky subtraction Note that observations under marginally photometric conditions are reasonable since GNIRS spectra will not normally have high absolute photometric accuracy anyhow see section 3 4 Water Vapo
29. controller health indicator to change color while mechanisms are operating If it does do not be concerned unless it does not change back to green within 2 3 minutes 3 2 1 Instrument Configuration The opto mechanical configuration of the instrument can be controlled from a single Mechanism Control screen Figure 3 2 This is opened by first opening the Instrument Sequencer screen from the top level screen above and the opening the Mechanism Control screen from the IS screen The IS screen can then be closed gmSeqMechCommands d GNIRS Mechanism Control EER Select configuration then APPLY to execute cc state Filter Wheel 1 ee Selected filter Grating Cover e Position Gratings Selected grating Filter Wheel 2 Acq mirror Figure 3 2 Mechanism Control screen This provides control for all mechanisms Selected filter Decker CEE 1 Selected slit Slit Selected width GNIRS Users Manual Prism gt short_xd Selected Prism 34 TNE Position camera gt Short_Red Selected camera The user selects mechanism positions for each of the discrete mechanisms The grating is positioned using a wavelength and order The focus is positioned by step number One enters all desired changes into the configuration displayed on the screen and then uses the apply button to reconfigure the instrument If another configuration or an observation with the science array is in pr
30. d compare the results of a 60 second exposure with the relevant figure from Appendix B Compare both the overall intensity of the night sky emission lines and the continuum level at the long wavelength end of the spectrum Both levels may differ significantly from what is shown in the figure If the test is done at the start of the night the night sky emission levels will tend to be higher than the reference values possibly by a factor of 2 The thermal continuum is affected by ambient temperature and will be higher if the telescope temperature is higher than the value at the time the data in Appendix B were taken see the figures Throughput test Observe a standard star with a wide slit in any of the configurations documented in Appendix B ideally one you are intending to use that night A simple extraction will produce values within 20 of those you would get with a complete preliminary reduction section 3 5 this provides a quick check of performance Outside of regions of atmospheric absorption the variations in measured counts should be no more than 10 4 8 Nightly Start Up Normally the nightly start up only needs to reverse the steps from the previous night s shut down section 4 9 e Ifthe EPICS screens have been closed restart by typing nirsSeqStart ina terminal window See 4 1 above Check the top level screen to make sure temperature and pressure health are still good 4 5 Re initialize the detector including
31. d for extreme positions In addition the spectra themselves run at an angle on the array i e constant slit position is not constant x position on the array The angle is such that wavelengths seen in more than one order are at the same x position The separation of the two polarizations in the Wollaston mode is a weak function of wavelength decreases with increasing wavelength In general these tilts should not have a substantial effect on the data However in some circumstances it may be useful to shift spectra slightly before co adding them In the case of spectra using the full slit length it may also be necessary to interpolate between multiple wavelength calibrations if precise wavelengths or velocities are required 2 2 4 Spatial Resolution In general the optics in GNIRS provide spatial resolution along the slit as good or better than that implied by the pixel sampling For the most part even the short camera 0 15 arcsec pixels samples the delivered image quality Therefore the instrument s spatial resolution should be considered to be that defined by the delivered image quality except for 20 percentile image quality or better For this level of image quality the resolution along the slit will be limited to 0 3 arcsec for the short cameras and will continue to be limited by the actual image quality for the long cameras The spatial resolution for the IFU is similar to that with the short cameras 0 12 x 0 15 arcs
32. e pick off mirror is tilted at 45 degrees and is therefore exactly at the telescope focus only in the center of the slit The mirror widens away from its center to ensure that the spectrograph field is unvignetted The field available for guiding is discussed in section 2 1 3 The Offner relay is serves two functions it produces an image of the telescope secondary on the Offner secondary where a cold stop is located and it re images the telescope focal plane onto the spectrograph slit The scale at the slit is the same as at the telescope focal plane GNIRS Users Manual 4 The spectrograph contains two filter wheels Each wheel can accommodate 9 filters in addition to an open position The first wheel uses several of the positions for focus masks a dark position and a lens used to view the telescope pupil during alignment see below These can be used in series with filters in the second wheel The remaining positions in the first wheel can be used for back up filters The current filter complement for the instrument is listed below prospective observers should verify e g via the Gemini web site that these are the filters that will be available at the time they wish to observe Table 2 1 GNIRS Blocking Filters Position Number Filter Wheel 1 Filter Wheel 2 0 open open l pupil viewer lens 1 order 1 4 4 6 0 um 2 cross disp 0 9 2 5 um order 2 2 90 4 25 um 3 order 2 2 90 4 25 um order
33. e section describes a complete start up after installation for daily start up see section 4 8 Software troubleshooting is covered in the Software Maintenance Manual 4 1 Software Start Up The Engineering screens are started by typing nirsSeqStart ina terminal window This script is located in GNIRSROOT bin solaris If this is not in your path use the whole path to invoke it This command will eventually bring up the top level instrument screen as well as a small screen labeled DM2 4 Both screens must not be closed although they can be minimized From the top level screen you can bring up the following additional screens see 3 2 Mechanism control screen brought up from instrument sequencer screen which can then be closed GNAAC top level screen overall control GNAAC observing control screen No other screens are required for routine observing Screens required for diagnostic purposes are described in the Service and Calibration Manual these are generally accessed through the components controller main screen accessed from the top level screen 4 2 Initialize Spectrograph Mechanisms The mechanisms can be datumed home position is located using the global datum command on the CC screen This will take a couple of minutes to execute depending on where the mechanisms happen to be At the completion of the datum all mechanisms will be left near but not quite at the zero step position You should then posi
34. e star used for this purpose does not need to be a photometric standard though it must have a relatively featureless spectrum For lower resolution F and G stars work well because they have relatively weak features At higher resolution more lines become evident and early type stars may be preferable outside the H I lines A list of recommended standards will need to be developed In many cases it will be possible to arrange the observations so that the same star is used for both flux calibration and telluric correction This essentially requires one standard observation for each object observation unless objects are close together in the sky and are fairly bright e g stars in a cluster 3 1 3 Slit Orientation Two factors affect the orientation of the slit and the IFU One is the structure of the object to be observed and the other is differential refraction The orientation relative to the object is dictated by the science to be done and is therefore specified by the observer The optimum orientation from the point of view of differential refraction may be quite different observers need to understand the effects involved Differential refraction smears the spectrum of an object by an amount that is significant on the scale of the narrow slits used in GNIRS GNIRS Users Manual 30 For example the differential refraction between 0 9 and 2 5 um roughly the wavelength range of the cross dispersed mode of the spectrograph is
35. e unvignetted GNIRS patrol field is effectively a 3 arcminute diameter circle minus a strip roughly 20 arcsec across the area is approximately 6 square arcminutes Models of the guide star distribution on the Gemini WFS web page suggest that stars of the required magnitude will be relatively common at galactic latitude 30 degrees but less frequent at higher galactic latitude roughly half the all fields Fainter stars can in principle be used at the price of losing full tip tilt and focus correction For large high latitude samples guide star availability may be useful way to select smaller sets of objects for observation 2 4 2 PWES Performance See the Gemini web pages follow Science Operations link to Telescope then to WFS and Guide Stars GNIRS Users Manual 28 3 Observing With GNIRS This section describes how to observe with GNIRS The first sub section 3 1 discusses preparation for observing Some of its content may be needed to prepare a proposal and all of it is relevant to preparing the actual observing program The next sub sections discuss the interfaces to the instrument section 3 2 describes the engineering interface which is what the early user must confront and section 3 3 will eventually describe the user interface Calibrations are discussed in section 3 4 and preliminary data reduction is discussed in section 3 5 3 1 Preparation for Observing This section assumes that you have already specified the inst
36. ec samples so the IFU resolution will also be seeing limited until image quality reaches approximately 20 percentile GNIRS Users Manual 23 2 2 5 Flexure GNIRS undergoes flexure as the orientation of the instrument changes There are three types of flexure that are relevant Flexure of the instrument as a whole relative to the telescope instrument support structure e Flexure of the OIWFS relative to the spectrograph slit or PWFS if OIWFS cannot be used Flexure between the spectrograph slit and the detector The flexure of the instrument relative to the ISS has components of tilt and displacement The displacement under 2 arcsec maximum is taken out by the OIWFS and is small enough that effects on the image quality or plate scale are negligible Note that if no guide star is available to the OIWFS so that the PWFS must be used flexure will be significant and will require compensation using look up tables Recentering may need to be more frequent in this case The effects of tilt are to decenter the image of the telescope secondary on the instrument s cold stop in the Offner relay This leads to some vignetting and loss of signal The maximum decenter is approximately 1 of the pupil diameter which leads to an equivalent loss of signal This effect is not significant either in terms of overall sensitivity or photometric accuracy Flexure of the OIWFS relative to the spectrograph slit leads to progressive decent
37. eca e a sano eee oes ar E A 44 4 6 Configuration Checks issiacyscsslesiacenscsesesvactsaetnaaptdiaiaseatiecapsarncaavigeas 44 Ao Night Tim Tests si cscsncusn tow tehleecnie ta ia e ts 46 4 8 Nightly Start Upa eean caus n e E E N EER 47 4 9 Sht DOWN itt os se teed cud Sadun idan tac nase oai a aA ES 47 4 10 Nightly Shut DoWwn ci iscise ie cien tees esate ea tatias Vea eects te Saas 48 5 Basic Trouble Shooting sccccssccccsseccscsscccsssccccssescssssseesssseesees 49 APPENDIX A Supplementary Information for Exposure Time Calella toS eee Cnn Pie ite ate Ata oe eh eee TSP RC enn wee aE arn eo ten ee ere eee 51 APPENDIX B Representative Calibration and Night Sky Spectra 56 GNIRS Users Manual 1 1 Introduction This is the Users Manual for the Gemini Near Infrared Spectrograph GNIRS GNIRS is a 0 9 5 um spectrograph that supports a variety of observing configurations including long slit cross dispersed IFU and polarization analysis modes as well as two different pixels scales and several different spectral resolutions The GNIRS Users Manual is organized in several major sections plus this introduction The first major section comprises an overview of the instrument It is intended to provide the information a prospective user of the instrument might need first to determine whether the instrument is suitable for his or her scientific needs and then to write a proposal to use the instrument on Gemini
38. ecker with the H sorter 4 filter Use the short blue camera The acquisition field should be roughly centered on the array and the field should not be vignetted Configure with any grating and any of sorters 3 6 a blue camera and the short camera long slit decker Use a moderately narrow slit 180 microns Specify a wavelength corresponding to one of the figures in Appendix B and verify that the wavelengths are about the same to within 20 pixels as in the Appendix and that the slit is centered on the array to within 20 pixels Confirm that the FWHM of the lines is 2 2 pixels or less Note at low resolution some lines are blends check the labeling in Appendix B to avoid these Configure to a cross dispersed mode Verify that orders 3 7 fall on the array and that the lines remain narrow FWHM 2 pixels for short camera closer to 3 pixels for long camera 4 6 2 Repeatability The repeatability test is fairly simple set up a configuration like one of the ones listed above take an exposure move one or more mechanisms out of position and then back and then take another exposure The apparent position of spectral lines in both directions should move by less than 10 pixels when the prisms gratings and cameras are moved and by less than 1 pixel when the filter slit decker and focus are moved These effects are at least 3 times larger with the long cameras than with the short cameras so the tests are best done with a long c
39. ee 2 1 1 4 or to use the pupil viewer for the latter a second lens is placed in the beam used in series with the lens in filter wheel 1 GNIRS Users Manual 6 Table 2 3 GNIRS Decker Lengths Decker Position Configuration Usable Length arcsec Spare 1 2 Long Camera Cross Dispersion 3 1 Short Camera Cross Dispersion 6 1 Integral Field Unit Wollaston Prism both scales 14 3 Long Camera Long Slit 49 4 Short Camera Long Slit 99 Acquisition 99 Spare 1 2 Pupil Viewer Not Applicable The partially vignetted lengths of the slits are all approximately 0 5 arcsec longer than the values given above There is also a decker position for the pupil viewing configuration The next element after the slit and decker is the collimator an off axis paraboloid of 1500 mm focal length The collimator mount includes a system of adjustable weights which provide partial compensation for internal flexure in the instrument This is a passive system where gravity acts on a set of weights and levers to tilt the mirror slightly with varying orientation of the instrument The largest corrective tilt of the mirror is less than 7 arcseconds After the collimator a mirror can be inserted in the beam to direct the light into the spectrograph cameras without being dispersed This acquisition mirror allows the observer to identify acquire or recenter objects via broadband imaging without the need to
40. efront sensor guider The spectrograph can be operated in a variety of different observing mode including a choice of 2 pixel scales 3 spectral resolutions different cross dispersion options and an integral field mode The two pixel scales are provided by cameras with different focal lengths The light flow through the instrument is shown in Figure 2 1 OPTICAL SCHEMATIC POM an SLIGHTLY OUT OF FOCUS DEFINES SLIT LENGTH ee WAVE FRONT SENSOR 111 Unio Focus OIWES on oo a ALON PP 8 SLITS COLLIMATOR OFF AXIS PARABOLIC FOCAL LENGTH 1500mm HELP SET IMAGE FOLD A e 21FU positions AT COLOSTOP MIRROR 1 USED COLD STOP OFFNER RELAY FILTER WHEEL IMAGE TELESCOPE TILTED TO AVOID GHOST IMAGE ACQUISITION MIRROR SECONDARY ON REDUCE SCATTERED LIGHT IN BEAM POSITION BYPASSES COLD STOP FROM SLIT Kos PRISM AND GRATING TURRETS REAMAGE TELESCOPE e 2 WHEELS FOR ACQUISITION MODE FOCAL PLANE ON SLIT Sere a a OUT OF BEAM POSITION FOR ar aT SPECTROGRAPH MODE kS PRISM TURRET 0 05 ARC SECIPIXEL LONG 2 CROSS 0 15 ARC SECIPIXEL SHORT DISPERSION AT wOLlASTON CAMERA TURRET SCHEMATIC OF POM 1 FOLD MIRROR Ca CHANNEL OIWFS CHANNEL POM OVERSIZED FOR EASE OF ALIGNMENT amp TO MINIMIZE DIFFRACTION DRIVERS FOR BEAM SIZE SHORTRED SHORT BLUE REO LONG BLUE IMAGE QUALITY AT ENDS OF SLIT MAX RESOLUTION CAMERAS MUST BE PARFOCAL IMAGE QUALITY 85 GEOMETRIC NO SLIT DETECTOR WITHIN TRAVEL OF FO
41. emission components of the background are slow typically on scales longer than reasonable integrations The airglow component can vary by a factor of 2 3 There is a tendency for emission to be higher at the start of the night and then slowly decrease but in addition there are variations on a variety of time scales some of which can be quite short Because of this one should be cautious about planning observations with very long individual exposures since systematic errors in the sky subtraction may result A good rule of thumb for GNIRS until experience with the site and instrument indicates otherwise is to limit individual exposures to 15 minutes GNIRS Users Manual 55 Appendix B Representative Calibration and Night Sky Spectra many lovely spectra appear here GNIRS Users Manual 56
42. en the different cameras and potentially other small focus changes produced by other changes in configuration The detector is a 1K x 1K ALADDIN III InSb array which is operated at a temperature of approximately 31K The detector and its controller can be operated at frame rates in excess of 1 sec allowing operation at 5 um with either camera at any spectral resolution imaging at 5 um for acquisition purposes is probably not possible Individual frames can be co added and then sent to the Gemini Data Handling System DHS to ensure a more manageable data flow For these high background observations the main concern is minimization of overhead since the principal noise source is photon noise from the background At shorter wavelengths especially at higher spectral or spatial resolution detector read noise can be significant even for relatively long exposures In these situations the detector can be read out non destructively so the read noise is reduced by multiple sampling Fowler GNIRS Users Manual 9 sampling For long exposures and low background the improved noise performance more than compensates for the increase in overhead involved 2 1 1 3 Detector and Controller Properties The two detector controller properties that directly affect signal to noise are read noise and dark current Read noise can be reduced by multiple sampling Fowler sampling at the cost of additional overhead in the form of time spent on the ex
43. er should be constant within a few pixels The recommended procedure is as follows Configure for acquisition with the filter camera combination required and the acquisition decker and slit Set up the WFS and start guiding Image the field If necessary take two images with the object displaced use the WFS for the offset and difference them Put in the slit decker combination required for the spectroscopy Take an image and determine the slit center e Using the WFS offset the object from its position measured in the first step to the slit center position As a check you can take two images with the object displaced along the slit and difference them to verify this step Complete the spectroscopic configuration If you are offsetting from a reference object to a very faint target you can either center and check the reference object then offset or combine the offsets Centering checks should be done for long sequences The recommended time interval depends on the camera choice somewhat TBD probably 1 2 hours The procedure that affects the wavelength zero point least is to insert the acquisition mirror and change the filter if desired then check that the object is still on the slit as above For a more accurate check a wider slit should be inserted and one should more or less repeat the acquisition process given above though any offsets should be very small If necessary the camera can be changed although the
44. ering of the object on the slit This effect is predicted to be relatively small with the dominant effect being flexure by the OIWFS mechanisms The shifts on the slit produced by flexure are less than 20 microns for gravity light losses exceed 5 with the long camera for a 10 micron shift which implies that recentering should be done approximately every 2 hours The short camera can tolerate decentering 2 3x larger so one needs to recenter for this purpose only every 4 hours or so Flexure between the slit and the detector leads to smearing of the spectrum in both directions for long accumulated exposures The spectrograph collimator has a passive mechanical compensator that slightly adjusts tilt with varying gravity to minimize this effect The residual flexure is approximately 0 3 pixel hour or less This is very small but for some observations cannot be neglected Specifically Observations requiring very accurate wavelength calibration velocity measurements for example should ensure that the wavelength calibration is an average for the observation not just data from the beginning or end For very long observation sequences several hours it may be useful to group the observations and then shift and add them to retain maximum resolution This decision can be made after the fact by examining shifts in the night sky lines GNIRS Users Manual 24 Observations requiring cancellation of telluric absorptions which require o
45. even for the shortest wavelengths scattered moonlight is not significant The only exception will be for objects located in the ecliptic where moonlight scattered off the telescope itself will cause problems GNIRS Users Manual 20 Air Mass should generally be modest as background image quality and differential refraction all get worse with increasing airmass All calculations assumed an airmass of 1 0 since the standard sky background and image quality specifications refer to zenith 2 2 1 5 Sensitivity Tabulations The following table provides magnitudes referred to Vega 0 0 reached at 5 sigma in 1 hour of integration For low background situations the measurements are assumed to comprise 4 exposures of 15 minutes for high background situations the lengths of the individual exposures are not relevant The values tabulated are all for the long slit mode i e with no prisms present These are for a resolution element 2 pixels and extraction of an optimal window but not optimal i e weighted extraction Table 2 12 GNIRS Sensitivity 5 sigma in 1 hour 20 percentile IQ Resolution Background Order Scale 6 x SCJ 4C H 3 CK 2 CL 1CM 1700 0 05 ave 21 50 21 42 20 36 19 43 15 28 12 03 90 21 23 21 13 19 96 19 11 14 90 11 59 5100 0 05 ave 20 09 19 97 19 24 18 47 14 35 11 21 90 19 98 19 84 18 94 18 20 13 97 10 78 17800 0 05 _ ave 19 16 19 04 18 54 17 77 13
46. ined by two mechanisms The width of the slit is defined by the one of several slits in a photo etched mask located in the slit slide while the length of the slit is defined by one of several openings in the decker slide The integral field unit IFU is also mounted in the slit slide there is a location for a second IFU unit currently occupied by a dummy module of similar mass The slit mask is located at the re imaged focal plane while the decker apertures are slightly ahead of it and therefore somewhat out of focus by a few pixels The decker sizes are matched to the full width of the array in long slit mode or to the minimum spacing between adjacent spectra when the prisms are used The slit mask in the instrument can be changed although it should not be considered a routine operation The slit widths currently available are listed below as are the decker lengths Tables 2 2 and 2 3 Table 2 2 GNIRS Slit Widths Slit Width Mask Position arcsec short camera long camera pixels pixels l dark 2 3 0 20 60 3 1 0 6 7 20 4 0 20 1 3 4 5 0 15 1 0 3 6 0 10 0 7 2 7 acquisition Figure 2 2 8 0 30 2 6 9 0 45 3 9 10 0 60 4 12 Widths in pixels are for lowest resolution grating and the acquisition mirror Projected widths with the 32 I mm and 110 I mm gratings are reduced by 5 and 22 respectively In addition the slit slide can be positioned to use the integral field unit s
47. ition mirror position to limits open closed and in out so limit messages on these two mechanism are not errors Hard limit OT messages are errors on all mechanisms including these two Note also that the rotary mechanisms turrets and wheels do not have mechanical limits so any limit message is an electrical or software problem not a mechanism problem If the mechanism gives an error on doing a datum try with the back up home switch and or do a datum test see service manual GNIRS Users Manual 50 Appendix A Supplementary Information for Exposure Time Calculations A 1 Throughput Throughput is affected by two main factors instrument throughput which is a function of configuration and slit losses which are a function of slit width and image quality Telescope and atmospheric transmission are also factors A 1 1 Slit Losses Slit losses are calculated assuming a gaussian image profile centered on the slit Detailed calculations show that a small amount of decentering does not affect throughput greatly The table below A 1 gives the percentage of light transmitted through the slit for both 2 pixel and 3 pixel slits for both pixel scales for the 20 50 and 85 percentile seeing cases listed in Table 2 11 The values listed are for the K band there are slightly greater losses at shorter wavelengths and smaller losses at longer wavelengths The highlighted values are those used for the calculations summarized in secti
48. lass in the cross dispersion prisms absorbs significantly in the K window The absorption averages about 80 between 2 0 and 2 4 um GNIRS Users Manual 53 A 2 External Background There are two main contributors to the external background sky background and emission from the telescope itself including warm parts of the instrument A 2 1 Average Background The sky background includes airglow which is dominant at shorter wavelengths thermal emission from absorbers at lower altitude which is important at wavelengths beyond about 2 4 um and some amount of continuum The telescope emission is essentially a gray body proportional to the effective telescope emissivity which is typically lt 10 For a well baffled instrument such as GNIRS it is close to 1 minus the telescope transmission given in Table A 3 As noted in 2 2 1 3 the structure in the night sky emission means that one needs to use caution in applying average flux values to exposure time estimates In the absence of detailed night sky spectra taken with GNIRS the following procedures have been applied Start with broad band sky brightness estimates Compare with model night sky emission spectra taken in this case from the Gemini web site e Apply corrections for non standard filter pass band and for flux distribution with wavelength This allows one to estimate both an average flux value per resolution element as well as flux levels at various percenti
49. les in the distribution It turns out that the average values correspond roughly to the ice percentile for the different windows i e 75 of the resolution elements in the window have sky fluxes less than or equal to the average In addition the flux values corresponding to 90 percentile are typically 2 3 times the average For the calculations presented above 2 2 1 a uniform factor of 2 3 has been applied Table A 5 summarizes the sky background values applied Note that the fluxes do not include the corrections applied for atmospheric absorption or telescope and instrument throughput The values tabulated are for R 1700 values at other resolutions are inversely proportional to the resolution GNIRS Users Manual 54 Table A 5 Mean Model Night Sky Emission Resolution 1700 Grating Order Background Background mag arcsec2 photons sec arcsec2 resolution element 1 M 0 2 91 300 000 2 CL 5 4 890 000 3 K 13 2 1560 4 H 14 1 1100 5 J 16 35 206 6 x 16 6 215 A 2 2 Variability The background will vary for several reasons The telescope background will change with telescope temperature emission at 2 4 um will change by a factor of 3 or so over a 20 degree C temperature variation which is more or less characteristic of seasonal extremes The thermal emission component of the atmospheric emission will also change by similar amounts In general variations of any thermal
50. m to ensure that the structure is maintained at a constant temperature 1K or better The window at the front of the cryostat has a motor operated cover that is closed to protect the window when the instrument is not in use As it is an external warm device it cannot be used as a dark slide There is a manual override on the cover so that the window can be protected even in the event of a power failure The window can cool below the dew point in conditions of high humidity so the window mount provides a slow flow of dry air across the window to avoid condensation This also helps to keep dust off the window GNIRS Users Manual 13 Figure 2 5 Cutaway view of instrument assembly showing cold structure inside cryostat electronics boxes and trusses The instrument is shown mounted on a face of the Gemini ISS Note that the bench structure is upside down compared with Fig 2 2 2 1 5 Electronics Enclosures As with all Gemini instruments the GNIRS electronics are mounted on the instrument to facilitate instrument changes and minimize the complexity of the connections between the instrument and telescope The electronics are mounted in two thermal enclosures which are insulated glycol cooled boxes that minimize heat dissipation from the electronics into the telescope environment One of the enclosures lower right in Fig 2 5 contains the science array controller which handles operations of the ALADDIN array including initial pr
51. me as Fig 2 6 but for the K band Note the rising thermal background beyond about 2 4 um GNIRS Users Manual 18 The situation in other bands and at other resolutions is generally similar although it is the case that at longer wavelengths there is more continuum flux so maximum contrast is less Also at higher resolutions the spectrum increasingly breaks up into resolution elements with and without lines at R 18000 roughly 90 of the spectrum does not contain strong lines There will still be enough lines for a reasonable wavelength calibration It is clear therefore that a sensitivity estimate based on the average background within a spectral window is not necessarily that useful though it is easy to calculate Therefore for this manual 90 percentile sensitivity is also calculated which is the value where 90 of the resolution elements in the spectrum will have greater than the specified signal to noise and 10 will have less It turns out that due to the nature of the night sky spectrum average background is roughly equivalent to 75 percentile background The ITC should be used when available as this provides the most accurate estimates This is particularly true when what is of interest is a specific feature or features rather than a complete spectrum 2 2 1 4 Standard Observing Conditions Since many Gemini observations are taken in queue mode where observing conditions are specified it is necessary to indicate which
52. mode If two or more windows are desired then the loss of sensitivity due to limited slit length limited number of different dither positions is normally offset by the ability to observe several orders at once e IFU vs slit The IFU is the only efficient way to get 2 dimensional coverage so it is clearly the choice when observing objects with structure or in very crowded fields It is possible that in poor seeing the IFU could be used as an image slicer for point source observations provided only a single atmospheric window is of interest otherwise the multiplex gains from cross dispersion should offset the gains from image slicing 2 3 2 Observation Configuration Trades This section discusses guidelines for optimizing one s observations in terms of number of exposures Fowler sampling and dithering positional sampling The emphasis is on measuring faint objects for bright objects total time is dominated by overhead so one should mainly ensure enough measurements are made at slightly different positions to provide adequate sky subtraction and flat fielding For faint objects one would like to measure at different positions along the slit ideally enough so that the noise contribution from sky subtraction is negligible see 2 2 1 At some point the number of positions required will become too large to fit along the slit and the overheads involved relative to the decreasingly short exposure times will also lead to diminishing re
53. n both polarization yet it is not clear that there is an appropriate unpolarized flat field to be used for this purpose It may be that the only solution is to rely on standard stars as unpolarized light sources and accept the fact that the flat field normalization in the two polarizations may be somewhat different GNIRS Users Manual 40 3 5 Preliminary Data Reduction The following steps are recommended to carry out preliminary data reduction They may in fact suffice for final data reduction as well but don t delete raw data until you have had the chance to assess the reduced data calmly Linearity correction Dark subtraction Flat fielding Sky reference generation Sky reference subtraction Spectrum extraction Absorption standard division optional Flux calibration optional The first three steps remove the detector signature the second two steps remove the night sky and telescope emission while the final steps produce a spectrum with the instrumental signature removed 3 5 1 Linearity Correction Dark Subtraction and Flat Fielding These procedures are essentially automatic if there are always matching dark frames and flat fields The linearity correction is optional if all count levels are low below 10 000 TBD ADU If used the recommended coefficients for the IRAF procedure irlincor are a 1 0 b xxx c yyy TBD actually will need separate values for each bias value See the IRAF documentation for an expl
54. n unit GCAL The flat field produced by this unit is flat to 1 TBD over the longest slit used in GNIRS which is adequate for nearly all programs For more accurate calibration of the long slit the best alternative is building up a flat using twilight sky exposures or daytime if the background is not too high The sky flat should be viewed as a supplement to the GCAL flats and will not be needed for most programs For exposures done using GCAL a flat field determination requires measurements with the light source on and off which are differenced For details of the operation of GCAL consult the relevant manual REFERENCE The flat field is relatively independent of wavelength so a separate flat field for each wavelength setting is not a necessity if there are many similar settings It is preferable though to have a flat field associated with each setting This is especially important in the low resolution and prism modes where a given configuration will not produce signal on the entire array This means that portions of the image will need to be masked off to produce a reasonable flat field which will not necessarily be useful for another configuration It is convenient to divide each flat field by an average value to normalize the corrections Darks and flat fields should be produced using the average or median of a large number of frames typically 20 50 for each dark time or flat field configuration GNIRS Users Manual 3
55. n wavelength units is nearly constant for a given order regardless of tilt The quoted resolutions are all for 2 pixels at the detector specified as AA The detector is 1024 x 1024 pixels so there are roughly 512 resolution elements in the dispersion direction For the R 1700 mode this corresponds to coverage AA A of roughly 30 From the grating turret light then passes to the camera turret The camera turret contains four cameras Table 2 6 GNIRS Cameras Camera Wavelength Range Focal Length Pixel Scale Long blue camera 0 9 2 5 um 1305 mm 0 05 arcsec Long red camera 2 9 5 5 um 1305 mm 0 05 arcsec Short blue camera 0 9 2 5 um 435 mm 0 15 arcsec Short red camera 2 9 5 5 um 435 mm 0 15 arcsec The blue cameras will not work at longer wavelengths the red cameras can be used at shorter wavelengths but with somewhat degraded image quality and transmission The main short wavelength use of the red cameras below 3 um is for acquisition of targets in the K band sorter 3 See 3 1 4 details All four cameras are close to parfocal the longer focal lengths are achieved by folding the beam with a combination of mirrors in the camera barrel and external to the turret The light path shown in Figure 2 2 is for one of the long cameras so one can see the folded light path The detector is mounted at the output of the cameras on a focus stage The focus stage provides correction for the small focus differences betwe
56. nd to verify centering during the course of long observing sequences The sensitivity of the acquisition mode should be comparable to that of NIRI in its f 6 configuration Note though that sorter 3 K band extends to longer wavelengths than a conventional K filter and will therefore have higher background and lower sensitivity by a few tenths of a magnitude It is not necessary to acquire the object with the filter through which you intend to take a spectrum but allow for differential refraction see 3 1 3 Although the red cameras don t perform well below 3 um it is possible to center objects with these cameras using the H or K band order sorters It is also possible to acquire using the equivalent blue camera center the object then change camera and filter Because the acquisition mirror is inserted without disturbing any of the dispersing elements it is possible to check centering on long observing sequences without losing the wavelength zero point Changing the slit or camera during the check will change the zero point GNIRS Users Manual 31 For sufficiently faint objects it may be desirable to center on something brighter nearby and then perform a precision offset with the WFS to center the object itself In all cases it is important to realize that flexure and positioning repeatability are such that the object should be centered relative to the slit itself and not on an absolute position on the array though the slit cent
57. ocessing steps such as non destructive reads and co addition The second enclosure upper left in Fig 2 5 contains all the other instrument electronics including the OIWFS electronics instrument motor and temperature controls and the instrument s VME crate GNIRS Users Manual 14 The instrument fits barely within Gemini s allowed instrument envelope and is approximately 2 2m long x 2 5 m wide x 1 3 m high side looking orientation The instrument weight is just under 2 metric tons ballast is added to bring its weight up to 2000 kg to balance it against other instruments on the Gemini instrument support structure GNIRS Users Manual 15 2 2 Science Channel Performance The performance of the science channel depends on the instrument configuration and details of the observations In general though all observations with GNIRS can be thought of as comprising a measurement of the object and a measurement of the background which are then differenced The signal is the flux from the object minus any light losses slit losses absorptions and reflections etc converted to detected electrons The noise is the combined effects of photon noise from the objectt sky photon noise from the subtracted background photon noise from dark current and internal background and detector read noise Summary sensitivity tables are given below 2 2 1 for some standard configurations and conditions details are provided in subsequent secti
58. ogress an error message will be generated You will need to use the apply button again when the current activity stops Note that there is no common sense filter applied to the configurations requested In particular if you are making major changes in the configuration check the following Make sure that only one filter is in use that is if you select a filter in Filter Wheel 2 make sure Filter Wheel is on open except for diagnostics Make sure that the camera corresponds to the filter going from xJHK to LM requires going from the blue to the red cameras and vice versa e Ifyou change pixel scale make sure you change both the slit and the camera Remember also that spectral resolution changes by a factor of 3 when pixel scale changes Remember also that the mechanisms are not perfectly repeatable so once you change from a given configuration you will not be able to go back to it exactly but you will be within 10 pixels or less Don t forget the window cover needs to be open Consult section 2 for useful configurations Tables 2 8 2 10 and elsewhere 3 2 2 OIWFS Configuration Configuration of the OIWFS is normally done through the telescope s acquisition and guiding system and is done by the telescope operator rather than the instrument user It is possible if needed to access the OIWFS from the top level screen Fig 3 1 by opening the OIWFS top level screen and then the O WFS CC screen 3 2 3
59. on 2 2 1 The sampling in the IFU corresponds to a slit width of 0 15 arcsec One should recognize that a spatial resolution element is effectively two samples which should be combined in considering overall signal to noise Table A 1 GNIRS Slit Transmission as a Function of Seeing Slit width arcsec amp 20 Percentile 50 Percentile 85 Percentile pixels 0 30 arcsec 0 50 arcsec 0 75 arcsec 0 10 long 2 pixels 31 19 13 0 15 long 3 pixels 44 28 19 0 30 short 2 pixels 76 52 36 0 45 short 3 pixels 92 71 52 The signal is directly proportional to the slit transmission so one can easily estimate sensitivity under difference image quality conditions The tables in section 2 2 1 for the short cameras provide estimates for 50 percentile image quality For 20 percentile image quality the ratio of slit transmission is 1 46 resulting in a sensitivity gain of 0 4 mag The loss in sensitivity at 85 percentile image quality is also 0 4 mag In those cases where the object is much fainter than the sky in the slit the integration time is roughly proportional to the square of the slit transmission hence one gains a little over a factor of 2 going from 50 to 20 percentile image quality Widening the slit to increase transmission also increases background For the case of faint objects going from 2 to 3 pixels increases background by 50 while increasing signal by a lesser
60. ons and yet more information in Appendix A In general the Gemini integration time calculator ITC should be used when it becomes available for best estimation However the following sections and the appendix can be used until then and also serve to show the considerations that go into the ITC 2 2 1 Sensitivity Summary 2 2 1 1 Baseline Observing Sequence For observations of objects comparable in size to the slit length or IFU dimensions the background observation is actually a separate observation taken by moving away from the object onto blank sky For observations with a small object or a long enough slit it is possible to position the object successively at different positions on the slit and thus use a single observation for both object measurement and background measurement For example if one can measure 5 different slit positions at any one position there is one object measurement and 4 background measurements which can be combined In the limiting case of very many slit positions the noise contribution from the background determination become negligible and has taken no additional time Compared with the simple on off case the multi position observation will take roughly 1 4 the time to achieve the same signal to noise This discussion assumes that it is possible to provide the same type of spectral extraction for the different cases If doing a smaller number of positions provides enough signal to noise for optimal ex
61. or fainter guide stars fast correction is limited to tip tilt and does not include focus The OIWFS filter wheel contains standard JHK filters In principle users should select the filter that provides the best signal to noise on the guide star normally H since the telescope s acquisition and guide A amp G system adjusts the OIWFS gimbal mirror position to compensate for differential refraction between the guide wavelength and the observing wavelength See section 2 4 1 for OIWFS performance and 3 2 2 for information on O WFS operation 2 1 4 Cryostat The cold structure shown in Figure 2 2 is contained within a larger cryostat shown in cutaway view below Figure 2 5 The internal structure is maintained at a temperature of approximately 60K using 4 Leybold RG 5 100 cryocoolers The initial cool down of the instrument can be done either with the cryocoolers alone which takes over a week or with the assistance of a liquid nitrogen pre cool system which allows the instrument to reach operating temperature in 3 days The cryostat is a large vacuum vessel which contains two aluminum passive shields and a third shield that is connected to the cryocoolers These act to minimize heat radiated and conducted into the cold structure which would otherwise produce unacceptable temperature gradients and gradient variations under varying ambient conditions In addition a temperature control system adds varying amount of heat to the cooling syste
62. pidly If so repeat to see if images return to normal Excess background observed Has the instrument cooled sufficiently The optics will not fully cool for 12 hours after the bench itself initially reaches 60K Something is in the way dome PWFS etc Vignetting should produce excesses only at KLM Pupil viewing may help find the culprit if only a long wavelength excess is seen Temperature health abnormal yellow or red GNIRS Users Manual 49 Open the temperature display screen If only one sensor is reading abnormally the diode may be bad A very low temperature 23K is an indication of an open circuit A single temperature that is slightly out of range should be monitored If multiple temperature readings are out of range check that the bench temperature control is on temperatures low or that the cryocoolers are functional temperatures high check operation also check individual head temperatures for high values Check that dewar pressure is within range poor vacuum will conduct excess heat into the cold structure In this case the pressure health should also be bad probably red Mechanisms give errors on positioning or datuming If the system has been rebooted or reinitialized you must redatum all the mechanisms before you can configure anything If the mechanism gives an error on going to a position for example hitting a limit redatum and try again Note that the window cover and acquis
63. r affects much of the L and M bands and the edges of atmospheric windows near water bands You should investigate the exact circumstances of your observations look at atmospheric spectra in the ITC to determine whether drier conditions will make enough difference to request them See Figure 2 7 below for an example of calculated transmission for dry conditions Water vapor at Gemini South has a strong seasonal dependence so observations requiring dry conditions will be far more difficult in the summer essentially 20 percentile conditions will occur much less than 20 of the time then Because the affected regions typically contain strong saturated lines observations that require dry conditions will be best done with the driest conditions realistically available Transmission oO h z I T es ee ee Se Sas I r 1 0 12 1 4 1 6 13 2 0 2 2 4 2 6 Wavelength ftm Figure 2 7 Calculated transmission for 0 9 2 7 um for 1 6 mm precipitable water on Mauna Kea This corresponds to very best conditions on Cerro Pach n One can see which regions might be usefully observed in dry conditions but not otherwise Sky Background in the near infrared effectively divides into night 80 percentile and twilight Background is both higher and variable during twilight so observers will seldom gain much by trying to observe at these times All calculations use standard night sky conditions 2 2 4 Note that there is no reason to request gray time
64. ration of the array Saturation of the array occurs relatively rapidly in the M window order 1 where exposure times under a second will be required with the short cameras to avoid saturation For these observations one would co add multiple short exposures at each slit position to keep the quantity of data manageable Because sky subtraction is critical one would want to limit the time at each position to tens of seconds or a few minutes at most Experience with actual conditions at Gemini will be helpful here One would then cycle through the slit positions multiple times to accumulate signal to noise For high background situations the detector bias can be increased to increase the well capacity The read noise is also higher at higher bias so it should be used only in these high background situations basically any case where you are limited to exposures of seconds or less Always use the same bias for the same configuration In the next order between 3 and 4 um backgrounds are substantially less but still high enough to limit exposures to under a minute at low spectral resolution Depending on the spatial and spectral resolution one might co add a small number of images but still keep time at any slit position to a few minutes at most and cycle through the positions as needed At the lowest resolutions use of the high bias may be required TBD At still shorter wavelengths exposure times could be tens of minutes or even hours
65. re is correct low enough Configuration incorrect Did you forget to configure a mechanism including the window cover Set the configuration again and see if mechanisms move observe again if they do If all mechanisms are alleged to be in the correct position do a global datum and reconfigure If this does not solve the problem try a configuration with the acquisition positions Further mechanism diagnosis is covered in the service manual Noise on image This may be due to grounding problems Also consider Was the OIWFS being reconfigured during the readout If so try again and see if the noise is absent when the OIWFS mechanisms are not active Image does not appear to be the correct configuration Check Did you make a mistake in selecting the configuration especially grating wavelength If one mechanism appears to be the source of the problem try a repositioning it and if that does not work try b datum and re position If this is not correct and especially if the configuration is still wrong but different try mechanism diagnosis procedures in the service manual Ifthe mechanism positions correctly it may be worth checking repeatability move away then back to confirm reliability Images are fuzzy or otherwise bad Configuration is otherwise correct Check focus setting Check instrument temperature Was the telescope slewing or was the instrument rotator changing angle very ra
66. re one wants to observe both an object and a reference star in the same identical configuration probably so as to divide out telluric absorptions In this case it would be possible to set the grating and measure orders 3 6 on the object by changing filters and then go to the reference star and the repeat the same sequence If on the other hand one wanted to observe at several grating tilts perhaps at higher resolution one would have to observe the object at one tilt then the reference star then change the tilt then observe the reference and the object at the new tilt and so on Any calibrations that cannot be done later need to be done as well GNIRS Users Manual 25 2 3 Observing Mode Trades 2 3 1 Instrument Configuration Trades This section discusses alternative configurations for some types of observations The discussion is general so it is useful to compare situations for specific programs using the ITC e Long vs short cameras The long cameras will give better performance only when image quality is good enough so slit losses don t exceed the gains due to reduced background Unless maximum spatial or spectral resolution is required the short cameras will probably be preferred Note that for a resolution of 18 000 the long camera must be used e Long slit vs cross dispersed Cross dispersion provides cover of multiple orders but at the expense of slit length If only one atmospheric window is of interest use long slit
67. rument configuration you require spatial and spectral resolution filters grating tilts and the like This section considers some additional aspects of the observations that need to be specified prior to carrying them out These comprise Guide Stars e Standard Stars Slit Orientation 3 1 1 Guide Stars Observations with GNIRS cannot be carried out efficiently without a suitable guide star The only probably exception to this is observations of bright standards where exposures are short and the guider set up time will be substantially longer than the added exposure time required to compensate for lack of a guide star This is a reasonable assumption but it should be verified and amplified after we go to the telescope If a guide star suitable for the OIWFS can be found see 2 4 this is preferred since the OIWES provides both tip tilt correction and full compensation for instrument flexure relative to the telescope Failing this a PWFS guide star will provide good compensation though the tip tilt correction will be less effective because of the greater distance between the guide star and the object 3 1 2 Standard Stars Two types of standard stars may be needed flux standards and telluric absorption standards Flux standards are needed if you intend to measure relative fluxes at different wavelengths or absolute fluxes A flux standard is not needed if you are only measuring GNIRS Users Manual 29 the intensi
68. slightly over 0 2 arcsec at a zenith distance of 45 degrees airmass 1 4 and is over 0 4 arcsec at a zenith distance of 60 degrees airmass 2 0 If the dispersion is perpendicular to the slit particularly in good seeing the losses at the extreme wavelengths can be considerable Ideally the slit should be oriented in the same direction parallactic angle as the refraction or an average over the observation in which case the spectrum may be slightly skewed but no light will be lost For point sources this is not a problem but for objects where a particular orientation is desired for scientific reasons there can be a conflict Several approaches are possible to mitigate the problem e Plan the observation so the parallactic angle and the desired position angle on the sky more or less coincide This may involve observing the object while rising or while setting The desired angle may not always be such that this can be done e Plan the observation so that observations are done only near zenith Use the short camera instead of the long camera Don t work in cross dispersed mode and center the object in each order of the grating for the observation in that order The dispersion across any one order is less than 0 1 arcsec for reasonable zenith distances so slit orientation is far less critical 3 1 4 Acquisition and Centering The acquisition mode of GNIRS allows observers to find objects and center them on the slit efficiently a
69. spectrographs There are two main sections a fore optics section which provides field and pupil stops to limit excess background and a spectrograph section which disperses light from the object of interest GNIRS Users Manual 3 2 1 1 1 Fore Optics The fore optics comprise the following elements Entrance window Pick off mirror Entrance fold mirror Offner relay primary and secondary mirrors Exit fold mirror Filter wheels The entrance window also acts as a weak lens in order to ensure that the telescope secondary is imaged on the Offner secondary mirror which is where the cold stop is placed The pick off mirror acts as a crude field stop defining the field accessible to the spectrograph Light from the rest of the instrument field roughly 3 arcminutes diameter is available in principle to the OIWFS see 2 1 3 The unvignetted field defined by the pick off mirror is basically a 10 x 100 arcsecond strip with a superposed half circular field 15 arcseconds in radius centered on the optical axis as shown schematically in Figure 2 3 15 radius 10 E a 100 Figure 2 3 GNIRS spectrograph field of view The long dimension is 100 arcsec the width is 10 arcsec except for the additional 15 arcsec semi circle The purpose of the semi circle is to provide a somewhat larger field for target acquisition and identification while at the same time allowing use of guide stars close to targets of interest Th
70. t of the number of co adds and the number of images The only other parameter that may need to be changed is the array bias which is changes from the array set up screen not shown which is also accessed from the main screen Use bias of 0 3V for low background and bias of 0 6V for high background The low bias has fewer hot pixels but less well depth Do not change any other voltage on this screen In order to start an exposure one needs to use the main window shown above in Figure 3 3 The observe button the top level sequencer screen not shown has the same function This window has only one function in routine use which is starting an exposure or exposure sequence Note that with the appropriate configuration see above it may produce several images that are sent to the DHS It is possible to change the exposure time from this window as well as the observing set up window If one is doing a raster pattern of some sort for example multiple positions along the slit the observing sequence would consist of configuring the instrument and the array controller positioning the object to the first location clicking on the observe button waiting for the exposure s to finish offseting the telescope and OIWFS clicking on the observe button and so on Normally this is done automatically through the sequence executor and the screen shown above will not be used GNIRS Users Manual 37 3 3 User Interface The user interface to
71. ted and added For faint objects one can add images taken at the same slit position then do the extraction Note that distortions and tilts require one to determine an extraction profile even in long slit mode where the spectrum runs approximately along the rows of the array 3 5 4 Standard Stars The summed extracted spectra in many cases provide enough information to see whether the observation is successful Even if this is true it is still often useful to remove the atmospheric and instrumental signatures The following steps will produce an approximate calibration but not an optimal one Divide by a standard star extracted spectrum Multiply by an artificial curve having the nominal flux distribution of the standard For most purposes this can be a blackbody fit to the standard s broadband magnitudes This procedure will remove weak atmospheric absorptions but not strong absorptions It will correct for instrumental response It will also introduce spurious emission features wherever the standard has absorption features The result is also uncorrected for differential light losses at the slit GNIRS Users Manual 42 4 Set Up and Operation This section describes the procedures required to verify that the instrument is working and to get ready to observe with it This section assumes that the instrument has been properly installed on the telescope see the Service and Calibration Manual for installation procedures Th
72. tion all the mechanisms to the initial configuration that you want Important The prism turret in the zero position will partially vignette the beam Therefore even if your initial configuration is with the acquisition mirror you should position the prism turret to one of its four positions or to park GNIRS Users Manual 43 4 3 Initialize Detector The detector is initialized from the GNAAC Engineering Interface Individual detector parameters may need to be adjusted from subsidiary screens see 3 2 3 The most useful parameters to adjust are the number of digital averages low noise reads and number of co adds and images The region of interest area to be read out can also be modified as well as other array control parameters The bias is changed when one switches between low background small full well low noise and high background large full well higher noise In general it is advisable to leave most parameters unchanged as this greatly simplifies subsequent data management and data reduction The array is initialized using the three buttons on the lower right of the GNAAC interface screen 4 4 Initialize OTWFS The OIWFS is started as part of the instrument start up procedure described above Any further initialization e g datuming mechanisms is done by the A amp G system These procedures are fairly slow 15 minutes so make sure they are carried out promptly You can initialize the OIWFS if necessary from its
73. total integration time should be allowed for these purposes This will increase for longer wavelength observations especially M band where it may approach 50 Note that the calculated observing time required for standards will approach half an hour for typical observations even though the actual integration time will be a few minutes The manual should include an estimate of time requirements for standard star observations once we have experience at the telescope 2 2 1 3 Performance Completeness The night sky background is not a smooth function of wavelength as demonstrated in Figures 2 6 and 2 7 which show low resolution night sky spectra GNIRS Users Manual 17 NOAD IAAF 2 10 SBETA aS ater Men 13 45 46 23 Jan 95 Sky papara tgn At Bend ai s4 1s T T T T T 25 2D 15 Sia UL 10 o f l l 1 4 1 5 1 6 1 7 1 8 Wavelength micrans Figure 2 6 CRSP H band night sky spectrum at resolution 1400 Replace with R 1700 GNIRS spectrum when we have one At this resolution the background varies by well over an order of magnitude between different wavelengths in the band As a consequence signal to noise will also vary by substantial factors possibly approaching a factor of 10 in this case NOAD TIAAF 2 10 SBETA tahoe UP man Mon 13 48 09 23 Jan 95 aration st Sky aoeparaitg K Bond at s4 1s 25 20 15 Sia UL 10 D l l l 2 2 1 2 2 2 3 2 4 2 5 Wavelength fImicrans Figure 2 7 The sa
74. tra reads In addition the effective well size limits how much signal can accumulate The relevant properties are tabulated below for the GNIRS ALADDIN III array Ser 410793 Table 2 7 GNIRS Detector Properties Read Noise single read 37 electrons Read Noise max useful multiple reads 7 electrons Single Read time 0 185 sec Max Useful Read Time 36 sec Mean Dark Current 0 1 electrons sec pixel Well size lt 10 non linearity 110 000 electrons Gain 13 electrons ADU Further details on this subject are found in section 2 3 1 2 1 1 4 Integral Field Unit The integral field unit IFU is an additional optical system provided for GNIRS by the University of Durham UK The IFU takes a rectangular input field of approximate dimensions 3 3 x 4 8 arcsec and divides it into 22 slices 0 15 arcsec in width The IFU optics see Figure 2 4 eventually map the slices of the rectangular field onto the input plane of the spectrograph aligning the slices more or less along the regular input slit position the slices are offset from each other by roughly 2 pixels slice The optics also change the input scale to 0 12 arcsec pixel along the slit and 0 075 arcsec pixel in the dispersion direction The IFU is intended to feed the short cameras and therefore can be operated at a maximum resolution of 5900 Figure 2 4 here Figure 2 4 Integral Field Unit Optics Layout GNIRS Users Manual 10
75. traction but this is not possible for a large number of positions then the smaller number may be better Also for the shorter slits cross dispersed modes in particular there may not be enough positions except in very good seeing Again use of the ITC should help in these decisions All the tabulations below assume that data were taken at a large number of positions along the slit Therefore for objects comparable in size to the slit observation times must GNIRS Users Manual 16 be increased by a factor of 4 For objects where only a small number of positions can be observed 2 or 3 there is a smaller increase in the time required 2 2 1 2 Overheads The tabulations do not include any allowance for overhead This should be allowed for explicitly The following rules can be applied provisionally Initial set up This includes moving the telescope to the object acquiring the guide star in the OIWFS acquiring the object on the slit and configuring the instrument Allow 15 minutes Instrument reconfiguration This occurs when the instrument configuration is changed but the object to be observed is the same For most changes filters grating tilt the time required will be lt 1 minute General overhead This includes allowance for moving the telescope between positions on the slit or on and off detector readout and write time periodic checks of object centering for long observation sequences 10 of the calculated
76. turns For most applications 5 different positions is a reasonable compromise For the cross dispersed long camera where the usable slit length is 3 arcsec Table 2 3 it may not be possible to use this many positions unless image quality is quite good 20 percentile or better GNIRS Users Manual 26 Alternatively for programs requiring very high signal to noise flat fielding may prove to be a limitation and additional positions presumably in long slit mode would then be useful A second set of choices relates to the read out mode used whether one does a single read or multiple reads Note that a exposures actually consist of an array reset an initial read out integration and then a final read out from which the initial read is subtracted The read noise given in Table 2 14 is for such sequences The use of multiple reads decreases the effective read noise by up to a factor of 4 but at the price of increased overhead The difference in overhead between the minimum number of reads and the maximum useful sampling is about 4 seconds so if one is in fact read noise limited multiple sampling is justified for any exposures longer than a few seconds In practice observations below 2 5 um will benefit from multiple sampling and observations at longer wavelengths will not The maximum exposure time to be used is set by one of three considerations the total integration time required for the object sky variability and satu
77. ty of individual features as equivalent widths whether in absorption or emission Normally a flux standard is used to produce a curve of response vs wavelength for each configuration If only narrow slits are used this is affected by slit losses and potentially by differential refraction see 3 1 3 Wide slit observations of both objects and standards can be used to convert observation to an absolute flux scale true spectrophotometry There are really no spectrophotometric standards in the near infrared Instead observers customarily use photometric standards with spectral types that indicate they should have relatively featureless spectra The broadband magnitudes are then fitted to produce a smooth flux vs wavelength curve for the stars Provisionally use the LCO standards which are mostly G type in the magnitude range 10 12 Flux standards are not necessarily measured at the same time and airmass as the objects they are calibrating and therefore will be affected differently by absorptions in the Earth s atmosphere If one s program looks only at features in relatively clean regions of the atmospheric windows such as the centers of the H and K bands dividing by a standard will provide adequate correction For messier regions it is useful to observe a correction star at the same time at a location near the object and then construct an atmospheric absorption correction normalizing the clean portions of the stellar continuum Th
78. without danger of saturation so the primary limitation is instead sky variability A 15 minute exposure time would allow one to cycle through 5 slit positions in under 90 minutes allowing for overheads which is a reasonable limit if there is still a need to subtract sky emission lines For higher resolution especially R 17800 it may be possible to ignore the lines in which case somewhat longer exposures would be reasonable GNIRS Users Manual 27 2 4 WES Performance For guiding either the On Instrument Wavefront Sensor OIWFS or the telescope s Peripheral Wavefront Sensors can be used The OIWFS is preferred if a suitable star is available see 3 1 1 2 4 1 OIWFS Performance The OIWFS performance can be calculated from first principles or can be assumed to be very similar to the performance of its equivalent in NIRI Both assumptions lead to similar predictions For use with the Shack Hartmann prism which is permanently installed the 5 sigma magnitude per spot for a 10 msec integration is approximately magnitude 13 8 for J or H and 13 1 for K This calculation is for 50 percentile image quality Performance for 20 and 85 percentile image quality is estimated to be roughly 0 7 mag better worse respectively Since image quality is slightly better at H this is probably the best choice for OIWFS filter under most circumstances In unobscured regions stars of the required brightness correspond roughly to R lt 16 Th

GNIRS Users Manual - National Optical Astronomy Observatory

Contents

Download Pdf Manuals

Related Search

Related Contents