Home

Old User`s Manual - WM Keck Observatory

1. clean no Detect and replace bad pixels skybox 1 Box car smoothing length for sky saturat INDEF Saturation level readnoi 4 Read out noise sigma photons 58 HIRES Manual Data Reduction gain 2 38 Photon gain photons data number lsigma 4 Lower rejection threshold usigma 4 Upper rejection threshold functio spline3 Fitting function for normalization spectra order 3 Fitting function order sample Sample regions naverag 1 Average or median niterat 3 Number of rejection iterations low_rej 3 Lower rejection sigma high_re 3 High upper rejection sigma grow 0 Rejection growing radius mode ql If you do it interactively as you should the first time through you will have the opportunity to play with all the fitting parameters Try to use the lowest order spline as possible when fitting out the quartz s low frequency variations or you will risk introducing ripples into your flat fielded spectrum Again be careful how you answer questions about resizing and editing apertures etc If you want to stick with the apertures you just found in apall be sure not to find recenter resize or edit the aperutres When finished you should do a sanity check using the task display to display flat imh and visually inspect for unexpected results etc You should have nice uniform intensity quartz spectrum orders with lots of clear dark space in between If you
2. 0 7 TRANSMISSION HIRES Manual Instrument Description Figure 4 Red Blocking Filters e NI tor trimmer o o a 0 3 0 2 0 1 0 3 04 05 06 07 08 0 9 10 11 12 WAVELENGTH microns 19 REFLECTIVITY Instrument Description HIRES Manual Figure 5 Collimator Reflectivities x red collimator enhanced silver blue collimator 2 layer enhanced Al 0 3 0 4 0 5 0 6 0 7 20 08 09 1 0 WAVELENGTH microns 1 1 HIRES Manual Instrument Description thermally insensitive provided the environment is isothermal i e it will remain aligned at any temperature provided the temperature is stable The echelle is mounted in a precision rotation stage Rotation of the stage allows the echelle format to be positioned as desired left right on the CCD and looks to the user as though one is moving the CCD left right around a fixed echelle format The echelle mosaic is housed in a dust tight enclosure Since these gratings can never be cleaned one should never attempt to touch them or even to get near them And their cover should be kept closed when not in use Cross Disperser CD The cross disperser hereafter CD is a mosaic of 2 12 by 16 gratings mosaiced such that the effective length of any ruling is 24 and the total ruled width is 16 The mosai
3. 55 Data Reduction HIRES Manual shift yes Use average shift instead of recentering RESIZING PARAMETERS llimit INDEF Lower aperture limit relative to center ulimit INDEF Upper aperture limit relative to center ylevel 0 1 Fraction of peak or intensity for automatic widt peak yes Is ylevel a fraction of the peak bkg no Subtract background in automatic width r grow 1 1 Grow limits by this factor avglimi yes Average limits over all apertures TRACING PARAMETERS t_nsum 8 Number of dispersion lines to sum t_step 16 Tracing step t_nlost 128 Number of consecutive times profile is lost befo t_funct spline3 Trace fitting function t_order 3 Trace fitting function order t_sampl Trace sample regions t_naver 1 Trace average or median t_niter 10 Trace rejection iterations t_low_r 2 5 Trace lower rejection sigma t_high_ 2 5 Trace upper rejection sigma t_grow 0 Trace rejection growing radius EXTRACTION PARAMETERS backgro none Background to subtract skybox 1 Box car smoothing length for sky weights none Extraction weights none variance pfit fitld Profile fitting type fitld fit2d clean no Detect and replace bad pixels saturat INDEF Saturation level readnoi 4 Read out noise sigma photons gain 2 38 Photon gain photons data number lsigma 4 Lower rejection threshold usigma 4 Upper rejection threshold nsubaps 1 Number of subape
4. LA Ito Gece Pee 8 Lamp Filters Er 10 HIRES Decker 13 FIIBES EIIGIS 2 dale roe EC 26 bas oec 15 Some useful filler combinations 16 viii HIRES Manual Introduction Chapter 1 Introduction HIRES was conceived in early 1987 in response to a call for instrument proposals for first light of the Keck Ten Meter Telescope It went through three round of proposals before being selected as one of the initial complement of Keck first light instruments It took about 5 years to build at a total cost of about 4 0 million It was designed and built in the technical laboratories of the UCO Lick Observatory at the University of California at Santa Cruz The name HIRES stands for High Resolution Echelle Spectrometer HIRES was designed to take advantage of the Keck telescope s large collecting area to push high resolution optical spectroscopy out to about V 20 0 at typical spectral resolutions of 30 000 to 80 000 A discussion of the various key science drivers which weighed heavily in the design of HIRES can be found in the HIRES Phase C proposal Vogt 1988 and will not be discussed here A preliminary overview of the as built instrument was published by Vogt 1992 HIRES is a fairly standard configuration in plane echelle spectrograph with grating cross dispersion It resides permanently at the right nasmyth focus of the Keck telescope HIRES is designed primarily to go quite faint by traditional high re
5. Star rows 13 Sky rows 56 Enter dark count electrons unbinned pixel hour 2 000 Enter readout noise electrons pixel 4 300 Enter binning factor dispersion 1 Enter binning factor cross dispersion 4 Enter magnitude of star 19 00 What type of magnitude Johnson 1 AB 2 1 Lunar phase days 0 Sky brightness at 5500A is 21 9 magnitudes arcseconds 2 Enter airmass 1 300 Enter exposure time seconds 3600 71 HIRES Exposure Estimator HIRES Manual Star counts 284 16 9 Sky counts 19 8 6 Dark counts 26 55 1 Readout 2 42 Net star 284 20 33 Net S N 14 per 41 mA pixel 26 per 142 mA resolution element The entries within the square brackets are the default parameters which will be used if you just hit lt return gt on each line If you enter new values the defaults will be updated to your latest values In the summary table of counts at the bottom the middle column represents signal level from each source and the right column lists the relative contribution of each source to the total noise One normally exits the S N estimator with a cntrl d If you wish to generate a file of the results you create when running the S N estimator before you run the program type script filename where filename is the name of the file where your S N estimator will be saved Then remember to exit from that script after terminating the S N estimator program with a cntrl d 72
6. apscatl Fitting parameters across the dispersion apscat2 Fitting parameters along the dispersion mode ql Note that we ve set all the find recenter resize edit and trace aperture parameters to no since we will be using the apertures found from the quartz spectrum The line 900 was set to try to avoid the dark blob but the routine started anyway at column 1024 so this didn t seem to matter It will start by giving you a cut down column 1024 Identify what you think are the scattered light points and fit them interactively using order low high and niter commands to adjust the order the low and high reject threshholds etc In general you ll want to set the high threshhold pretty low to reject high points which come from edges of orders etc And you ll want to set the low threshold pretty high so as not to reject many low points because most of them will contain useful dark information Again try to stay with as low an order as you can to avoid introducing ripples When satisfied with each column s fit type q to quit and you will be prompted for a new column value You must reply with a col 100 to fit along column 100 etc I suggest fitting at every 100 columns across the CCD You can go across the CCD once fitting every 100 columns and you can return to places later for re fitting if necessary until you ve got it just the way you want There is presently a diffuse halo in the scattered lig
7. camera mirror Instrument Description HIRES Manual An Iodine absorption cell can be moved into position directly in front of the slit for very precise wavelength calibration Quartz halogen incandescent lamps and hollow cathode lamps located up near the ceiling of the slit area provide for flat fielding and wavelength calibration Light from these calibration lamps gets fed into the HIRES optical axis by reflection off a feed mirror which slides into place when calibration is desired The calibration lamp system has a filter wheel for chromatic and intensity control and one position of that filterwheel contains a Fabry Perot etalon for producing Edser Butler fringes along the echelle orders to aid in wavelength calibration Immediately behind the slit are two filter wheels mainly for filters required for blocking unwanted cross disperser orders Behind the filterwheels is a shutter for controlling the start and stop of an exposure The f 13 7 beam then expands and gets collimated to a 12 diameter beam by either of two red blue optimized collimator mirrors The collimated beam is then sent to an echelle grating 1 x 3 mosaic and then to a cross disperser grating 2 x 1 mosaic It makes a 40 turn off the cross disperser and into a large 30 diameter entrance aperture prime focus camera The camera features two large corrector lenses with very special sol gel anti reflection coatings a large light weighted Hextek primary mirr
8. 1 265 X Tilt deg 0 000 HamHeightzssssessm Deck Hgt m 0 001000 Deck Size 1 379 Slit Wid m 0 001000 Slit Size 1 379 SlitWid pix 4 614 SlitVel m s 9619 7 x Ord 1 Xa deg 25 34 X binning 1 Curs psn M 5403 9 EchOrd 66 Ab A 5401 7 Disp px 0 038 Disp A mm 1 57 Sep 12 97 Sep pix 68 0 Loc mm 31 66 Len mm 52 288 Chip Pixel 1 XPixel 193 YPixel 2361 DeckHgt pix 7 224 1 3820 32 077 Section 3 Graphical Interaction Most of the user interaction with the program is accomplished with single keystrokes or mouse clicks while the graphics are displayed Many of the possible commands are visible on the menus at the left side of the screen The scarcity of screen real estate on some types of displays prevents all possible 35 The HIRES Spectral HIRES Manual Format Simulator commands from being visible There are 3 methods by which the user can graphically interact with the program 1 Accelerator Key 2 Mouse Drag 3 Menu Click A glance through the following tables will reveal that some parameters may be modified using more than one of these methods Accelerator keys are single keystrokes An accelerator which is associated with a Boolean parameter will toggle that parameter from one state to the other An accelerator which is associated with a string or numeric parameter will prompt the user for a new value If the display is an X11 se
9. 16 Chapter 7 Chapter 8 Appendix A Appendix B Appendix C Appendix D Appendix E Appendix F Appendix G Appendix H Appendix Attach Dispersion Solution to the Solar Spectrum 67 Applying Dispersion Correction to Object SSC EEN 68 Flattening the Continuum 68 Show and Tell of the Final Result 69 HIRES Exposure Estimator 71 References a as 73 Some useful numberS 74 Spectrograph Technical Data 75 Detector Technical Data 76 Telescope Technical Data 77 Tables of Spectral orders 78 System efficiency 82 Special considerations for low S N and or long integration observations 85 Future HIRES upgrades 86 Acknowledgments 88 vi Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Figure 8 Figure 9 List of Figures HIRES Schematic 5 TV Camera Field acs WS dd eae 9 Blue Blocking Filters 18 Red Blocking Filters 19 Collimator Reflectivities 20 Cross disperser efficiency curves 22 Tektronix CCD Quantum Efficiency 26 Typical appearance of the Simulator during interactive USE ss odo wea a Se gt 35 Spectrometer Telescope Efficiency 84 Table 1 Table 2 Table 3 Table 4 Table 5 List of Tables TV Filters
10. Chapter 3 Section 1 Section 2 Section 3 Topic 1 Topic 2 Topic 3 Section 4 Chapter 4 Chapter 5 Section 1 Section 2 Section 3 1 2 3 Section 4 1 2 3 Section 5 1 O1 W nnm Enclosure Electronics Bay and Clean Room Ante Ohamber usa pam uc Sod ad ok Ae ee 27 Electronics Control System 28 Software Control System 29 The HIRES Spectral Format Simulator 30 Before starting some words about Configuration FSS oie rr kee ch Lol ees IN A 31 Starting the format Simulator 33 Graphical Interaction 35 Modifying the display 36 Modifying the Setup 38 Modifying the Optics 39 Command Line Interaction 40 Preparation for Observing 42 Observing at Keck 43 Arrival at the Mountain top 43 Starting Up the Instrument 43 Selecting a Spectral region 43 Collimator choice o 43 Cross disperser choice 43 Order blocking filters 44 Choosing the Entrance Aperture 44 SIG ie WEE EE 44 slit length ie uos ai AA 44 decker vs slit considerations 45 CCD Readout Mode 45 WINdOWING sa tm alt e E t die oe e de 45 A urhe in Peas alex E 45 fast slow readout modes 45 MPP full well vs dark current modes 45 Mul
11. Image Reduction and Analysis Facility PACKAGE echelle TASK refspectra input solar ds ec List of input spectra referen thar ec List of reference spectra apertur Input aperture selection list refaps Reference aperture selection list ignorea no Ignore input and reference apertures select match Selection method for reference spectra sort Sort key group Group key time no Is sort key a time timewra 17 Time wrap point for time sorting overrid no Override previous assignments confirm yes Confirm reference spectrum assignments assign yes Assign the reference spectra to the input spectr logfile STDOUT logfile List of logfiles verbose no Verbose log output answer yes Accept assignment mode ql This now takes the solution computed for thar ec imh and attaches it to solar ds ec imh At this point though you only know the true wavelength for each pixel in each order You will generally want to linearize or logarithmize or whatever the dispersion to some convenient sampling scale with the task of the next section 67 Data Reduction HIRES Manual Attaching wavelength references is generally far more complex than simply attaching a single solution to a single frame Often one will have pre and post Th Ar spectra and will want to interpolate between these Or one may wish to attach a solution from a group of reference spectra and to perhaps a group of program objects Or o
12. Low rejection in sigma of fit high_re 0 High rejection in sigma of fit niterat 10 Number of rejection iterations grow 1 Rejection growing radius in pixels markrej yes Mark rejected points graphic stdgraph Graphics output device cursor Graphics cursor input ask YES mode ql Note that we ve set high_rej 0 and low_rej 2 in order to try to avoid having the continuum fit being pulled down by absorption lines But these fitting parameters should be played with according to the user s judgement Again though try to stick with the lowest order possible on continuum fits Show and Tell of the Final Result You will now have in solar final ec a respectable solar spectrum There are several ways to explore this final result and make hard copies You can certainly use splot to plot order by order I like to use specplot which can display all orders at once Here is the parameter file Image Reduction and Analysis Facility PACKAGE echelle TASK specplot spectra solar final ec List of spectra to plot apertur Apertures to plot bands 1 Bands of 3D images to plot autolay yes Use automatic layout algorithm autosca yes Scale to common mean for automatic layout fractio 0 Fraction of automatic minimum separation step units wavelength Coordinate units scale 1 Default intensity scale 69 Data Reduction HIRES Manual offset Default intensity offset step Default
13. O UU vs vs uS GA GA C C C C CO NO IO PO PO PO DN PO PO IND SD SD 320 00033000044 4 CO CO P2 ND ND PP IP E29 CO QO QO o Xo WOW c 00 00 00 336 543 758 981 4213 454 704 965 236 9 19 813 120 440 774 122 487 869 268 686 124 583 065 5172 104 665 254 876 533 225 958 733 555 426 351 381 497 689 962 326 788 358 048 837 966 275 79 167 876 866 922 050 251 531 894 343 885 2524 267 e 119 087 178 401 764 276 948 790 816 039 473 135 044 220 NNNNNNNNNNNNNDN ID NH FORMAT IN 2ND ORDER OF THE CROSS DISPERSER PKckckckckckckokck ck KK ck ckckckckck ck kck ck KK KK KA ck ck ck ckck ck ck k ck AAA ECHELLE grooves mm 52 68 Blaze Angle 70 4 Theta 5 0 DIAMETERS Collimated Beam 0 3028 m Telescope 10 90 m Collimator Focal Length 4 1547 m Camera Focal Length 0 7627 m CD GRATING 250 gr mm ORDER 2 Order Blaze A FSR A DEL mm DEL asec HEIGHT mm LENGTH mm DISP A mm 9 2994 8 25 2 0 995 7 916 4 863 28 94 0 87 8 3020 2 25 6 011 8 051 5 866 29 18 0 88 7 3046 0 26 0 028 8 189 6 885 29 43 0 88 6 3072 3 26 5 046 8 330 74 922 29 68 0 89 5 3099 0 26 9 064 8 476 8 9 TT 29 94 0 90 4 3126352 27 4 082 8 625 10 050 30 20 0 91 3 319349 2159 10 8 778 11 141 30 47 0 92 2 3182 0 28 4 120 8 936 124 251 30 74 0 92 L 3210 7 28 9
14. by first light In the future as further funding becomes available and providing science needs dictate I expect to add a number of useful features For instance one can add new cross dispersers to best match the order separation wavelength coverage required of any given project It may prove worthwhile in the future to consider adding other echelles particularly if detector formats evolve consider ably For example an R 1 5 echelle optimized for wide wavelength coverage in the ultraviolet in conjunction with a first order uv cross disperser would be quite useful for QSO work Or a coarser echelle could be used to provide shorter orders in the red and near IR to avoid gaps in the spectrum The addition of each new echelle or cross disperser is relatively expensive but easy to fund from individual research grants if the science warrants The first addition will be probably an image rotator This is quite necessary for highest performance of the instrument when doing long exposures on faint objects both to compensate for field rotation and to eliminate light losses at the slit from atmospheric dispersion The prob lem of atmospheric dispersion of point like objects can be overcome by using the rotator to set the HIRES slit along the parallactic angle such that the dispersed image lies along the slit Field derotation will also be required for long exposures on gravitationally lensed QSO s Ultimately an atmospheric dispersion compen sator must a
15. for this run we will not be extracting sky since there really is none in these data We will do a simple sum along columns in the aperture window for this extraction but of course more complex extractions such as optimally weighted summation can also be done here You should review each extracted order interactively or use splot to again sanity check for correctness Wavelength Calibration We will now use the task ecid to do a wavelength calibration using the thar ec imh spectrum IRAF has a catalog of some 3000 Th Ar lamp emission lines and will use this information in conjunction with the many line positions found in your thar reference spectrum to compute a very accurate wavelength solution for each order You might wish to obtain a copy of A CCD Atlas of Comparison Spectra Thorium Argon Hollow Cathode 3180 A 9540 A Willmarth 1987 from NOAO It is helpful for recognizing and indentifying features in your Th Ar spectrum Unfortunately it doesn t look that similar to the specific lamp used in HIRES so don t expect line strengths or line strength ratios to be the same Only line spacings can be trusted and some lines won t even be there A better aid is the Th Ar atlas we are currently working on taken with the actual HIRES lamp Graduate student Mike Keane at UCO Lick is championing this Hopefully it will soon be available as a UCO Lick Technical Report You may also wish to run a session of the HIRES echelle simulator
16. query mode Il Image Reduction and Analysis Facility ccdred setinstrument hires keck ccddb yes ql Instrument ID Site ID Instrument directory Review instrument parameters Instrument ID type q to quit type for a list CCDRED You are now in the epar mode in the parameter list for the package ccdred This is what you ll see PACKAGI TAS Di N pixelty verbose logfile plotfil backup instrum ssfile graphic cursor version mode Il Il Il Image Reduction and Analysis Facility imred ccdred real real subsets stdgraph ccddb keck hires dat 2 October 1987 87 ql Output and calculation pixel datatypes Print log information to the standard output Text log file Log metacode plot file Backup directory or prefix CCD instrument file Subset translation file Interactive graphics output device Graphics cursor input A crucial parameter here is pixelty which must be real for both output and 51 Data Reduction HIRES Manual calculation types This does eat up more disk space but you run the risk of data overflow if you don t do this Note that the instrum parameter is already set up for you When done with this parameter list exit by typing lt g gt to exit and execute This will move you over into the epar mode in the ccdproc task of the ccdred package CCDPROC You will now be in the epar mode of th
17. 140 9 098 13 381 31 02 0 93 0 3239 9 295 5 160 9 264 14 531 31 30 0 94 09 3269 6 30 0 18 9 435 15 702 31 59 0 95 08 3299 9 30 6 203 9 610 16 894 31 88 0 96 07 3330 7 Ils 2225 9 791 18 107 32 18 0 97 06 3362 1 3164 248 9 976 19 343 32 48 0 98 05 3394 1 32 3 27 0 167 20 602 32 19 0 99 04 3426 8 32 9 1295 0 364 21 885 33 11 00 03 3460 1 33 6 320 0 566 23 192 33 43 00 02 3494 0 34 3 345 0 774 24 524 33 76 01 01 3528 6 34 9 Od 0 989 25 882 34 09 02 00 3563 9 35 6 398 1 210 27 267 34 43 04 99 3599 9 36 4 426 1 437 28 679 34 78 05 98 3636 6 Sl 455 1 672 30 119 35 14 06 97 3674 1 31 9 484 1 914 31 588 35 50 07 96 3 T3 23 3 38 7 514 2 163 33 087 35 87 08 95 3751 4 39 5 546 2 421 34 616 36 25 09 94 3791 3 40 3 578 2 687 36 178 36 63 10 93 3832 1 41 2 611 2 961 33s 713 37 02 s ih 92 3873 8 42 1 646 3 244 39 401 37 43 413 91 3916 3 43 0 681 34537 41 064 37 84 14 90 3959 8 44 0 718 3 839 42 764 38 26 15 89 4004 3 45 0 756 4 152 44 500 38 69 16 88 4049 8 46 0 795 4 476 46 276 39 13 18 87 4096 4 47 1 836 4 810 48 091 39 58 T9 86 4144 0 48 2 878 5 057 49 948 40 04 20 85 4192 8 49 3 921 5 516 51 847 40 51 woe 84 4242 7 5055 966 5 887 53 790 40 99 23 83 4293 8 Dit 2 013 621 3 55 7719 41 49 225 82 4346 2 53 0 2 061 6 672 57 816 41 99 26 81 4399 8 54 3 Sech 7 086 59 902 42 51 28 80 4454 8 5547 2 163 7 516 62 039 43 04 429 79 4511 2 57744 24213 7 962 64 228 43 59 v3 78 4569 0 58 6 2623 8 426 66 473 44 15 133 77
18. 4628 4 60 1 2 331 8 908 68 774 44 72 34 76 4689 3 61 7 2 392 9 409 71 135 45 31 133 6 75 4751 8 63 4 2 454 9 930 73 558 45 91 38 74 4816 0 65 1 2 520 20 472 76 044 46 53 40 4882 4949 5019 5091 5165 5241 5319 5399 5482 5568 5656 5748 5842 5939 6040 6144 6252 6364 6479 6599 J JO GR Oy d O0 dS DO VU tO O0 9 O O I9 UO CO N 0 ad 1 o 4S OO 1 00 O 4 00 4 P 0 JIJI SEER aS YA WWW WWWWNNNNN 588 659 4432 809 974 061 153 249 349 454 564 680 802 929 064 205 354 4512 678 037 626 239 879 547 245 974 EFST 935 SS 247 166 130 143 209 329 509 753 065 451 81 597 220 914 684 933 464 481 587 786 084 484 992 613 353 217 212 344 622 053 646 Appendix F System efficiency A plot of the overall system efficiency is shown in Figure 9 This plot shows the results of several attempts at measuring system efficiency on several different dates with different flux standard stars Since conditions were not always perfectly photometric some variation is expected The ordinate is the combined absolute efficiency of telescope HIRES Tektronix CCD It does not include losses from the slit from an ADC from an image rotator or from atmospheric absorption Solid line curves represent measured efficiencies based on actual measure ments of flux standards through a wide open slit Dotted line curves
19. A solid state laser is also available It produces a very intense beam and is used only for alignment and scattered light experiments The laser actually produces a spectrum of intense lines spanning about one full order of the echelle Should an observer be so foolish as to attempt to observe the light from this laser with the CCD they can expect to suffer the consequences not the least of which may be a very prolonged residual image All calibration lamps are housed in a light tight thermally insulated housing above the slit area The lamps are mounted on a translating table which runs parallel to the slit HIRES Manual Instrument Description Figure 2 TV Camera Field 45 arcsec diam clear f o v 4 4 60 arcsecs gt 45 arcsecs Instrument Description HIRES Manual The calibration lamp optical system was designed to ensure that calibration light enters the spectrograph as similarly as possible as that coming in from the telescope at all wavelengths This is crucial for accurate measurement of instrumental profiles and flat fielding Light from the calibration lamps is first collected and collimated by a cemented doublet HTF1 Fused Silica lens located just above the filter wheel The collimated beam then passes through a defining stop which sets the size of the projected pupil and adds a central obstruction The beam then passes through a 12 position
20. Az The mirror blank weighs only about 183 lbs It is supported axially at 6 points which attach to the mirror s honeycomb structure at the center of 23 Instrument Description HIRES Manual gravity plane Radial support is accomplished through a diaphragm ring structure glued to the center rear surface of the mirror The mirror is enclosed in a dust tight housing and its doors should be kept closed whenever possible for obvious reasons Eventually two identical mirrors will be available one overcoated with enhanced aluminum and the other overcoated with a multi layer silver recipe At present only a single mirror is available overcoated with a standard telescope grade aluminum coating Switching between mirrors or removing the one presently available mirror is done with a manually operated overhead crane The mirror cell assembly is kinematically located on three ball feet and held in place by both gravity and hold down clamps Field Flattener Dewar Window The field flattener is a thick meniscus lens made of Corning 7940 fused silica As with the corrector lenses it is AR overcoated with sol gel so must never be touched at any time with anything for any reason period This lens also functions as the dewar vacuum window It is sealed to the dewar with a single o ring and positioned axially by a precision machined surface Radial locating is accomplished with three radial retainer clips The lens must be properly centered to wi
21. Filters UGS Imm UGT imm BGI2 mmy 10 HIRES Manual Instrument Description Table 2 Continued Lamp Filters TB my The light from the calibration system is fed into the HIRES optical axis by a feed mirror which automatically slides into place in front of the slit when calibration lamps are requested When stowed this mirror retracts into a dust tight housing off to one side The mirror has angular adjustments which allow the calibration system pupil to be aligned with the telescope pupil The adjustments must only be made by a qualified technician Iodine Absorption Cell An iodine absorption cell can also be slid into position directly in front of the slit This cell is basically a sealed glass bottle with a small amount of iodine crystal within When heated to a temperature above 35C the iodine sublimes and the gas then produces an absorption spectrum on the beam from the telescope as it enters the spectrometer The iodine absorption spectrum is a rich forest of deep very narrow lines This forest of lines starts at about 4800A and ends near 6000A The absorption spectrum thus yields a very stable zero velocity reference spectrum superimposed on the spectrum of the object being observed It is intended to be used primarily for very accurate radial velocity studies involving asteroseismology and searches for planetary companions of stars A detailed description of the iodine cell and its use for
22. HIRES Manual References Chapter 8 References Epps H W and Vogt S S 1993 Applied Optics 32 6270 Leach R 1988 Publ Astronomical Society of the Pacific 100 1287 Marcy G W and Butler R P 1992 P A S P 104 270 Massey P 1992 A User s Guide to CCD Reductions with IRAF National Optical Astronomy Observatories Tucson Az Vogt S S 1992 ESO Workshop on High Resolution Spectroscopy with the VLT ESO Garching 11 13 February 1992 p 223 Vogt S S 1988 HIRES Phase C proposal UCO Lick Technical Report No 57 Willmarth D 1987 A CCD Atlas of Comparison Spectra Thorium Argon Hollow Cathode 3180 A 9540 A F National Optical Astronomy Observatories Tucson Az 73 Appendix A Some useful numbers This section not yet completed 74 Appendix B Spectrograph Technical Data This section not yet completed Appendix C Detector Technical Data This section not yet completed Some items for potential inclusion format flat filed response dark current cosmetic defects noise and gain full well capacity cosmic ray rate orientation and useful amplifiers OO sl QN UA eg UA bas 76 Appendix D Telescope Technical Data Sa Qu Us This section net yet completed Some items for potential inclusion Effective light gathering area Image scale at f 15 nasmyth 1 3789684 arcsec mm at nasmyth see p 6 5 of HIRES book V of 4 30 92 notes Typical pointing accuracy Typical guiding accura
23. apsum input solar ds thar List of input images output List of output spectra format echelle Extracted spectra format referen quartz List of aperture reference images profile List of aperture profile images interac yes Run task interactively find no Find apertures recente no Recenter apertures resize no Resize apertures edit no Edit apertures trace no Trace apertures fittrac no Fit the traced points interactively extract yes Extract apertures extras no Extract sky sigma etc review yes Review extractions line INDEF Dispersion line nsum 10 Number of dispersion lines to sum backgro none Background to subtract none average fit weights none Extraction weights none variance pfit fitld Profile fitting type fitld fit2d clean no Detect and replace bad pixels skybox 1 Box car smoothing length for sky saturat INDEF Saturation level readnoi 0 Read out noise sigma photons gain 1 Photon gain photons data number lsigma 4 Lower rejection threshold usigma 4 Upper rejection threshold nsubaps 1 Number of subapertures per aperture mode ql 62 HIRES Manual Data Reduction By not putting explicit names in the list of output spectra we will be accepting the default whereby extracted spectra keep the same name but with addition of a ec exptension Again we will be using the quartz imh image as the reference for apertures Also
24. are predicted efficiencies based on the measured Ist order efficiency and knowledge of the wavelength dependence of the optical system efficiency Since these curves also include the efficiencies of the three telescope mirrors they may be expected to decline as the telescope gets dirty By way of passing interest HIRES by itself peaks at about 13 efficiency But there are three aluminum telescope mirrors out there in the telescope ahead of HIRES in the photon path Each telescope mirror primary secondary and tertiary when cleaned has only about 85 efficiency So 1 0 85 or almost 40 of the light is lost to the telescope and thus never makes it to the HIRES entrance slit HIRES is attached to effectively a 7 7 meter telescope Clearly there are gains to be had with using silver on some or all of the telescope mirrors but at the price of losing the region below about 3400A Keck II will apparently have all silver mirrors It is my hope that we may be able to borrow the silver tertiary for HIRES on Keck I occasionally Also worth noting is that all of the HIRES lenses and mirrors transmit ex tremely well down to the atmospheric cut off The rapid fall off in system effi ciency towards the ultraviolet is due partly to the roll off of the cross disperser s blaze function in Ist order and also to the roll off in Q E of the first light engineering grade Tektronix CCD A uv blazed first order CD is under construc tion to improve this
25. attempt to touch them or even to get near them And their cover should be kept closed when not in use 21 Absolute Efficiency Instrument Description HIRES Manual Figure 6 Cross disperser efficiency curves SCD MR122 1 2 gr mm egree blaze angle 2nd order 1st order 40 deg included angle average of S and P polarization 0 2 03 04 05 06 07 08 09 1 0 wavelength microns 22 HIRES Manual Instrument Description Camera The camera is an all spherical f 1 0 polychromatic catadioptric sys tem It uses two corrector lenses an f 0 76 primary mirror and a thick meniscus field flattener which also serves as the dewar vacuum window This style of camera is extremely achromatic The camera delivers 21 6 micron rms diameter averaged over all field angles and colors images over a 5 2 diameter field of view over a spectral range of 0 3 to 1 1 microns with no refocussing A de tailed description of this camera was presented by Epps and Vogt 1993 Some sacrifice of image quality was necessary with this final design to accomodate at the last minute the unanticipated curved surface of the CCD 65 radius of curvature Backup designs featuring 12 6 micron rms image diameters over a 6 7 field of view with flat focal plane are also in place for the time when the flat CCD s become available Retrofitting to the flat focal plane design requires only fabricating and insta
26. color filters provide some capability for distinguishing and or guiding on different color sources For example a colored filter may be necessary for accurate guiding if the wavelength being sampled by the spectrometer is not the same as that sensed by the CCD TV In particular the Photometrics CCD is not sensitive below about 0 4 microns so ultraviolet spectral observations require special offset guiding especially if significant atmospheric dispersion is present Instrument Description HIRES Manual Table 1 TV Filters o y ND OI AT OA om The aperture and focussing of the TV camera lens is also under computer control The aperture is generally to be left wide open but can be stopped down to increase the dynamic range of the camera Re focussing will be necessary as one switches from guiding off the slit jaws to guiding off the decker plates Some refocus is also necessary if different total thickness filter combinations are used The TV camera system automatically refocusses for the different filter thickness combinations but presently assumes that the focus for guiding the deckers and for guiding off the slit are the same Calibration Lamp System A series of lamps are provided for wavelength calibration and flat fielding A Thorium Argon hollow cathode lamp is provided for the former and a quartz halogen 3400K incandescent source is provided for the latter A Deuterium lamp is also provided for flat fielding in the deep ultraviolet
27. defined are plotted on the Echelle format twice The position of the spectral line closest to the blaze is drawn with a filled dot and the position next closest to the blaze is drawn as an open dot It 33 The HIRES Spectral HIRES Manual Format Simulator will usually be best to choose to observe a spectral line in the order where it is closest to the blaze since that will be where most of the light at that wavelength is located The secondary line locations are shown for cases where the Echelle format is large compared with the detector Information concerning the blaze wavelength and or order number of each echelle order can be toggled onto or off of the display by clicking on appropriate menu items The simulator also draws a schematic of the detector s properly positioned on the Echelle format Any bad spots on the detectors are indicated by rectangular regions on the display If you are windowed down to some subset of the CCD the readout regions are indicated by dotted lines The simulator also displays text lists which give the details about the optics their settings and the current location of the cursor At this point the simulator is ready for interactive graphical use 34 HIRES Manual The HIRES Spectral Format Simulator Figure 8 Typical appearance of the Simulator during interactive use Other Menus Optical HW All Setup DetectorMosaic Refresh All ZoomOnDetector o e SetupName twoamp Observer E Tilt deg
28. required to get proper focus Focus errors of 0 001 may degrade spectral resolution unacceptably A powerful focussing algorithm is being developed which should assist the user in this task At some point it will probably be possible to fully automate the focus procedure The camera frame is a thermally stabilized design and camera focus should thus not be a function of temperature During the first 4 months of HIRES use we have not seen any reason to refocus the camera but it should be periodically checked Focussing can also be done using the collimator focus and refocussing for various thickness filters etc will generally be done using the collimator focus Changes of AX in collimator focus are equivalent to changes of AX 30 in camera focus Enclosure Electronics Bay and Clean Room Ante Chamber The spectro graph is enclosed in a modular insulated light tight dust tight housing This housing provides about a 9 hour thermal time constant between inside and outside temperatures There is no attempt to thermally control the interior temperature Rather it is expected to track the dome interior temperature from day to day but not from hour to hour A slow flow of filtered and dried dome air is continually forced into the enclosure through a hose and adjustable valve that is the hissing noise you always hear when inside All attempt has been made to keep all sources of heat out of the interior of HIRES and not to dump heat into th
29. separation step ptype 1 Plotting type labels user Type of labels ulabels User labels file xlpos 1 02 X label position fraction of range ylpos 0 Y label position fraction of mean sysid yes Include system banner and step value yscale yes Draw Y axis scale title u vogt IRAF demo solar final ec imh Plot title xlabel wavelength X axis label ylabel intensity Y axis label xmin INDEF X axis left limit xmax INDEF X axis right limit ymin 0 Y axis bottom limit ymax INDEF Y axis top limit logfile Logfile graphic stdgraph Graphics output device cursor Cursor input mode al If you set fractio 0 the continuum level for all orders will remian at a constant ordinate value all the way across the spectrum plot The spectrum will be very highly compressed in the wavelength direction but can be expanded about any point with the lt wx gt command as many times as you need and then wl and wr to pan left and right Unfortunately there doesn t seem to be an unexpand key stroke so you have to wa to redraw the entire plot if you ve overexpanded or get tired panning left right in too small increments It is instructive to look carefully at the order overlap regions to see how well these overlap regions agree These are independently observed and reduced spectral regions and offer a useful consistency check Some disagreement is expected in the co
30. used for spectra from a multi fiber input feed but again such a fiber feed will not be provided at first light One other important point is that the HIRES slit is fixed with respect to the telescope and thus its position angle rotates on the sky as the telescope tracks Thus it may not be possible to get a particular desired position angle on an object unless the object is observed at the proper position in the sky And the slit position angle will rotate as the telescope tracks causing potential problems on extended objects with long exposure times Finally at high zenith distance atmospheric dispersion will spread the image out into a tiny spectrum and in general this spectrum will not be aligned along the entrance slit causing potentially large light loss over certain regions of the spectrum if not carefully accounted for in guiding etc An image de rotator is under consideration but will not be provided at first light An atmospheric dispersion compensator would also be helpful but is also not to be provided at first light So the prospective user is advised to carefully consider the consequences of a rotating slit and lack of atmospheric dispersion compensation on the proposed observing program HIRES Manual Instrument Description Chapter 2 Instrument Description The HIRES instrument sits permanently on the right nasmyth platform of the Keck telescope It is enclosed in a thermally insulated light tight dust tight room which is
31. which contains your images This directory will also end up containing as subdirectories all the databases etc generated by IRAF when it munches on your data I like to break my directory up as IRAF in the top level and then subdirectories containing groups of data files which are to be combined together in a reduction For this example I will assume we have obtained a group of exposures of the day sky solar spectrum plus calibration spectra at some place in the echelle format and that these data files have been put into the directory u vogt IRAF demo The data frames to be used in this sample reduction are solar fits a 100 second observation of the solar spectrum quartz fits the spectrum of a quartz halogen incandescent lamp dark fits an 1800 second observation of the dark level in the spectrograph zero fits a zero length lt 1s exposure on dark to determine bias levels AS E thar fits a 1 second observation of the Thorium Argon hollow cathode lamp used for wavelength calibration Note that for full blown data reductions one might well have several zero frames several quartz frames and several dark frames These would then be combined into more noise free calibration frames by median filtering out cosmic rays by simple averaging to reduce readout noise and by suitable interpolation if necessary between bracketing calibration exposures There are a number of strategies which can be used within IRAF to combi
32. 387 44 938 334 776 OL 01 3528 6 34 9 0 704 5 494 44 240 34 09 02 00 3563 9 35 6 0 718 5 605 43 529 34 43 04 99 3599 9 36 4 0 133 bi 4 19 42 804 34 78 05 98 3636 6 3T 1 0 747 5 836 42 064 35 14 06 97 3674 1 37 9 0 763 5 957 41 309 35450 07 96 37112 3 38571 0 778 6 082 40 539 35 87 08 95 3751 4 39 5 0 795 6 210 39 753 36425 09 94 3791 63 40 3 0 811 6 343 38 950 36 63 10 93 3832 1 41 2 0 828 6 480 38 130 37 02 11 92 3873 8 42 1 0 846 6 622 31 2 93 37 43 13 91 3916 3 43 0 0 865 6 768 36 438 37 84 14 90 3 95 98 44 0 0 884 6 920 35 564 38 26 215 89 4004 3 45 0 0 903 7 076 34 670 38 69 i6 88 4049 8 46 0 0 924 1 238 334 757 39 13 18 87 4096 4 47 1 0 945 7 405 32 823 39 58 19 86 4144 0 48 2 0 966 7 578 31 868 40 04 220 85 4192 8 49 3 0 989 7 758 30 891 40 51 122 84 4242 7 50 5 14012 7 944 29 891 40 99 23 83 4293 8 5157 1 036 8 136 28 867 41 49 ect 78 4346 4399 4454 4511 4569 4628 4689 4751 4816 4882 4949 5019 5091 5165 5241 5319 5399 5482 5568 5656 5748 5842 5939 6040 6144 6252 6364 6479 6599 6724 6853 6987 7127 RER 7424 15824 7747 2919 8099 8288 8485 8692 8909 9138 9378 9632 9899 OY O Oy P2 OY GA OO NUNN NN t0 OY 0 40 Oy iS O0 4S N 10 O1 o OO lOO O I9 01 00 O o 00 0 BON 00 00 MN CQ W 0 CO O QO JJ OG GA AJ OY O 00 x0 2 00 Oy OO 1 00 Ob OO 4S PO 40 40 P 4 J 0 P G6
33. CO Lick The underlying mathemat ics are described in the textbook by the original author of the code D J Schroeder Astronomical Optics Academic Press 1987 These algorithms provide a com plete description of the Echelle format within the constraints of the 2 dimensional grating equations They do not handle the more general problem of modelling the 3 dimensional grating equations Other important algorithms used in the code are based upon the text by E Hecht amp A Zajac Optics Addison Wesley 1974 The code is able to model a spectrograph where an Echelle grating is fed by a collimated beam The beam leaving the Echelle grating may be cross dispersed by at most 1 cross dispersing grating and or up to 9 cross dispersing prisms This simulator also serves as a convenient means by which set up files for most all of the HIRES parameters can be created off line i e before going to the mountain while preparing for a run etc These set up files can then be uploaded to Mauna Kea and stored in the instrument computer prior to starting the observing run A detailed description of the HIRES echelle format simulator can be found in Steve Allen s user s manual which is published as UCO Lick Technical Report No 68 But if you don t happen to have a copy of that manual in hand TI simply reproduce with my own comments added much of the user interaction section from Steve s manual here 30 HIRES Manual The HIRES Spectral Format Sim
34. GE echelle TASK ecidentify images thar ec Images containing features to be identified databas database Database in which to record feature data coordli linelists thorium dat User coordinate list match 1 Coordinate list matching limit in user units maxfeat 20 Maximum number of features for automatic identif zwidth 10 Zoom graph width in user units ftype emission Feature type fwidth 4 Feature width in pixels cradius 5 Centering radius in pixels thresho 10 Feature threshold for centering minsep 2 Minimum pixel separation functio chebyshev Coordinate function xorder 4 Order of coordinate function along dispersion yorder 4 Order of coordinate function across dispersion niterat 5 Rejection iterations lowreje 3 Lower rejection sigma highrej 3 Upper rejection sigma autowri no Automatically write to database graphic stdgraph Graphics output device cursor Graphics cursor input mode ql Another much easier way to wavelength calibrate is to use the solution from a previous calibration run and use ecreidentify to match to the previous solution making slight shifts etc This works quite well if the reference spectrum is near to the one you are working on I have not yet tried to see how far one can reach out to reference spectra which are significantly offset Eventually we hope to build a library of Th Ar reference spectra which can be used at any place around the HIRES
35. HIRES USER S MANUAL UCO Lick Observatory Technical Report No 67 UCO LICK OBSERVATORY Steven S Vogt Santa Cruz California May 4 1994 file u hires PUBS manual2 manuall PREFACE This manual is intended as an introductory guide for users of the Keck Ob servatory HIRES spectrometer I personally detest having to read users manuals And I dislike writing such manuals even more The best manual is indeed simply a menu driven self prompting set of options which can guide the user through set up and operation However as much of this user friendly interface will probably not be ready at or soon after first light I shall attempt herein to give the reader enough explanation and description to operate the instrument This is only a preliminary draft of the final HIRES user s manual Some sections have not yet been written and it will probably be found lacking key information here and there I will depend on feedback from first light users to help improve and enhance this user s guide as necessary Please send all your comments via e mail to vogt lick ucsc edu and I will try to incorporate suggested changes as time and energy permits Since this is an evolving document you may wish to check the date on the cover page of your copy If it is not the most recent please discard your copy and get a more up to date one Since this manual is intended primarily for users it will not include much technical detail Detailed technica
36. RAF IRAF is the reduction environment recommended by the P I for HIRES An optimized version of IRAF utilities which have been set up to know about many of the actual instrument parameters lives on the CARA network and will be maintained by the project s software office First time users of IRAF and or echelle spectrometers would be very well advised to obtain a copy of A User s Guide to CCD Reductions with IRAF Massey 1992 This is an excellent guide to basic IRAF data reduction of echelle CCD images Read it carefully before starting any reductions In fact 47 Data Reduction HIRES Manual a careful read through before observing will give you a much better feeling for what kind of calibration frames to obtain with your program object data and how they will be used I will now try to walk you through a very basic set of reductions on a typical data set from HIRES This will be only a simple example but will illustrate many of the most important aspects of the data reduction I will assume you are generally familiar with running IRAF in using its eparameters feature in finding your way around within its libraries of routines and understanding what kinds of data it creates and where such data lives Getting set up with data in the appropriate directory I will also assume that you have logged onto some Sun Sparc station running X11 windows with the latest version of IRAF installed and that you have moved over into some directory
37. This plot is simply meant to be a rough guide as to which order to choose of the cross disperser and what the approximate throughput will be Efficiencies are for the center of the echelle free spectral range at any order For more accurate efficiency estimates one must include also the slit losses for given seeing and atmospheric absorption as well as the effects of sky background dark current readout noise and binning on the final signal to noise of the data This is easily 82 done using the HIRES S N estimator program described in a previous section Our best estimate of actual system efficiency has been incorporated into this simulator Feedback from observers though on their measured efficiencies are always welcome not only to aid in converging on the true efficiency numbers but also to check for and guard against system efficiency decline with time 83 System Efficiency 96 Figure 9 Spectrometer Telescope Efficiency 7 25 93 HR 5501 10 8 93 HR 9087 M 4st order CD 7 18 93 HR 8634 order 02 03 04 05 06 07 08 09 1 0 Wavelength microns 84 Appendix G Special considerations for low S N and or long integration observations This section not yet completed 85 Appendix H Future HIRES upgrades The instrument described thus far is simply the core version of the final in strument There were not enough funds available to build the entire instrument
38. al converter process the CCD video signal All of the electronics are housed in a separate thermally insulated enclosure adjacent to the spectrometer This electronics enclosure is cooled via the ob servatory s recirculating coolant system Since we were obliged to use standard observatory VME electronics modules which are not rated for use below 0 C we will actually be holding the electronics enclosure at a temperature of 5 C 28 HIRES Manual Instrument Description Software Control System The software for instrument control at Keck Obser vatory is written in the C programming language and runs under UNIX on a network of Sun computers The observer controls a given instrument through a software user interface which allows both command line input through keywords and scripts and window style graphical input using X11 windows with the MO TIF toolkit Both types of input can be intermixed The user interface also allows for multiple invocation of control processes which is important for distributed observing Here the primary observer can be quite remote from the telescope i e in Waimea or back in California while graduate students and or technical observers at other sites or at the mountaintop can cooperate in the set up and running of the instrument during an observing run The primary tool for interacting with HIRES is a graphical user interface called xhires It is a self explanatory click on icon pop up menu style co
39. and yorder to decide upon the best compromise Try always to use the lowest order number possible consistent with the desired or expected fitting accuracy In my limited experience I have found 66 HIRES Manual Data Reduction that 4 works well for both xorder and yorder and rms fits of 0 0022 Angstroms are commonly achieved You can also display your fitting residuals in many other meaningful and entertaining ways by using the lt x gt and lt y gt keys to redefine the abcissa and ordinates of the residual plot Try lt yo gt and then lt xp gt to show a map of where all your identified features were located Check for areas clusters where many points may have been deleted If necessary you may have to go back in and pin down more lines in these areas Try lt yv gt and lt xw gt to show velocity residuals with wavelength etc When convinced that the solution looks solid and correct quit out and save what you ve done to the database Next time through near this position on the format you can use this spectrum as a reference and use ecreidentify to do the wavelength solution quickly and painlessly You are now done computing the wavelength solution for the Th Ar reference spectrum Attach Dispersion Solution to the Solar Spectrum We must now use the task refspec to attach the dispersion solution just found for the Th Ar reference spectrum to our program object spectrum solar ds ec imh Here is the parameter file for this
40. at A list of all the setups found in these directories is presented and the user is asked to choose which setup Only two options are presently of relevence to HIRES users the others are for other developmental experiments Pick either 2 for the first order CD format or 3 for 2nd order CD format Once selected using the information contained in the setup file the simulator draws a picture of the Echelle format At this point it would be wise to position this relatively large graphic window such that it does not completely obscure the prompt line of your present window You may sometimes be asked to enter data from this window and unless it s prompt line is visible you may forget that this window exists In the large window showing the spectral format for each Echelle order within the specified wavelength limits one free spectral range FSR centered on the Echelle blaze is drawn Most of the light in any Echelle order is within one FSR of the Echelle blaze wavelength for that order There is some light in each order more than one FSR away from the blaze but the intensity drops rapidly On a display which supports color the simulator extends the length of each Echelle order by drawing another FSR in grey on either side of the blaze In the case of an Echelle spectrograph with grating cross dispersers the simulator displays the selected order of cross dispersion and several nearby orders of cross dispersion Any spectral lines which were
41. c concept is quite similar to that of the echelle The cross disperser ruling is 250 grooves mm The collimator to camera angle is 40 The intended blaze angle of this CD was supposed to have been 5 343 but came out 4 3 The effect of this error was to put the first order blaze peak near 0 56 microns rather than the intended 0 7 microns and the 2nd order blaze peak at 0 28 microns rather than the intended 0 35 microns This CD is intended to be used in 1st order in the visible and in 2nd order in the ultraviolet blue Note that the orders get uncomfortably close together down in the uv with the CD in 1st order but the spacing doubles in 2nd order Appropriate order blocking filters will have to be used to eliminate unwanted CD orders In 2nd order one will generally be limited to a wavelength span per observation of 3 2 times the bluest wavelength observed because of the need to block 3rd order In Ist order a wavelength span of twice the bluest wavelength will be possible if there is enough CCD real estate A plot of the efficiency of the CD is presented in Figure 6 Because of the above mentioned blaze angle error the efficiency of this CD below 0 35 microns will be somewhat lower than hoped A new Ist order CD blazed for 0 39 microns and with 395 grooves mm is being manufactured to correct this situation Like the echelles the CD is housed in a dust tight enclosure Since these gratings can never be cleaned one should never
42. ce then try playing around with parameters And you can always do an end run around the auto order finding routines and just mark them by eye interactively if you wish It is very important at this stage to verify that your apertures are reasonably well fit to the orders so I suggest looking at them in detail in the interactive mode of apall Be sure that apertures from adjacent orders do not ever overlap and that all orders except perhaps for the first and or last which are often partial since they fall off the chip edges are present and accounted for When finished you will be asked if you want to write the apertures to the database You should answer yes and IRAF will create a subdirectory called 57 Data Reduction HIRES Manual database where it saves all the aperture information along with lots of other information to come Now you can extract and review each quartz order to check that they all look reasonable If you wish some hard copies just type snap and you ll get a laser print of the Tektronix plot window If the orders all look good you have successfully located and traced all the echelle orders and are done with apall Generate the Flat Field Image The next task is to generate a flat field image from the quartz which can remove pixel to pixel predominantly high spatial frequency variations We do this using the task apnormalize This task will take the quartz image remove the low spatial frequency va
43. cker plane is about 3 16 above the slit plane The instrument control system automatically refocusses the collimator for the particular slit decker filter thickness collimator mirror combination used Table 3 shows the complete selection of available deckers 12 HIRES Manual Instrument Description Table 3 HIRES Deckers Projected Projected Heihi paai height eer Comments arcsec arcsec use slit EE use slit E m moo le m moo ES COC NI EUM I m m m moo le fa m 0 75 1 36 0 574 146 5 E em 000 28 for 13 Ea A UE 075 10 136 pS PS o 20 140 80 A Instrument Description HIRES Manual Table 3 Continued HIRES Deckers Projected Projected height width Comments pixels pixels Height Width arcsec arcsec R 45 000 3 5 for R 45 000 7 0 for 0 861 Ge Ie 000 14 for um Dom id i 1 148 146 5 XE 28 for sky reo ea e D4 14 0 1 722 73 2 R 23 000 14 for sky DS 0 119 0 179 0 623 0 624 projects to 15x15 microns for tests Slit The slit is a bi parting mechanism which means that the slit centroid should not change position as slit width is varied The slit jaws cannot be closed 14 HIRES Manual Instrument Description completely since this would damage their sharp edges Slit width can be specified either in microns in seconds of arc as projected on the sky or in pixels as projected on the CCD A wide opened
44. cy Zenith blind spot limits 1 1 Altitude limits 33 3 nasdeck region 5 3 to 146 2 azimuth 15 elsewhere dome shutter starts vignetting Field rotation at nasmyth yes indeedy 77 Appendix E Tables of Spectral orders FORMAT IN FIRST ORDER OF THE CROSS DISPERSER KK KK KKK KK KK KKK KKK KKK KKK KKK ck ckck ck kok KKK ck ck ck KKK ECHELLE grooves mm 52 68 Blaze Angle 70 4 Theta 5 0 DIAMETERS Collimated Beam 0 3028 m Telescope 10 90 m Collimator Focal Length 4 1547 m Camera Focal Length 0 7627 m CD GRATING 250 gr mm ORDER 1 Order Blaze A FSR A DEL mm DEL asec HEIGHT mm LENGTH mm DISP A mm 9 2994 8 25 2 0 510 3 958 55 026 28 94 0 87 8 3020 2 25 6 0 518 4 025 54 512 29 18 0 88 7 3046 0 26 0 0 527 4 094 53 989 29 43 0 88 6 3072 3 26 5 0 536 4 165 53 458 29 68 0 89 5 3099 0 26 9 0 545 4 238 52 917 29 94 0 90 4 3126 2 27 4 0 555 4 313 52 367 30 20 0 91 3 315349 21 9 0 565 4 389 51 807 30 47 0 92 2 3182 0 28 4 0 575 4 468 51 237 30 74 0 92 1 3210 7 28 9 0 585 4 549 50 658 31 02 0 93 0 3239 9 2945 0 595 4 632 50 068 31 30 0 94 09 3269 6 30 0 0 606 4 717 49 467 31 59 0 95 08 3299 9 30 6 0 617 4 805 48 855 31 88 0 96 07 3330 7 31 1 0 629 4 895 48 233 32 18 0 97 06 3362 1 Ia 0 640 4 988 47 598 32 48 0 98 05 3394 1 3253 0 652 5 084 46 952 32 19 0 99 04 3426 8 32 9 0 665 5 182 46 293 36 LI 00 03 3460 1 33 6 0 678 5 283 45 622 33 43 00 02 3494 0 34 3 0 691 5
45. e However it is generally prudent to make the slit 44 HIRES Manual Observing at Keck length as long as possible to measure as much sky as possible but short enough such that one still leaves rooom for some rows of dark between orders decker vs slit considerations The guide star image looks much worse when guiding off the slit decker A combination than it does guiding off a simple decker Thus for faint object work use of the deckers alone will be preferable For work requiring a longer slit or narrower slit than 0 6 arcsecs requires use of the normal slit jaws In any case once either the slit or the decker is selected the collimator will automatically be refocussed properly Basically the system looks to see if decker plate A is being used If so it assumes you are to be using the slit jaws If not it wil refocus for the decker plates There is about a 1 8 difference in the focal planes of these two entrance apertures Section 5 CCD Readout Mode windowing binning fast slow readout modes MPP full well vs dark current modes Multiple vs single amplifier modes Section 6 Focussing Section 7 Taking Calibration Exposures Flat fields Wavelength calibration and instrumental profile Th Ar hollow cathode lamps 45 Observing at Keck HIRES Manual E B bands Iodine Absorption Cell Spectrum Dark Frame s Bias zero Frame s Day Twilight Sky and or Moon Spectra Section 8 End of the Evening Final cal
46. e and the camera aberrations blur circle Effectively this 0 6 arcsec slit translates to a spectral resolving power of about 54 000 60 000 A 0 9 arcsec slit width projects to about 3 pixels width and results in a resolving power of about 45 000 A 1 1 arcsec slit projects to 4 pixels width and yields a resolving power of about 34 000 In the limit of larger slits where camera aberrations and finite pixel widths are small the effective throughput slit width times resolving power product is about 39 000 arcsecs The slit must be used with one of the notches in Decker Plate A to define slit length and keep orders from overlapping If instead the user wishes to use one of the apertures from Decker plates B D which define both slit width and length the slit must be opened fully to keep it from blocking any light This is done at present by doing the command m slitwid 11 1 but will eventually be an option in xhires slit length Slit length defined either by one of the notches in Decker plate A or by one of the apertures on the other decker plates must be chosen according to the available order separation at the spectral region of interest and the need for sky subtraction Consult the echelle format simulator for the minimum order separation available in your chosen spectral region and set the slit length to be less than this If sky is not important to measure a slit length only a bit bigger than the seeing disk is adequat
47. e to use Consult the curves in Figure 5 for this choice Cross disperser choice The next decision concerns both efficiency for the spectral region and desired order separation At present we have only one CD but this can be used in either 1st or 2nd order Most applications redward of 0 4 microns will use the Ist order while most blueward of here will use 2nd order Consult both the efficiency curves of Figure 6 and the HIRES spectral format simulator for this decision 43 Observing at Keck HIRES Manual Order blocking filters Once the CD order decision is made you will have to think carefully about how to block unwanted orders from the CD Here Table Sand Figures 3 and 4 will help in this decision Once these filters are selected the collimator will be automatically refocussed for the new filter thickness combination Section 4 Choosing the Entrance Aperture slit width The user must set the slit width to give the desired spectral resolution Basically the de projection factor from actual slit width to projected slit width in the echelle dispersion plane at the CCD is about 1 8 715 which results in a scale of about 12 44 arcsecs mm at the CCD in the echelle dispersion direction The present CCD pixel size is 24 microns or 0 024 mm Thus a 2 pixel projected slit is about 0 60 arcescs wide as projected on the sky The resultant spectral resolution will be some thing like a gaussian quadrature sum of the projected slit the pixel siz
48. e dome Most of the control electronics are contained in a separate Electronics Bay a similar style thermally insulated enclosure and heat inside the electronics bay is carried away by the observatory s recirculating liquid coolant system Electronics inside the HIRES enclosure which drive the CCD and other electronics which control the TV camera are contained in their own thermally insulated footlockers These footlockers are also cooled by the observatory s coolant system A psuedo clean room ante chamber is also provided Personnel entering HIRES will be required to don appropriate clean room garb in this ante chamber Dust accumulation is a serious concern for HIRES and personnel entering HIRES are expected to do all they can to eliminate dust and dirt particularly that brought 27 Instrument Description HIRES Manual in on shoes Sticky mats get a lot of it and must be renewed frequently But even the mats do not get it all Clean room suits with booties are thus mandatory Electronics Control System The control system for HIRES is a VME based system which uses only Keck Observatory standard modules HIRES is one of an initial complement of five first light instruments which connect to a scientific instrument LAN at the mountaintop Each instrument is controlled by its own VME bus based Sun 3E 68020 CPU real time controller running VxWorks which connects over the scientific Ethernet LAN to either of two SUN Sparc series i
49. e task ccdproc PACKAGI TASK Fl images ccdtype max_cac noproc fixpix oversca trim zerocor darkcor flatcor illumco fringec readcor readaxi fixfile biassec Zero fringe scantyp i I 3 O p D i order naverag Image Reduction and Analysis Facility ccdred ccdproc solar quartz thar dark zero object 32 no 2100 2112 23220107 List of CCD images to correct CCD image type to correct Maximum image caching memory List processing steps only in Mbytes Fix bad CCD lines and columns Apply overscan strip correction Trim the image Apply zero level correction Apply dark count correction Apply flat field correction Apply illumination correction Apply fringe correction Convert zero level image to readout correction Convert flat field image to scan correction Read out axis column line File describing the bad lines and columns Overscan strip image section Trim data section Zero level calibration image Dark count calibration image Flat field images Illumination correction images Fringe correction images Minimum flat field value Scan type shortscan longscan Number of short scan lines Fit overscan interactively Fitting function Number of polynomial terms or spline pieces Sample points to fit Number of sample p
50. ed during intial design studies of new Echelle spectrographs It is possible to modify the properties of many of the optical elements while the program is running Under normal circumstances these capabilities are not desired by the user and they are disabled COLFOCLN ColFocL m 39 The HIRES Spectral HIRES Manual Format Simulator ECTHETA EC0 dcg ECTHETA ECO rad XDSIGMA XD groov mm aes ECDELTAD ECblaze deg EST ECDELTA ECblaze rad Section 4 Command Line Interaction When the user has quit from the graphical interaction by simply typing q while in the graphic window the program enters another mode where the interaction is done on the text screen All commands in this mode must be followed by a carriage return lt CR gt 40 HIRES Manual The HIRES Spectral Format Simulator ret graphics and rem o graphical eri Oi tie program 5 ite KICS setup feo dik and optionally nere i 41 Preparation for Observing HIRES Manual Chapter 4 Preparation for Observing This section is not yet written Some possible items for inclusion are checklist for caveats on program object observability etc finder charts coordinates offset stars blind offsets and faint object acquisition and guiding choice of targets choosing wavelength ranges resolutions sky sampling estimating exposure times co
51. filter wheel to a cemented triplet Fused Silica NaCl Fused Silica lens and then off a retractable folding flat which directs the beam into the spectrometer The triplet lens produces a beam of proper numerical aperture f 13 7 focussed at the slit plane and a virtual pupil of the correct size and distance 58 diameter 785 ahead of the slit to accurately mimic the telescope s exit pupil The HTF1 element a glass very similar in dispersive properties and transmission to CaF but without hygroscopic problems and NaCl element were required in order to control pupil distortion and pupil walk over the very wide chromatic range 0 3 to 2 microns of the spectrometer The NaCl element was encapsulated between the fused silica elements to avoid hygroscopic problems The optical system provides a 2 1 magnification so the typically 3 5 mm diameter spot of light produced by hollow cathode lamps is only 6 10 mm or 8 14 arcsec at the slit far too small for longslit wavelength calibration For longslit work the lamp is simply scanned along the slit direction A list of filters available in the comparison lamp system is given in Table 2 Position 1 vignets the beam and should never be used Position 11 contains the Fabry Perot etalon used for wavelength calibration Position 12 is presently open and available for public use though loading one s favorite filter can only be done by a qualified technician and does take some effort and time Table 2 Lamp
52. find for the Sun to see that obvious fetures like H a Na D etc ended up with accurate wavelengths In splot use the lt gt and lt gt keys to move among the orders and lt wx gt lt wl gt and lt wr gt to expand in x and move left and right around any feature You ll see that you have a pretty decent looking spectrum at this point but that there is still the strong echelle blaze profile dominating the continuum shape We will remove that in the next section Flattening the Continuum To flatten the continuum we will use the task 68 HIRES Manual Data Reduction continuum We will use solar ds ec imh as the input and save the continuum flattened version as solar final ec imh Here is the parameter file IRAF Image Reduction and Analysis Facility PACKAGE echelle TASK continuum input solar ds ec Input images output solar final ec Output images lines Image lines to be fit type ratio Type of output replace no Replace rejected points by fit wavesca yes Scale the X axis with wavelength logscal no Take the log base 10 of both axes overrid no Override previously fit lines listonl no List fit but don t modify any images logfile logfile List of log files interac yes Set fitting parameters interactively sample Sample points to use in fit naverag 1 Number of points in sample averaging functio spline3 Fitting function order 1 Order of fitting function low_rej 2
53. format and thus always allow wavelength calibration to be done by ecreidentify If you had a reference spectrum called thar ref ec this is what your parameter list for ecreidentify would look like Image Reduction and Analysis Facility PACKAGE echelle TASK ecreidentify images thar ec Spectra to be reidentified referenc thar ref Reference spectrum shift Shift to add to reference features cradius Centering radius thresho Feature threshold for centering refit SE Refit coordinate function databas database Database logfile STDOUT logfile List of log files 64 HIRES Manual Data Reduction mode al Anyway here we go now in ecid The first thing which will be displayed is a plot of the 1st aperture Move the cursor to a line you think you recognize and type lt m gt to mark If 1t beeps without marking try shifting the cursor a tiny bit to the right often you have to position the cursor slightly right of line center to get it to mark that line It will then respond with the column number and await an input wavelength If you change your mind and do not want to mark that line just hit return and then without moving the cursor hit lt d gt for delete then lt r gt for redraw the plot Using the lt m gt and lt d gt keys you can mark or delete and enter wavelengths for as many features as you wish A lt gt will give you the position of the feature nearest the cursor When done marking li
54. hing else In this case you would want to go in and edit the appropriate keywords to keep IRAF from getting confused over which files are which You can use ccdlist to get a brief review of your files or imhead with long yes to check out what s in their headers followed by hedit if necessary to actually make any changes For example to add the keyword IMAGETYP and set it equal to object in the header for file solar imh you would do hedit solar imagetyp object ver or else do it from the epar route on hedit 50 Setting the Instrument Parameters HIRES Manual Data Reduction Once you are satisfied that you have all the necessary files read in and converted to IRAF imh files with appropriate names and keywords you are ready to begin the actual data processing The first step is to inform IRAF what instrument set up configuration you are using and get it loaded Type setinstrument to load the setinstrument package for HIRES You will be asked the question 1 Assuming it is the HIRES instrument file you will be using hitting hires Instrument ID type for a list a simple lt return gt will load the default option hires and then move you onward to the ccdred task in the imred package You will not see the parameter list for setinstrument but it can of course be accessed by epar ing on setinstrument It looks like this PACKAGE TASK Il instrume site directo review
55. histogram equalize the image you may also be able to see the meteor which is a scattered light feature of the spectrograph It generally runs diagonally across the chip and is brightest when crossing each order You will also see the prominent dark blob near CCD center Again if you need hardcopy output use the print button under the etc menu in SAOIMAGE Flat Fielding the object images With the normalized flat field image in hand from the previous step you are now ready to use the task flatten located in the generic package to remove the high spatial frequency pixel to pixel response variations We will flat field correct both solar imh and thar imh using the flat field image flat imh The flattening routine resides in a package called generic so load this package by typing generic and then epar into flatten The parameter list for flatten should look like this IRAF Image Reduction and Analysis Facility PACKAGE generic TASK flatten images solar thar Images to be flattened flatfiel flat Flat field minflat INDEF Minimum flat field value 59 Data Reduction HIRES Manual pixtype real Flattened image pixel datatype keeplog _ keeplog Keep log of processing logfile _ logfile Log file imlist tmp ims7334a tmp ims7334a flat flat ql m ini O E B a o wd dw dw Ho Wo og Note that flatten will write the flat fielded images back over the originals so if you want to go back to unflattened sola
56. ht like the ring nebula at very low light level centered roughly on the center of the CCD You may notice the double peaked signature of this halo as your cuts move across the chip It may require even 5 7 orders to fit well But out near the edges away from this 61 Data Reduction HIRES Manual halo a lower order spline should suffice Hopefully anti reflection coating the field flattener lens will reduce this halo Once you ve got a full set of columns fit quit out of the routine It will then go away for a long time many mnutes calculating the proper smoothed fit to the background in the row direction When it returns it will display a row cut across the image at row 1024 and you must repeat the above process for a series of rows across the image Again it is good to do a fit every 100 rows across the CCD keeping the order as low as possible When finished quit out of the routine You should then use display and or implot to examine your de scattered light result to see that all looks as expected Extracting Orders and Compressing to 1 d Spectra You are now ready to use apsum to extract the orders and compress them into 1 d spectra We will extract both solar ds imh and thar imh for this demo I didn t bother correcting the thar imh for scattered light since I will only use it for wavelength calibration anyway Here is the parameter file for this Image Reduction and Analysis Facility PACKAGE echelle TASK
57. ibration frames Shutdown Dewar auto fill Section 9 End of the Observing Run Section 10 Observing Checklist Section 11 Observing Log Sheets 46 HIRES Manual Data Reduction Chapter 6 Data Reduction Section 1 FIGARO FIGARO is the CARA standard for quick look data reduction at Keck Obser vatory The HIRES data frames are normally written to disk in FITS format and FIGARO can read these FITS format files The FITS format is the default standard for HIRES However there is a switch selectable option which allows the user to write the data in FIGARO style format and thus bypass the FITS to FIGARO conversion if so desired At this time due partly to lack of a well supported and fully developed FIGARO system and partly to lack of infinite software manpower resources a FIGARO HIRES reduction suite of routines is not available from the HIRES development team No doubt existing FIGARO echelle packages will work with enough patience and tuning Section 2 IDL KHOROS and Others Since the HIRES data is in FITS format many other data reduction analysis packages can be used Gibor Basri and co workers at U C Berkeley prefer IDL and have many useful scripts and routines written to do echelle data reduction Others like the Khoros package All should work just fine though each will have its own particular bugs and irregularities to overcome before becoming user friendly for a task as complex as echelle data reduction Section 3 I
58. ight characteristics Furthermore features like the meteor would require quite high order to fit accurately and may be quite tricky to remove In many cases it may be better depending on the data and application to simply steer clear of the regions contaminated by the meteor So here is the parameter list for apscatter set up to use solar imh as the input quartz imh as the reference image for the apertures and to write the output as solar ds imh solar de scattered Note that we are again getting our aperture information from the quartz image 60 HIRES Manual Data Reduction Image Reduction and Analysis Facility PACKAGE echelle TASK apscatter input solar List of input images to subtract scattered light output solar ds List of output corrected images scatter List of scattered light images optional referen quartz List of aperture reference images interac yes Run task interactively find no Find apertures recente no Recenter apertures resize no Resize apertures edit no Edit apertures trace no Trace apertures fittrac no Fit the traced points interactively subtrac yes Subtract scattered light smooth yes Smooth scattered light along the dispersion fitscat yes Fit scattered light interactively fitsmoo yes Smooth the scattered light interactively line 900 Dispersion line nsum 10 Number of dispersion lines to sum buffer 1 Buffer distance from apertures
59. ine as possible that fits the baseline reasonably well When done type lt q gt to quit out of each interactive fitting session and move on to the next frame to be baselined When all finished in ccdproc go take a look at your files with ccdlist and you should see all the processing operations that have now been done on them along with the dates of when the operations were done This may begin to give you a warmer feeling that something useful is now happening to your data Tracing the Echelle Orders The next step is to use the quartz spectrum to locate the positions and track the shapes of all the echelle orders We generally use the quartz since it is a nice high S N smooth spectrum with easy to find orders But one could use a spectrum of a star or some other reference if desired The 54 HIRES Manual Data Reduction point is that you want the echelle orders of this reference frame to correspond as closely as possible to where your object orders will be To find and trace the echelle orders we will use the task apall Here is its parameter list PACKAGE TASK input output format referen profile interac find recente resize edit trace fittrac extract extras review line nsum dispaxi lower upper apidtab b_funct b_order b_sampl b_naver b_niter b_low_r b_high_ b_grow width radius thresho nfind minsep maxsep order a
60. it if atmospheric dispersion is not correctly accounted for in the guiding A guiding option which calculates and correctly offsets for atmospheric dispersion is available though of course is not as effective as the eventual ADC unit will be Image Rotator An inage rotator has also been designed for HIRES but has not yet been funded So no image rotation will be available at first light The consequences of not having control of the position angle of the slit on the sky should be carefully considered when planning and executing observations TV Acquisition Guide Camera A Photometrics CCD TV is provided for object acquisition and guiding A Canon 200 mm f 1 8 lens provides a 45 arcsec by 60 arcsec field of view centered on the entrance slit CCD TV pixels are about 1 6 arcsec square but can be made bigger by on chip binning if desired with no increase in the field coverage of course A view of the slit area with the TV camera field is shown in Figure 2 Here one sees a portion of decker plate A overlying the widely open slit Two 8 position filter wheels are also provided at the CCD TV for brightness and color control One wheel contains neutral density filters while the other contains colored glass filters Table 1 shows the available ND and colored filters for the TV The neutral density filters in combination with the TV lens aperture stop control and tv integration time give the camera some 20 stellar magnitudes of dynamic range The
61. kept under strict clean room conditions A word of warning is necessary here Access is restricted solely to authorized Keck personnel All personnel entering the HIRES enclosure are required to don appropriate clean room garb full suit booties cap and mask It should never be necessary for astronomers or other users of HIRES to enter the enclosure Indeed entry by unauthorized or untrained personnel is highly likely to result in damage to the HIRES optical components and also to the person entering there are powerful remotely controlled mechanisms which can move without warning HIRES is designed to be run totally remotely either from the control room at the telescope from Hale Pohaku or Waimea and also from just about anywhere in the world these days over the Internet It can also be run from multiple locations simultaneously such as shared observing by a collaborating group Section 1 Summary of Characteristics Before discussing the principal components of HIRES it seems useful to briefly list a summary of HIRES characteristics and first light expected capa bilities Spectral range 0 30 to 1 1 microns Spectral resolution up to 67 000 Slit length up to 70 arcsecs Defined by a selection of deckers Typical spectral span per exposure 1200 to 2500 A Order separation 8 to 43 arcsecs Resolving power times slit width 39 000 arcsecs Detector Tektronix 2048x2048 CCD 24 micron pixels CCD readout noise 5 6 elec
62. l Instrument Description Figures 3 and 4 show the transmission of the various HIRES filters provided for order blocking Note that different combinations of filters require refocussing of the colli mator and this refocussing is now handled automatically whenever new filter combinations are selected The focus change will be approximately T 3 where T is the total thickness of all filters in the beam Adding filters requires moving the collimator farther from the slit by a distance T 3 All Schott filters in Filter Wheel No 1 are 3 mm thick The CuSO filter is 5 mm thick The OCLI de tector trimmer d t is 0 533 mm thick All Schott and CuSO filters have been AR coated with an optimized multi layer broad band overcoat Exposure Control Shutter During observing the CCD is normally left open to the room Its darkcover is a relatively slow mechanism intended primarily to protect the sol gel coating on the field flattener Timing of exposures requires a much faster mechanism so starting and stopping of exposures is controlled by a fast shutter behind the slit This shutter is actuated by a signal from the CCD controller crate The minimum exposure time is 1 second Collimators The f 13 7 beam is collimated into a 12 diameter beam by either of two identical collimators These collimator mirrors are spherical with matched radii and tilted by 1 75 such that the beam is reflected up towards the echelle at an angle of 3 5 One of the collimat
63. l descriptions of the mechanical electronic and software subsystems of HIRES will be provided in other reports delivered to Keck Observatory with the instrument Copies of this user s manual can be obtained from the UCO Lick Observatory Ask for as UCO Lick Technical Report No 67 11 PHEFAGE EE EE ii LISEOPFIgUIGS icc te aya ear S xo Se ks vii List of Tables sie eee vs eomm xxm XS UR xn viii Chapter 1 Introduction ue decer RD en fog ore 1 Chapter 2 Instrument Description 3 Section 1 Summary of Characteristics 3 Section 2 Description of the Light Path 4 Section 3 Detailed Description of Principal Components 6 1 Entrance Hates eua erra REC ea RA EOS 6 2 Atmospheric Dispersion Compensator ADC 7 3 Image Rotator ura adm paces o e EE e 7 4 TV Acquisition Guide Camera 7 5 Calibration Lamp System 8 6 lodine Absorption Cell 11 7 Decker Tray and Deckers 11 8 SI A E ere 14 9 Behind The Slit Filter Wheels 15 10 Exposure Control Shutter 17 11 Gollimatots a ae TAS 17 12 Eclielle ta 26 is UE A 17 13 Cross Disperser CD aaa 21 14 Camera alada a rad 23 1 Corrector LenseS oo 23 2 Hextek Primary Mirror 23 3 Field Flattener Dewar Window 24 4 D tection 14 c NS E AE OS E a 24 5 Dewar EE 27 Contents iii 15 16 17
64. lling a new field flattener Corrector Lenses These lenses are made of Corning 7940 fused silica The front corrector lens corrector No 1 is a biconvex element and corrector No 2 is a meniscus These two large corrector lenses are heavy enough and thin enough that they sag under their own weight Finite Element Analysis FEA was done to design mounting cells which would remove most of this sag The lenses thus have push supports just outside their clear apertures which remove the sag and must be properly adjusted when re installing these lenses The corrector lenses are anti reflection overcoated with sol gel This sol gel AR coat is a dip coat process and is a very fragile overcoat It must never be touched The slightest touch such as lightly brushing with a sleeve etc will damage the coating These sol gel coatings can only be done at Lawrence Livermore National Labs and are thus very difficult to re do They can be cleaned by a high pressure ethanol spray but only after removing from their cells and by qualified technicians These coatings also have 35 times the surface area of the part they are on so they are a very effective magnet for dust Thus their covers must be kept closed as much as possible and the HIRES enclosure must be periodically wiped down for dust Hextek Primary Mirror The camera mirror is a 44 diameter f 0 76 sphere It is fabricated from a lightweighted mirror blank manufactured by Hextek Corp in Tucson
65. lso be provided as well to provide dispersion compensation on lensed QSO s or other extended objects When probing the chemical abundances of globular cluster members a multi fiber input feed would be quite desirable and would yield enormous speed gains in the multiplexing Probably the fiber head would feed both HIRES and the Low Resolution Imaging Spectrograph LRIS HIRES fibers would terminate at the curved focal plane of a spherical collimator which would drop down in front of the normal collimators A cross disperser which gives much more order separation in the visible would also probably be used in this mode and could provide enough interorder space for perhaps 100 objects while still achieving good wavelength coverage Also a mirror could be installed in place of the cross disperser for multi object or longslit single order work Infrared arrays HgCdTe are also now becoming available which provide excellent sensitivity 86 and low noise imaging capability out to at least 2 6 microns The HIRES optical train is designed to be quite efficient and to produce quite good images out to these wavelengths and such an IR array could be easily installed in the camera in place of the conventional CCD s A different cross disperser would also be purchased for use with this detector HIRES is nominally designed to be used up to resolutions of about 100 000 without image slicers However in conditions of bad seeing or for much higher resoluti
66. mmon acquisition observation problems running the simulator creating and uploading set up files plan on arriving early 10 remote observing 11 pre observing run checklist CHNAKHRWN 42 HIRES Manual Observing at Keck Chapter 5 Observing at Keck Some of these sections are not yet complete Section 1 Arrival at the Mountain top Section 2 Starting Up the Instrument The mountain staff will take care of starting up the xhires and xpose control windows plus any other instrument status information windows desired by the user The HIRES CCD dewar gets filled automatically about once per day It is a good idea to note the dewar level and decide whether an automatic fill may occur during the eveining observing hours The autofill procedure does dump some cold into the spectrograph and probably moves the CCD dewar by a very small amount due to the added weight of the liquid nitrogen plus various thermal excursions in the surrounded metal structure For highest precision work I advise that one NOT allow an autofill during observing Thus if an auotfill during the night looks probable take time in the late afternoon to force an autofill before starting calibrations Section 3 Selecting a Spectral region Collimator choice The most fundamental decision a user wil have to make concerns optimizing the efficiency of the optical train for the desired spectral region The first choice in this regard concerns which collimator red or blu
67. ne calibration data frames 48 HIRES Manual Data Reduction to squeeze out the last bit of instability from the instrument as needed for the particular project Note also that these frames could and generally would end up with much more abstract names such as n0045 fits or data0131 fits when read into the subdirectory in which you plan to work If they do you might wish to consider changing their names before you get started to more obvious descriptions of what they represent so that you don t get confused later in the reductions For example quartz s might be labelled quartz1 quartz2 etc Darks of various exposure times might be called dark100s dark500s etc Try to stick with short names to minimize typing and keep names fairly distinct so that you can make most use of command line interpreters when wild carding to again save typing You should now start up from a separate xwindow the SAOIMAGE tool The reason we start this up from a separate xwindow is because it sends text at you from time to time and it is annoying to have that chatter break in on your IRAF text Now from a different xterm window than the one you used to start up SAOIMAGE move into your IRAF directory and type lt cl gt to get IRAF going Then cd into the demo subdirectory where the images are I don t understand why you can t just start IRAF from this demo subdirectory but that s ok for now Reading FITS files into IRAF The very first
68. ne may wish to attach the closest reference spetrum in time There are keywords and options for doing many such operations The point is that refspec is much more powerful than illustrated here Applying Dispersion Correction to Object Spectrum We would now like to use the task dispcor to linearize the dispersion of our program object spectrum solar ds ec imh Epar into the parameter list now for dispcor Image Reduction and Analysis Facility PACKAGE echelle TASK dispcor input solar ds ec List of input spectra output List of output spectra lineari yes Linearize interpolate spectra databas database Dispersion solution database table Wavelength table for apertures wl INDEF Starting wavelength w2 INDEF Ending wavelength dw INDEF Wavelength interval per pixel nw INDEF Number of output pixels log no Logarithmic wavelength scale flux yes Conserve flux samedis no Same dispersion in all apertures global no Apply global defaults ignorea no Ignore apertures confirm no Confirm dispersion coordinates listonl no List the dispersion coordinates only verbose yes Print linear dispersion assignments logfile Log file mode ql There are many options here but we want just a simple flux conserving linear interpolation After running this task you should do a quick sanity check and use splot to browse around the spectrum checking against known reference spectra easy to
69. nes in any order or at any time you may switch to other orders using the lt j gt and lt k gt keys to move backwards or forward among your orders Many other helpful options can be displayed as usual using lt gt in the plot mode I try to mark a few lines in the first several orders a few near the center and a few near the last few orders before letting the routine go to try finding its own lines When finished marking a reasonable sampling of identified lines across the orders you are ready for some initial fitting Type lt f gt in the plot window this stands fit dispersion One uses a combination of maxfeat and threshhold to control the number of peaks found It will only find peaks above the specified threshhold and will find up to maxfeat of these We have the maxfeat parameter set at only 100 right now so it will find a maximum of only 100 peaks above threshhold but that s ok for a first timid try Make sure they are reasonably well distributed across the full field of apertures you don t want it to find all of them near the beginning etc You will now be presented with a plot of fitting residuals vs pixel Use the cursor and the lt d gt key to delete bad points and the lt f gt to re fit the affect all features switch is already turned on If you ve done things right most of your residuals should be down well below a pixel by this point If you already know which order corresponds to a given apertu
70. nstrument computers The main instrument control computer for HIRES is makua keck hawaii edu Another Sparc station lanikai keck hawaii edu is also used during HIRES observing primarily for data reduction and analysis Since there are multiple instrument computers two or more separate instruments can be electronically on line at the same time as will often happen as one team prepares an observing run following another One instrument computer also serves as a back up for the other The instrument computers are then connected to the Keck Observatory Ethernet LAN which provides a link with similar computers at the headquarters down in Waimea The HIRES VME chassis includes one SUN 3E120 CPU card one SUN 3E340 Ethernet card eight Galil DMC 330 10 Motor controller cards three XYCOM XVME 212 input port cards three XYCOM XVME 220 output port cards and one XYCOM XVME 540 Analog logic card Most moving mechanical devices are driven by Galil DC servo motors Each optical instrument which uses a CCD has its own CCD controller system The CCD controller is based on the design described in Leach 1988 that utilizes a programmable digital signal processor to generate timing signals and manage communication with the host computer and allows remote programming of the timing waveforms and CCD clocking voltages The CCD clocks are generated with digital to analog converters while a conventional preamplifier dual slope integrator and 16 bit analog to digit
71. ntifying a decent sampling of lines throughout the orders I keep the Th Ar atlas by the keyboard at this point and just run through the format picking out the stronger lines and verifying that the wavelength predicted for each marked feature agrees to within 4 places with the atlas Whe done again type lt f gt for fit dispersion Again use cursor and lt d gt key to zap out outliers and re fit Once you feel sure that you have a solid preliminary solution you can let the routine find many more lines automatically But this time increase maxfeat to 3000 to allow it to find all the features contained in the linelist database But be careful if you haven t yet input enough correct line id s and or your match parameter is too large it can quickly find many incorrect id s and head off toward an incorrect solution Use lt y gt to find up to maxfeat features above threshhold and then lt l gt to match features to entries in the linelist database using the current dispersion solution Again when all the features have been identified type lt f gt to re fit the dispersion Now with so many features it becomes time to tune in the order of the fitting functions in x and y Again check your residuals and zap out outliers as needed Use xorder and yorder to vary the order of the fit in either direction Use show to see the fitting parameters and the rms fit You can watch the rms fit statistic as you play with xorder
72. ntinuum level because it is very hard to rectify the continuum of each echelle order right near the near the ends of each order splines have a way of heading off on their own at the end of a data set But the shapes and relative depths of all spectral features should match very closely if everything has been done properly Once you have a view of some piece of the spectrum you like you can get a hard copy by simply typing snap If you are correctly set up in the system this should output the hardcopy directly to the local laser printer That brings us to the end of our initial foray into IRAF data reduction of HIRES spectra You are now no doubt well aware that this was only a very simple but entirely respectable shot at reducing a data set IRAF is a very powerful data reduction environment with lots of rooms and corridors to explore Data sets which push the limits of dark current cosmic ray exposures low S N etc may require more calibration frames and more extensive reduction treatments 70 HIRES Manual HIRES Exposure Estimator Chapter 7 HIRES Exposure Estimator An exposure estimating program first written by UCSC graduate student Don Penrod and later enhanced by UCSC graduate student Michael Keane is available to help the user estimate signal to noise under various conditions This S N estimator is fully self prompting and very easy to use It knows about the overall efficiency of the telescope and spectrometer It also incl
73. ntrol interface It can be run in active mode where it will actually move spectrograph parts or in simulator mode not connected to the real spectrograph The simulator mode is very useful for practicing before actually going out for a run and instrument set ups can be generated and saved for use later during observing Al Conrad aconrad keck hawaii edu wrote xhires and is the contact for this software 29 The HIRES Spectral HIRES Manual Format Simulator Chapter 3 The HIRES Spectral Format Simulator Like most echelle spectrometers the spectral format is larger than the available detector real estate The HIRES optical system was designed to feed a 2 by 2 mosaic of Ford Loral 2048 CCD s This mosaic would have been some 61 mm on a side What we ended up with at first light was a 49 mm square Tektronix CCD When the length of the free spectral range of any echelle order is longer than the 49 mm dimension of the CCD holes will appear in the data since those regions of each order falling off the CCD imaging area will not be recorded These holes begin occurring redward of about 5100 A Avoiding these holes and ensuring that the desired spectral range falls properly on the CCD requires accurate positioning of the echelle format on the CCD For extragalactic objects redial velocity must also often be taken into account To aid the observer in optimal positioning of the CCD on the echelle format a simulator was developed by Steve Allen at U
74. nywhere after the starting in column 11 The principal difference between the Echelle Simulator configuration files and true FITS headers is the existence of carriage control Echelle Simulator configuration files contain carriage control and are intended to be edited by any text editor Each time the Echelle Simulator is run it outputs hidden versions of the three configuration files These are named ech spc ech det and ech set These can be compared with the original inputs and any changes made by user interaction to verify that the program is working as desired Upon request of the user the program also writes out an observation setup file in either of 2 formats The first format is identical to the inputs described below The second format contains FIORD commands designed to command the Keck HIRES spectrograph to the given configuration The user can also edit these files as desired without going back and rerunning the simulator using your favorite text editor There are several keywords which are defined by the Keck HIRES data acquisition system which are not used by the instrument simulator The instrument simulator accepts these keywords and carries their values from input to output unchanged An observer may have a number of key spectral features which need display ing at their respective positions on the echelle format The Echelle Simulator will accept a file containing the wavelengths of spectral lines and display th
75. oints to combine 52 HIRES Manual Data Reduction niterat 5 Number of rejection iterations low rej 5 Low sigma rejection factor high re 1 75 High sigma rejection factor grow 3 Rejection growing radius mode ql Input the names of the images you wish to correct i e reduce or the name of a file which contains a list of the images to be corrected For our example these will be solar imh quartz imh thar imh dark imh and zero imh Type them all in on one line with commas but no spaces between The ccdtype parameter being set to object tells the routine to perform dark current correction only to files with IMAGETYP object in their FITS headers The max_cac parameter can be increased as allowed by available memory to speed processing time Oversca yes means we will be correcting each image for row to row baseline variations using the overscan strip presently image columns 2070 to 2112 The baseline will be measured in the overscan area for each row and then a smoothed version of this baseline measure will be subtracted row by row from the image Trim yes will trim each image stripping off the first 22 prescan columns which do not contain real image pixels and the columns beyond 2070 which at present are reserved for overscan pixels Zerocor yes will cause the image with IMAGETYP zero in its FITS header i e our file named zero imh to be subtracted from each image to remove the bias level cu
76. on and dark current correction of each object image note that we are including our quartz and our thar images as object images which also need correcting The images will also be trimmed of their prescan and overscan regions When happy with the set up type lt g gt which will send you onward Note if IRAF has some trouble locating or otherwise deciding upon correct image types for your data files it may well return immediately to the command line prompt without actually doing anything nor telling you that it didn t do anything The processing will take a while so if it returns immediately something is probably wrong If all is well IRAF will begin trimming the frames and extracting baseline information Since we specified interactive baseline fitting it will put you into the interactive baseline fitting mode for each frame Try playing with the baseline fitting from the tektronix plot window displayed If totally lost as to what to do next from the tektronix plot window just type lt gt and you ll get a help summary But you must lt q gt twice out of the help summary to get back to interactive mode in the tektronix window Useful commands to play around with in interactive fitting of the baseline or for many interactive fitting tasks are high high reject level sigma units low low reject level niter number of iterations order and show to display these parameters Try to get as low order of a spl
77. on work image slicers can be added to maintain high throughput at the slit The collimators are oversized to accept the square beam from a Richardson style slicer and the camera s image quality will be good enough to provide resolutions of at least 200 000 CCD s with 7 5 micron pixels will also be required for such resolutions and appear to be now available in 4096 formats One could also envisage a double pass very high resolution mode with the cross disperser rotated to send the light back to the echelle but tipped slightly such that the collimator could also act as the camera and produce an image up near the slit This image could be picked off by a small mirror and sent to a detector mounted up near the slit Finally HIRES can also be extended out through the future expansion door in the wall near the cross disperser A symmetrical outrigger could be added to the optical bench structure and the cross disperser used to steer the light to perhaps a different focal length camera 1 Image rotator 2 Atmospheric Dispersion Compensator 3 Image slicers or adaptive optics 4 tip tilt system 87 Appendix Acknowledgments HIRES was built by a superb team of people at UCO Lick Observatory Former UCO Lick director Bob Kraft and present director Joe Miller contributed much useful scientific technical and managerial input Neal Jern was the overall project manager Jack Osborne and Bruce Bigelow were the mechanical engineer
78. or and a thick fused silica field flattener which also serves as the dewar vacuum window Inside the dewar at the camera s prime focus is a Tektronix 2048EB2 1 CCD A slowly actuated dark cover is also provided at the field flattener dewar window to keep this sol gel coated optic clean and to keep the CCD reasonably dark if lights must be turned on inside the spectrometer room The LN dewar near the CCD is filled automatically about once per day from a large LN storage dewar sitting outside the HIRES room The storage dewar needs manual re filling about once per week by qualified CARA technical personnel Section 3 Detailed Description of Principal Components Entrance Hatch The entrance hatch is a simple hinged door It is normally kept closed when not using the instrument for any extended period It serves to isolate the slit area from dome light such that calibrations can be made during the day or while someone else is using the telescope It also serves the important function of keeping dirt and airborne contaminants out of the slit area so please keep it closed when the instrument is not in use HIRES Manual Instrument Description Atmospheric Dispersion Compensator ADC An atmospheric compensator ADC will eventually be installed though there will not be one at first light Since the HIRES slit will not in general lie along the parallactic angle losses particularly in the ultra violet could become substantial at the sl
79. or mirrors has an enhanced 2 layer dielectric over aluminum coating and is for use over the 0 3 to 0 5 micron spectral region This is called the Blue collimator The other collimator is coated with an enhanced silver recipe the holy grail and features somewhat higher reflectivity in the 0 34 to 1 1 micron range but drops off sharply below 0 34 microns This collimator is called the Red collimator Figure 5 shows the reflectivities of both collimators Echelle The echelle is a mosaic of 3 of the largest echelles currently available The mosaic is 12 by 48 in size The ruling is 52 68 grooves mm and the blaze angle is 70 5 The collimator to camera angle 20 is 10 0 The echelles are in tentionally pistoned in the mosaic such that the two gaps are maximally shadowed minimizing light loss at the gaps The echelle mosaic alignment is intended to be passive and should never need adjustment Alignment is maintained by clamping the echelles to a large granite subplate minimally constrained so as not to intro duce any moments or unwanted forces The mounting scheme is intended to be 17 Instrument Description TRANSMISSION 0 9 0 8 0 7 0 6 0 5 0 4 0 3 0 2 0 1 HIRES Manual Figure 3 Blue Blocking Filters WG335 gt WG360 KV370 KV380 KV389 KV408 KV418 GG475 RG610 0 3 0 4 0 5 0 6 WAVELENGTH microns 18
80. ose lines with any desired velocity shift on its graphics The existence of such a file can be indicated using the WAVEF ILE keyword in the Setup Configuration file and it can also be indicated interactively during the execution of the program Each line of the file contains a description of one spectral line The program looks for a wavelength expressed in Angstrom a boolean value T or F which describes whether the line is telluric and thus should not be redshifted and a statistical weight The statistical weight is used by the program during the design of new 32 HIRES Manual The HIRES Spectral Format Simulator spectrographs It is intended to assist the program in choosing an Echelle groove spacing which places certain spectral lines near the blaze Section 2 Starting the format Simulator Starting the Echelle Simulator can be done by typing the command echelle or echelle amp if you want to run it in the background and keep the window available for other input If the user is running the X Window System and the user s environment defines the DISPLAY variable the simulator will assume that the graphics should be displayed in an X Window If the DIS PLAY variable is not set the program will prompt the user to enter one of the terminal types known to Lick Mongo The simulator then searches the current directory and the library directory looking for a setup file to be used to display the Echelle form
81. ow 0 38 microns This is rather a shame since the CCD effectively dies before the 2nd order CD efficiency curve rises to a peak thus producing a dip in the overall instrument throughput in the 0 35 micron region All the rest of the HIRES optics transmit very efficiently all the way down to below 0 3 microns One very prominent distinguishing cosmetic defect of this CCD is a large felt tip pin mark near the center of the CCD It was kindly added by some technician at Tektronix to remind us that this virtually flawless 100 000 CCD is only an engineering grade device Unfortunately the folks at Tektronix seemed to have forgotten how to make science grade devices so we are stuck with this annoying blob It is marked in the HIRES format simulator as a red square though it is irregular in shape Take care to avoid this region when positioning critical spectral regions on the CCD 25 Instrument Description HIRES Manual Quantum efficiency Figure 7 Tektronix CCD Quantum Efficiency Tek2k LRISEng2 T 130 0 2 03 04 05 06 07 08 09 1 0 wavelength microns 26 HIRES Manual Instrument Description Dewar Focus Fine focus of the camera can be done by using the dewar focus mechanism This mechanism moves the entire dewar CCD plus field flattener along the camera axis Total travel is only about 0 03 but with very high precision The camera is so fast f 1 0 that some care is
82. ow wish to check your directory to see that all the corresponding imh files have been created You might also want to check your disk space with a lt df gt to make sure you have enough room to keep going Once data files are read in with rfits and converted to imh files each imh file will actually be carried around as both a header file imh and a corresponding pix pixel file i e where all the pixel information is kept The pixel files live in a directory specified by the logical variable imdir In my setup imdir is set to HDR pixels This setup puts the pixel information in a subdirectory called pixels and you can cd there and Is them to see that they really do now exist The main point here is that IRAF image files are actually associated pairs of header and pixel files Thus when copying deleting etc such files it is generally much easier to use the commands imcopy or imdel etc since these commands also keep track of all the housekeeping for the associated image files Checking Header Information It is useful at this point to check your headers to see that all looks well and more importantly that you have the right keywords describing the type of images in each case Initially this is a must since we are still working out FITS keyword assignments but soon this step will not be necessary unless of course you screwed up and recorded say an object frame which was really a dark frame or somet
83. pertur npeaks IRAF Image Reduction and Analysis Facility chebyshev 1 28 21 21 28 3 0 O 0299 40 40 10000 30 50 1000 increasing INDEF List of input images List of output spectra Extracted spectra format List of aperture reference images List of aperture profile images Run task interactively Find apertures Recenter apertures Resize apertures Edit apertures Trace apertures Fit the traced points interactively Extract spectra Extract sky sigma Review extractions Dispersion line umber of dispersion lines to sum etc 4 DEFAULT APERTURE PARAMETERS Dispersion axis 1 along lines 2 along columns Lower aperture limit relative to center Upper aperture limit relative to center Aperture ID table optional DEFAULT BACKGROUND PARAMETERS function function order sample regions average or median rejection iterations lower rejection sigma upper rejection sigma rejection growing radius Background Background Background Background Background Background Background Background APERTURE CENTERING PARAMETERS Profile centering width Profile centering radius Detection threshold for profile centering AUTOMATIC FINDING AND ORDERING PARAMETERS Number of apertures to be found automatically Minimum separation between spectra Maximum separation between spectra Order of apertures RECENTERING PARAMETERS Select apertures Select brightest peaks
84. r imh and thar imh you will have to imdel these files and rfits in both solar fits and thar fits again It is also a good sanity check now to display the newly flattened solar and thar files They should look clean with nice dark spaces between orders and perhaps a bright line along the edges of each order Note that they have only had the high frequency pixel to pixel variations removed at this point and will still show the low frequency blaze profile variation in intensity That will be removed later It is also humbling and scary to roam around a bit on the histogram equalized thar frame searching for ghosts You ll see a number of them as well as the meteor Removing scattered background light The next task is to measure and remove the scattered background light This is light which shows up between the orders and which results from scattered light ghosts and other reflections inside the spectrograph You will use the task apscatter in the echelle package of routines so bye out of the generic package if you aren t already out and get into the echelle package now Basically apscatter will allow one to interactively fit a function using to the interorder light both in the row and column directions This 2 d functional fit to the background is then subtracted from the given image Note here that scattered light will depend strongly on the illuminating source and thus each individual data frame will have different scattered l
85. re you can use the o option in fit to set this If you are not setting the order number directly check to see if it solved correctly for the order offset the offset between your aperture number and the true interference order at the echelle If it got the offset correct this is a good indicator that you are on your way to a solid and correct solution Be 65 Data Reduction HIRES Manual careful though it could still be off slightly in echelle order you may need more features to really nail this down Now you can type lt q gt to quit out of this fitting subroutine and return to displaying the orders again You can now either manually hunt down and mark other features to improve the fit or quit out and write your initial solution to the database To identify other lines quickly by hand say by looking at the Th Ar atlas simply move the cursor near a line and hit spacebar to mark the nearest line It will then prompt you with the pixel position and the computed wavelength from the current solution If this computed wavelength matches within a window set by the match parameter a line in the linelist database it will return the tabled wavelength Otherwise it returns with INDEF and waits for you to input a wavelength If you are happy with its tabled wavelength just hit return This enormously speeds the entering of many more lines to really pin down the dispersion fit Use lt j gt and lt k gt to move among the orders now ide
86. rently set near 1100 dn Darkcor yes will instruct the routine to use the image with the IMAGETYP dark keyword in the FITS header i e our file called dark imk to determine the dark current dn pixel sec and then scale that dark current to the correct exposure time for each image before then subtracting that dark current The other processing switches won t be used right now Most are fairly self explanatory Readaxi line tells the routine that the readout axis is along the row line direction We don t have a fixfile image yet for descibing and fixing bad lines or columns The biassec and trimsec parameters shown here are correct for defining the overscan and trim regions for the present single amplifier readout configuration They will be different for dual amplifier readout and they may change also as we do more optimizing of the CCD Check the locations of these araes yourself to see that things are where you expect them You don t need file names for zero dark flat etc since these files will already have the correct IMAGETYP FITS header keywords and the routine will be smart enough to recognize them as such For this example we will be fitting the overscan region 53 Data Reduction HIRES Manual interactively using a cubic spline So before moving on let s summarize what we are now about to initiate when we lt g gt out of this parameter list We are set up to do overscan correction bias zero frame correcti
87. riations i e blaze profile and create a normalized flat field image which I will call flat The reason we would like to normalize out the low frequency quartz spectrum variations is that doing a straight division by a quartz frame will give excess weight to those pixels where the quartz illumination happens to be low either away from blaze or away from the center of the aperture One could not go on to perform an optimal extraction if you flat fielded in this manner Here is the apnormalize parameter file for doing this using quartz imh as the input and flatimh as the output Image Reduction and Analysis Facility PACKAGE echelle TASK apnormalize input quartz List of images to normalize output flat List of output normalized images referen List of reference images interac yes Run task interactively find no Find apertures recente no Recenter apertures resize no Resize apertures edit no Edit apertures trace no Trace apertures fittrac yes Fit traced points interactively normali yes ormalize spectra fitspec yes Fit normalization spectra interactively line INDEF Dispersion line nsum 10 Number of dispersion lines to sum cennorm no ormalize to the aperture center thresho 10 Threshold for normalization spectra backgro none Background to subtract weights none Extraction weights none variance pfit fitld Profile fitting type fitld fit2d
88. rithm This blob is a mark from a felt tip pen kindly put on this 100 000 CCD by some technician at Tektronix to remind us that this is an engineering grade device The parameters lower and upper define the lower and upper limits of the tracing aperture relative to order center The parameter b_sampl defines the background sample region and has been set to run just outboard of the tracing window here The parameter nfind must be at least as big as the number of orders to be found The parameter minsep must be set correctly but the parameter maxsep can be any number much larger than your order separation So when ready type lt g gt to exit and execute apall and start answering interactive mode prompts It will ask questions about whether or not you want to resize apertures for quartz edit apertures for quartz etc Your answers will depend on whether this is your first time through or whether you have already run the routine and have previous aperture information stored away in the database subdirectory so answer carefully When satisfied with each fit to each order type lt q gt to quit and move on to the next order As with all interactive question and answer sessions within IRAF If you get tired of answering yes s just type YES all caps If all the parameters are set reasonably well apall should have no problem finding all orders If it does have problems and you should check by going through in interactive mode at least on
89. rtures per aperture mode ql Obviously there are a large number of parameters in this very generalized and powerful task The ones shown here worked well for a shot in 1st order of the cross disperser in the 4800 to 7000 angstrom range where the decker used projects to a height of about 38 rows You will probably have to modify several parameters if your order spacing and or order widths set by the decker length usually are different Some of the most crucial parameters to set properly are width and radius I m told that the width parameter must be just slightly bigger than the width in rows or lines of the widest echelle order In our example this is about 38 rows so I set width to 40 Also the radius parameter should be set to the same value Eventually we will be able to use the instrument keywords and or set up routines to set this parameter automatically but for now you must display your image in SAOIMAGE using the task display roam around looking at order widths and 56 HIRES Manual Data Reduction decide upon the correct value for width by measuring the width of the widest order Actually all orders should be very nearly all the same width save for slight anamorphic and distortion effects Here are some other noteworthy parameters The parameter line is the column where the order tracing routine starts its search Tve set it at 900 to avoid starting in the dark blob near chip center which may confuse the algo
90. rver the prompt will change the cursor into a question mark and the prompt will be visible at the bottom of the screen On other displays the prompt will appear on the text screen if one exists or on the graphics screen Again this is where you will want to be sure your text screen is not buried under the graphic screen Mouse drag can only be used on X11 displays Dragging can be done with fixed size objects or with rubber objects The Instrument Simulator allows the readout window if windowed down to some subset of the full CCD format to be indicated by dragging a rubber rectangle over the display The position of the detector s can be modified by dragging a fixed size rectangle of the same size as the detector When dragging a fixed size rectangle it may be grabbed at any of 9 locations defined by the corners and points halfway between Menu click can be used for items which are displayed in the lists of text at the side of the display It requires that the display have some kind of moveable cursor The cursor is moved over the menu item and any unassigned key or mouse button is hit The user will then be prompted for a change in the same manner as for accelerator keys Modifying the display The overall display can be modified by the use of single accelerator keystrokes or menu clicks 36 HIRES Manual The HIRES Spectral Format Simulator Show Display the complete menu of optical Optical HW hardware Show All Setup Di
91. ry searched is the current working directory Next if the environment defines EFDIR that directory is searched otherwise the program looks in the built in default directory The simulator accepts 3 kinds of configuration files The first 2 kinds of files rarely need changing they describe the telescope spectrograph optics spc and the detector at the focal plane det These files will be maintained and updated by Keck Observatory folks and will be write protected from general users The third kind of file contains the settings of all the moveable parts of the spectrograph which are expected to change from one observation to the next set It is this type of file that the observer will be creating modifying and storing for later use with HIRES 1 On the UCO Lick systems this is home hires sla echelle lib 31 The HIRES Spectral HIRES Manual Format Simulator The configuration files for the Echelle Simulator look like FITS files More detailed descriptions of these configuration files can be found in Steve Allen s UCO Lick Technical Report No 68 Configuration parameters are stored as keyword value pairs In accordance with FITS files the keywords are up to 8 characters long Many of these keywords are identical to the keywords which will be used by the Keck Data Acquisition System when it is documenting actual observation Each keyword is followed immediately by in columns 9 10 The values may be found a
92. s Harland Epps did the optical design optimization Carol Osborne assisted with drafting Master Optician Dave Hilyard did the optical fabrication with help from opticians Darrie Hilyard and Gerard Pardeilhan Terry Ricketts did the electronics design Lance Bresee and Cal Delaney assisted with the electronics fabrication and checkout Bob Kibrick Richard Stover Al Conrad Dean Tucker Steve Allen Kirk Gilmore and Mike Keane provided software support Erich Horn Jeff Lewis Terry Pfister Dick Kanto and Jim Ward did the mechanical fabrication Bill Brown assisted with optical coatings Lloyd Robinson and Mingzhi Wei developed the CCD detectors Marlene Couture and Joe Calmes did the accounting with Ted Cantrall providing help with the project scheduling and purchasing 88
93. slit is about 11 1 mm Behind The Slit Filter Wheels There are two 12 position filter wheels behind the slit These are primarily for blocking unwanted orders from the cross disperser Eleven positions are for 2 by 2 square or 2 diam round filters while position No 1 is a long narrow clear slot and cannot be loaded with a filter Table 4 shows filters currently available Filter positions marked clear are empty and can be used temporarily for a user s personal filters Personal filters may only be loaded by the Keck technical personnel and should be removed at the end of one s run If users wish any filters to be permanently added to the selection please contact S Vogt Table 4 HIRES Filters SERGE Sesso evs ETC 9 sesso esesso Proper choice of order blocking filters is absolutely crucial to one s success in isolating any particular spectral region of interest Ultimately the user must 15 Instrument Description HIRES Manual Table 5 Some useful filter combinations Order Range microns 1 Position 2 Position bear responsibility for choosing the correct combination of blocking filters and may have to refer to filter transmission curves In particular real filters do not have infinitely sharp cut on curves and this can make a difference in some cases However as a first cut guide to selecting appropriate filters the following table lists some useful combinations 16 HIRES Manua
94. solution spectroscopy standards on single objects and to give a relatively large throughput or slit width times resolution product without the need in general to image slice at the entrance slit The nominal throughput of HIRES when using a detector with 24 micron pixels is about 39 000 arcsecs which means that a 1 arcsec entrance slit yields a resolution of about 39 000 It achieves this relatively large throughput in spite of the very large diameter of the telescope primary by a combination of a large 12 diameter collimated beam a large 48 long echelle grating mosaic and very fast f 1 0 exquisitely achromatic camera The optics and image quality are optimized for use over the entire 0 30 to 1 0 micron spectral region without refocus and could readily be extended to 2 0 microns by replacing the optical CCD with an IR array detector A generous amount of room has intentionally been left between echelle orders to allow for adequate sky sampling a factor which can become quite important in bright or grey time when pushing to faint limits This interorder room could also be used for image slices though image slicers are not being provided at first light Indeed it is hoped that within a few years adaptive optics may provide a better solution to squeezing the target image down a very narrow slit in the Introduction HIRES Manual very high spectral resolution regime Alternatively this interorder room could be
95. splay the complete menu of instrumental setup Die Display more information about the detector s Default Setup Restore the originally displayed menus Refresh All OO R O Redraw everything ema z Zoom the display to show only the region where the detector is currently located Exit2CmdLinMod Quit the graphical interaction and begin command line interaction see Section 4 below EEN Wes Un Display the wavelengths of every fifth order e Un Display the order numbers of every Display Orders fifth order Un Plot a temporary outline of the detector s at the current location of the Ma Deer detector These outlines will be visible in a hardcopy control I Identify the spectral line nearest to the X Mouse 1 cursor Move the detector s over the Echelle cy Z Define the readout window of the detector s 37 The HIRES Spectral HIRES Manual Format Simulator Modifying the Setup DETFILNM None SPCFILNM Noe OBSERVER None ECANGLE ECangle deg M X Mouse 2 DANGLE XDangle deg M X Mouse 2 HAMHGT M X Mouse 2 ECANGRAW M X Mouse 2 DECKER None DECKRAW None DECKPOS None DECKNNAM None DECKHGT None DECKSIZE None DECKSPEC None SLITVEL None SLITRAW None ILTER2 None ILNAME None IL2NAME None 38 HIRES Manual The HIRES Spectral Format Simulator ColFocuslmE Cam Foetal Modifying the Optics The Echelle Simulator can be us
96. t to the farthest corners of the hexagonal primary beam then proceeds to focus at the HIRES slit plane The HIRES slit is actually 1 83 behind the nominal telescope nasmyth focal plane but well within the focus range and good imagery range of the telescope The slit plane is tilted such that light can be reflected up at an angle and re imaged onto a CCD TV acquisition and guiding camera This camera is a simple fixed CCDTV staring at a 45 arcsec by 60 arcsec field centered on the entrance aperture of the spectrometer The TV camera has both color and neutral density filters for brightness and chromatic control on the guide target and also has variable focus and aperture control At the nasmyth focal plane a bi parting precision slit is provided for adjusting spectral resolution A series of decker plates just above and in very close proximity to the slit jaws is provided for defining the entrance slit length Some of these decker apertures decker plate A slots are for defining slit length only and are used in conjunction with the slit while all other decker apertures define both slit length and width and are used in place of the slit jaws When using the latter the slit jaws must be fully opened to get them out of the way HIRES Manual Instrument Description Figure 1 HIRES Schematic collimators echelle mosaic rotator iodine shutter cell FW1 FW2 field flattener cross disperser mosaic corrector lenses
97. task is to get your FITS files read into IRAF as imh files Get into the parameter editing mode of the rfits task by typing epar rfits This is what you will see IRAF Image Reduction and Analysis Facility PACKAGE TASK dataio rfits fits_fil solar fits quartz fits thar fits dark fits zero fits FITS data source file_lis 1 File list iraf_fil solar quartz thar dark zero IRAF filename make im yes Create an IRAF image long he no Print FITS header cards short h yes Print short header datatyp ushort IRAF data type blank 0 Blank value scale yes Scale the data oldiraf no Use old IRAF name in place of iraf file offset 0 Tape file offset mode ql On the first line for fits_fil enter the names of the files to be read in I ve already typed in the fits_fil and iraf_fil lines for you These filenames will all 49 Data Reduction HIRES Manual be entered on one line with commas and no spaces between names They are solar fits quartz fits thar fits dark fits zero fits Using the same sequence of names for the iraf_fil parameter will give the IRAF files the same names but the extension will be imh I like to do it this way to avoid getting confused with renamed files It is also very important that the datatyp parameter be set to either ushort unsigned short or real and that the scale parameter be set to yes Once finished type lt g gt to exit and execute You may n
98. thin 0 02 A light shroud around the edge of the lens helps to keep the CCD reasonably dark but is not completely leak proof A slow operating dark cover consisting of a cap which moves into place over the field flattener helps to keep fingers and other foreign invaders away from the sol gel surface of the lens and provides some protection from flashlights or other low level lighting used by personnel entering the room Again this dark cover is not expected to be completely light proof so care should be taken when entering HIRES to avoid using any more photons than necessary unless one is prepared to suffer potential image afterglow effects And the dark cover should always be put back into place when the system is not in use to provide maximum mechanical and dust protection for the sol gelled lens Detector We will be using the engineering grade CCD from the LRIS instrument at first light This detector is a Tektronix 2048EB2 CCD This CCD is optimized for the visible and shows a pronounced roll off in quantum efficiency in the ultraviolet The science grade Tek CCD for HIRES has not yet been nor may ever be received The CCD features 24 micron pixels in a 2048 by 2048 format It is a thinned backside illuminated chip with surface treatment and AR coating 24 HIRES Manual Instrument Description Figure 7 shows the quantum efficiency for this CCD as measured at UCO Lick at operating temperature Note that the QE drops very rapidly bel
99. tiple vs single amplifier modes 45 Section 6 Section 7 1 S Go OO N N 5 Section 8 Section 9 Section 10 Section 11 Chapter 6 Section 1 Section 2 Section 3 1 0 ON OOF CO Iv ck O N FOCUSSINO EE eee Wo s Taking Calibration Exposures Flat ftieldS cios o Ecke S Es oe Wavelength calibration and instrumental profile Th Ar hollow cathode lamps E B Darids don do ER de ER lodine Absorption Cell Spectrum Dark Frame s i353 ESL Rea Bias zero Frame s Day Twilight Sky and or Moon Spectra End of the Evening End of the Observing Run Observing Checklist Observing Log Sheets Data Reduction we ee eee ete wee hw mn EIGAB Vua ar aU bust esto EE da IDL KHOROS and Others IRAE oak e iuto Ee aoo EE A o lee eda Getting set up with data in the appropriate directory exis ES eek hs ES Bad ate Reading FITS files into IRAF Checking Header Information Setting the Instrument Parameters CODAED bob ENNEN CCDPROC wa EEN E tg E ed Tracing the Echelle Orders Generate the Flat Field Image Flat Fielding the object images Removing scattered background light Extracting Orders and Compressing to 1 d Specta ENEE Wavelength Calibration V 13 14 15
100. to aid in finding features If you get it set up fairly closely to how the data was obtained you can read wavelengths off at any column in any order quite rapidly Wavelength fitting is a bit tricky since if you get a few wrong identifications near the start you can end up going down an incorrect path and arrive at an incorrect solution This is largely because there are so many Th Ar lines that the routine can almost always find one near where it thinks one should be And then if you let it start finding its own lines without properly constraining it enough it will quickly accumulate a list of incorrectly identified lines which will overwhelm your relatively small list of proper identifications and converge on the wrong solution So you have to start out slowly giving it a few tens of lines scattered across the format and letting it find small numbers of its own When confident that it is finding lines correctly then you can turn up maxfeat to the maximum of 3000 and 63 Data Reduction HIRES Manual really let it go hog wild But to begin with I d suggest setting maxfeat at about 20 and inputting 10 to 20 features manually Of course the easiest landmarks to identify are the very bright Argon lines which appear redward of about 7000 A but if you don t have any of these in your spectrum stick with the strongest features at first Here is a reasonable starting parameter list for ecid Image Reduction and Analysis Facility PACKA
101. trons rms 9 CCD dark current lt 10 e pixel hour 10 2 pixel projected slit 0 60 arcsecs 11 Image de rotation none 12 Atmospheric dispersion compensation none OO pace Instrument Description HIRES Manual 13 Acquisition and guiding fixed Photometrics CCD TV staring at a 45 arcsec by 60 arcsec field centered on the entrance aperture 14 Calibration sources quartz halogen Deuterium lamp Th Ar hollow cathode Iodine absorption cell Edser Butler FP diode laser Section 2 Description of the Light Path A simplified schematic of the HIRES instrument is shown in Figure 1 This figure was also used as the model for xhires the graphical user interface through which the user controls the instrument It is well worth studying this conceptual diagram and using it as a guide when setting up the instrument When configuring HIRES for a given observation I like to go through this diagram as a photon travels element by element checking the options for and settings of each element Light from the telescope enters from the left in this schematic coming to focus at the f 15 nasmyth focus An entrance hatch at the front of HIRES seals off the slit area such that the instrument can be run during the day under high light level conditions in the dome The hatch also protects the slit area from dirt contamination and should be left closed when the instrument is not in use for any prolonged period The f 15 it s f 13 7 actually ou
102. udes provisions for calculating slit losses under various seeing conditions absorption by the atmosphere at any input airmass and effect of moonlight on sky background subtraction This program resides on the UCO Lick network computer system as home umbra mk bin sparc sn Keck observatory should also have a version for public release Contact aconrad keck hawaii edu for assistance Be sure you get an up to date version there are some older versions floating around with incor rect efficiency numbers Then just answer the questions as prompted If in doubt about an input value the default will often suffice The program also remembers all previously used values and reuses these as new defaults to speed up repetitive inquiries Efficiencies used for the S N calculations are our best estimate from actual first light performance at the telescope during commissioning Here is a sample of the prompts you will receive when computing the S N for a given set up of HIRES Wavelength of interest A 5500 Blaze center A 5517 Free spectral range A 86 20 Enter cross disperser order 1 Spectrograph efficiency at 5517A blaze peak is 18 4 Blaze function is 87 9 Single order efficiency at 5500A is 16 1 Extinction at 5500A is 0 12 magnitudes airmass Enter slit height arcseconds 13 13 Enter slit width arcseconds 1 000 Enter seeing FWHM arcseconds 0 8000 Slit throughput is 74 1 Slit width projects to 3 5 pixels
103. ulator The executable code for the simulator will reside on the Keck Observatory computer system Observers will have to get guest accounts at Keck to access this code Alternatively binary versions of the code running under SunOS may become available from UCO Lick Observatory and can be FTP d to the user s institution The most common platform on which the instrument simulator will be used is probably a high resolution monitor running version 11 of the X Window System from MIT However it is essential to note that the instrument simulator is NOT an X11 based program The instrument simulator uses the Lick Mongo package to do its graphics and user interaction This allows the instrument simulator to be run on a variety of platforms dating back to Tektronix storage tube terminals Because of this the instrument simulator cannot do multiple popup windows and menus as would be expected of a modern X11 based user interface Section 1 Before starting some words about Configuration Files The simulator is a general purpose tool which requires configuration files to give it the information necessary to compute a simulated HIRES spectrometer and its particular detector The complete description of an Echelle spectrograph requires copious amounts of information Most of this information does not change and it is convenient to store it in configuration files The simulator searches for configuration files in several directories The first directo
104. ultra precise radial velocity work can be found in Marcy and Butler 1992 Decker Tray and Deckers Immediately above the slit is a tray containing a series of 4 decker plates These decker plates are highly reflective for guiding 11 Instrument Description HIRES Manual and made of type 420 Stainless steel which were EDM d to shape and then polished in the UCO Lick optical Lab These deckers define the length of the effective slit seen by the spectrometer Some of the deckers in plate A define slit length only while others define both slit length and width As such these latter deckers are to be used without the underneath slit i e the slit gets opened up wide so it is out of the way They may be more effective for guiding on faint objects since there is then only a single focal plane whereas using the deckers in plate A in conjunction with the slit results in two slightly separated focal planes with some resulting corruption to the reflected guide image If one is using a decker which also defines slit width i e spectral resolution and wavelength zero point one must bear in mind that wavelength zero point will change if the decker tray slide which moves perpendicular to the slit axis is repositioned Thus one may expect to have to take new wavelength calibrations if the decker tray is repositioned while using such deckers Also the collimator mirror must be refocussed when using deckers only rather than deckers slit since the de

Old User`s Manual - WM Keck Observatory

Contents

Download Pdf Manuals

Related Search

Related Contents