Essence/SuperMacho Observing Procedure

CWS Nov 5, 2002
Updates: 2003, 2004 by Chris Smith, Claudio Aguilera
Updates: 2005 by Tamara Davis, Chris Smith, Alfredo Zenteno
Update: 2005.11.08 by Chris Smith Update: 2006.09.13 by Chris Smith

How to observe

  1. Establishing the video link
  2. Starting up the MOSAIC imager
  3. Checking disk space
  4. In the ctioa0 VNC
  5. At ctioa8 Unix prompt
  6. Coping ocs files
  7. Taking Zero images
  8. Taking Dome flats images
  9. Doing Zero point
  10. Taking Standard Stars
  11. Running pipeline (*No Longer Necessary)
  12. Doing Focus
  13. Using mosocs mode
  14. SuperMacho observing
  15. End of the night
  16. Any problem during the observing?
  17. Fixing headers

------

INTRODUCTION

This document describes procedures for taking Essence/SuperMacho data. It assumes familiarity with the MOSAIC imager (see www.ctio.noao.edu/mosaic ).

Nights are shared between the two programs, but procedures are very similar. This document presents things you are to type between quotes, so when presented with

type "echo hello"

you are to type the stuff between the quotation marks.

------

ARCHITECTURE SUMMARY

A machine called ctioa0.ctio.noao.edu runs the camera electronics and another, ctioa8.ctio.noao.edu, actually accepts and stores the images. On both of these we use the username mosaic.

For scripting and other generalities, we use /ua80/mosaic/smsn as the home directory of the smsn project. This DOES NOT MEAN that the data is being written to /ua80! /ua80/mosaic/smsn is a LINK to a much larger disk. The actual disk we're using can and WILL change over time. Check the link to find out where we're actually writing the data this year.

For 2005, the directory we use to store a night's images resides on /ua85 disk on ctioa8.

The generalized path to a night's data is /ua80/mosaic/smsn/sm020930, for example. This is for the night of Sept 29-Sept 30 (Chilean time). This corresponds to UT 020930, which determines the directory name.

Images are transferred to the data analysis cluster in La Serena. This transfer is initiated at soon as an image is read out via the postproc mechanism. This can be toggled on and off (we'll see how later).

Each night's observing sequence is determined by a pair of files, with .ocs suffixes. The default observing plans reside in /ua80/mosaic/smsn/sm.ocs for SuperMacho and /ua80/mosaic/smsn/walpha.ocs (or wbeta.ocs) for ESSSENCE. These files are used by the mosocs utility in the data acquisition system.

------

PREPARATIONS

  1. Establish the video link between CTIO and La Serena
  2. Start up the MOSAIC imager, either on the mountaintop or in La Serena, using the VNC windows. To run the VNC windows in La Serena:
  3. We are observing from ctiokq in the Remote Observing Room this year.

  4. Check disk space:
  5. Either open a window in the ximtool VNC or open an xgterm and log into ctioa8.ctio.noao.edu as user mosaic. Be sure there is enough room on the data disk /ua85. Use "df -h". We need about 25 GB per night. Move and delete images if needed (Just delete fits files).

  6. In the ctioa0 VNC, right click on desktop and select End Arcon Session.
  7. At a ctioa8 UNIX prompt, run the ~/smsn/bin/startsm 06**** script.
  8. This takes an argument which is the night's date string, for example 060914, ***WITHOUT*** the sm in front. NOTE: the command should look like "~/smsn/bin/startsm 060913", the date should NOT have the typical sm in front.

    This script will then:

    Create a directory for the night's data, conforming to the convention /ua80/mosaic/smsn/sm020930. It will akso create links "tonight" in the directories /ua80/mosaic/smsn, /ua80/mosaic, and /ua00/mosaic, so from all of the "home" directories you just have to type "cd tonight" to get back to tonight's directory.

    Create a directory for the night's data downtown (on ctiokc) in the mscraw data directory. A "tonight" link will also be created there.

    Copy the "sm.ocs" and "wxx.ocs" (from /ua80/mosaic/smsn/) file into the night's directory.

    Create an observing log file for the night, conforming to the naming convention /ua80/mosaic/smsn/sm020930/sm020930.log.

    Run skycalc for the night, and put the alamanac info into the observing log.

    Append the two ocs files into the observing log: "cat *.ocs >> logfile".

    Use a symbolic link to maintain consistent IRAF parameters for our observations. On ctioa8 this is done with

    cd /ua80/mosaic rm uparm ln -s /ua80/mosaic/smsn/uparm.supermacho.2002 uparm

    Make a link to the SMSN postproc shell script file: rm /iraf/msgbus/bin/postproc ln -s /ua80/mosaic/smsn/sm.postproc /iraf/msgbus/bin/postproc

  9. In the ctioa0 VNC, select (Re)Start Arcon Session in desktop menu.
  10. Let Arcon boot up. This takes a few minutes. Answer yes to Synchronize parameters question in lower blue Arcon acquisition window. Then epar obspars, change comment to todays date and reset counter to 1.

    In the ctioa8:9 VNC, be sure that PostProc is enabled in the DCA GUI window (square should be yellow).

    Then back in ctioa0 VNC in blue Arcon acquistion window, type cd and then cd /ua80/mosaic/smsn/tonight to go to tonight's data directory. (Don't use the shortcut cd tonight, because then the data is saved by following links.) Take a test exposure (type "test" and take a type "zero"). Make sure of the following...

    - watch black status, make sure image reads out - watch ximtool, make sure bias looks OK - watch DCA window, make sure there are messages indicating image transfer

    Make sure that the PostProc option in the DCA GUI window is enabled, so that the images are automatically transfered to La Serena. (Note: this is updated from previous versions which said to disable this feature during calibrations. The net is now working well and zeros can be taken without getting behind in the data transfer).

  11. Now copy the appropriate essence ocs file into the night's
  12. directory. These files are usually called walpha.ocs and wbeta.ocs.

    You should edit the observing log file over the course of the night, noting seeing, clouds, malfunctions, bad images, etc.

    In particular, add in the night's essence observing file (the .ocs file) into the observing log, just under the skycalc data and before the sm ocs file.

    Note: timestamping log entries

    VIM (i.e. vi ): you can insert a time/date stamp by typing ":r read !date", or can set a function key to do this by:

    type "map " then hit the F2 key, for example, then type "read !date <cr><cr>" where the characters "<cr>" mean exactly that, namely 4 keystrokes that include less than and greater than symbols.

    Non-maschochists (i.e. emacs users) can hit 'Ctrl-C t' to get a time stamp.

  13. Ask mountaintop staff to turn off dome flat lights. Ask mountain
  14. staff to insert a DLT in the tape drive on ctioa8, and make sure that the 35Gb and compressed lights are on (at least temporarily... the compress light tends to blink three or four times and then shut off until tape starts writing).

    AFTERNOON CALIBRATIONS

    ...ZEROS

  15. Take a test image of type zero. Check to see that image looks clean and normal
  16. Upon getting permission from mountaintop staff, take a stack of 10 zeros. (Check PostProc is turned off so we don't get an rsync backlog.)
  17. BE SURE TO USE THE COMMAND "zero"!
    (Number of exposures to take: 10, Title of picture: Zeros aftnoon 050928.)

    ... DOME FLATS

  18. Ask the mountaintop staff to turn on the "color balance"
  19. dome lights, and to point the telescope at the white spot. Take an 8 sec test image in VR, should have around 8-12K counts.

  20. Do settemp, then
  21. Do "lpar doobs" and verify that the file has the settings given below.
  22. Take the following set of dome flats, 9 each, with the command doobs:

    Make sure that there are approximately 10,000 counts in the flat fields by looking at the values reported in the lower window of the DCA.

    TWILIGHT

    Verify that the mountaintop staff have the binders with guider finding charts in them, to get proper pointing.

    ALSO note in the observing log the sky conditions at sunset (cloudy, cirrus, photometric, other). Save an image from the webcam into /user/sm/environ2005/CTIO_webcam_050930_aftnoon.jpg.

  23. ZERO POINT
  24. As soon as the sky is dark enough, take a telescope zeropoint image to set up telescope coords.

    This can be a short exposure in twilight (as short as 0.1s if sky is very bright, usually ~1s). Usually just go to a bright star near zenith (the telescope operator will select one).

    There is an IRAF task called "zpt", which allows you to mark the centroid of the zeropoint star with a keystroke (for example, hit spacebar). Then hitting any key will print out the telescope offset needed to align telescope coordinates. It should look like "teloffset 16.7 40.5" (for example). Copy and paste this into the Arcon acquisition window to offset the telescope by this amount. Take another test image, confirm that the star is centered and then ask to have the telescope zeropointed. (note E is up N is left)

    If the display doesn't look good you can mscdisplay filename to get a clearer view.

  25. STANDARD STARS
  26. Ask telescope operator to slew to standard field near the current LST, take series of standards.

    In Oct, we're using MarkA. In early Nov, we use L113. In late Nov and Dec we use L92.

    BEST STETSON STD FIELDS

    Field IDRA(2000)DEC(2000)RA Size DEC SizeB StdV StdR StdI Std
    MarkA20 43 43.20-10 46 16.933.9 26.9723917749764
    L11321 41 34.20+00 43 47.063.358.1 441457122116
    SA9200 55 21.50+00 51 37.036.138.5 163205129201

    LANDOLT STANDARD FIELDS...

    SA11119:37:38.47+00:22:02.46
    SA11220:42:25.14+00:12:59.87
    SA11321:41:37.18+00:29:46.79
    SA11523:43:34.56+00:59:39.08
    SA9200:55:50.78+00:50:07.30
    SA9301:54:32.68+00:48:24.41
    SA9402:56:58.90+00:26:52.51
    SA9503:54:02.78+00:13:36.91

    When you get to the standard field, take VRI images at base position. doobs filters: B,V,R,I exptimes: 3,3,2,2 title: L113 (0,0) After this, offset in some pattern, take another doobs, offset again, take another doobs, etc until about 20 min before astronomical twilight.

    Offset sequences can be repeated offsets of 270,270 (diagonally stepping accross field of view), or -270,270, or other offsets.

  27. FOCUS
  28. NEW: Slew to first ESSENCE field. TO do this, just start up the mosocs script with tonight's SN ocs file. Answer "no" when asked "observe this field?" but the slew will happen anyway.

    Read the telescope temperature with settemp in the ARCON window.

    Take note of the temperature and suggested focus setting

    Take a focus frame: use the command "focus" in the Arcon window.

    **note that focus frame should be done in R, (VR for supermacho)**

    Exposure time (0.:) (10.): Title of picture (focus VR T=11.2 Z=1.87): focus R Z=1.94 Number of focus exposures (5): 9 Middle exposure (number 4) of sequence to have Focus value (14658): (insert value from settemp above) Focus increment (75): 75 Filter in wheel one (R):R

    Use mscfocus to evaluate the image. Use "m" to mark (always the top!) star in a sequence, and do this for 10-20 stars well distributed across the image. You can also use the "g" key if you want to see the plot of FWHMs at each star. (Use mscexam and type "r" to get profiles if the seeing is too bad for mscfocus to work.)

    Once enough have been clicked, type "q" to quit the ximtool window.

    This will bring up a graphical window. You can delete outlying points by putting the crosshairs on the bad point and typing "d".

    Once you're happy with the points, quit the graphical window with "q".

    Cut and paste the suggested focus with the mouse.

    Do "epar instrpars" and copy the telescope temperature from the temperature parameter just above basefoc into the reftemp parameter two lines below. Enter the best focus value into the spot between the two temperatures.

    As the night progresses, keep an eye on the PSF in the images. If it becomes non-axisymmetric, then you need to adjust the focus. We seldom need to refocus during a night, unless the temperature makes a dramatic change.

    Note if stars look square check whether the images are being binned. Use "show stdimage", if it returns imt2048 type reset stdimage=imt4400.

  29. Do "epar mosocs" in the ARCON control window, and point it at the night's "essence.ocs" file. No other changes are necessary.
  30. Verify that the correct file is listed in mosocs with "lpar mosocs" then start the observing sequence with "mosocs". This will bring up a slew request that the telescope operator must enable, and they will then line up the guide star according to the guider finder charts they have on the mountain. If for some reason the guider finder hardcopy is not there, there are also images (for both W and SuperMacho) on the wproject web pages.
  31. WHILE OBSERVING

  32. Monitor the image quality using seeing in the IRAF window on ctioa8. Be sure you don't use the blue Arcon window for this! This utility called "seeing" is in the "smsnscripts" package (so if "seeing" doesn't work, type "smsnscripts" to load the package). Run seeing, and then hit either 'a' or 'r' on stars around the field ('a' just selects the star and adds the fwhm to the average, 'r' does the same but ALSO PLOTS the radial profile of the star for you to see). Seeing converts from pixels into arcseconds. Copy the output of seeing into the observing log.

    Move REALLY bad images to imagename.fits_bad - this will prevent the pipeline from picking them up.

    Also monitor that the headers are OK various times each night. To do this simply do a
    hsel w*.fits[1] $I,RA,airmass,exptime
    and see if any of the images are missing key information. If they are, follow the instructions given below for fixing headers.

  33. As we approach the end of the first half of the night, there is a
  34. possibility of losing the long integrations on the deep fields, if we get behind. There is a file ***PUT FILENAME*** that shows the desired pacing of observations. It's fine if we get *ahead*, since then we just integrate longer. If, however, we slip behind then the suggested procedure is to drop the I band integrations for some of the shallow fields. The file gives desired time until SM handover... so judge accordingly based on middle of night hand-off time.

    ----------------------------------------------

    for SuperMacho hand-off at night center according to Skycalc:

  35. Switch to the VR filter. (Click on button in the MCCD configuration schematic.)
  36. Run mosocs and change the file from w***.ocs to sm.ocs
  37. This will start up the SuperMacho field sequence. START OBSERVING IMMEDIATELY.
  38. If focus looks bad THEN (and only then) do a focus frame...
  39. Run into twilight until the guider becomes unworkable.
  40. ##22. Try for sky flats, in the reverse order used in the evening twilight.

------------------------------------------

End of night procedure:

  1. Take final cal's, update log file, and write DLT:
  2. Ask telescope operator to close dome and turn off lights.

    Grab - red and blue movies (every 5 min) from TASCA - morning webcam image (from Tololo webcam) - environmental page screen shot (with xv) Save these in /user/sm/environ2005 (on ctiokq) as animation2_blue_05****.gif and _red_, CTIO_webcam_05****_morn.jpg and environ_05****.gif

    take 5 more zeros

    Update summary log file smsn05obslog.sum.

    Fill out night report at http://www.ctio.noao.edu/new/Tools/Forms/EON/Form.php?telescope=Blanco%204-m

    ls *.fits | fields RA DEC FILTER AIRMASS EXPTIME >> (obslog for the night)

    end Arcon session, save and close observing log.

    run ~mosaic/smsn/bin/endsm on ctioa8

    Restart Arcon (check it loads properly for the next observer)

    Back-up the data on tape. In IRAF window on ctioa8:9 VNC:
    make sure you're in mscred
    rew mtc
    smtape (will prompt for a date string)
    rew mtc

------------------------------------------

In case of trouble:

Fixing headers

Before fixing any headers, save a copy of the original image as imagename.fits_orig.

Removing bad images

Bad images just cause problems for the pipeline, so we should move them out of the main data directory.

As mosaic@ctioa8 in the observing directory (sm05****)
mv image ../sm05crap/

As sm@ctiokc type cdmscraw to move to /ukc3/sm/data/mscv6.0/SMSN/mscraw then cd to today's date.
mv image ../sm05crap/

Refocus mid-run.

About the telescope

The Blanco 4m has a 36 arcmin field of view, with 0.27" pixels. The camera consists of 8 chips, each 2048x4096 pixels. Each chip has two amplifiers.
The numbering pattern for the chips looks as follows:
east
5678
1234
north
The numbering pattern for the amplifiers looks as follows:
9,1011,1213,1415,16
1 , 23 , 45 , 67 , 8

Useful Tidbits

Testing headers: in IRAF type
hsel w*fits[4]
fields to be extracted ($I, RA, DEC, AIRMASS)
boolean (yes)

Seeing: from mscred in iraf run seeing. Select good looking stars with the mouse and type "r". Repeat several times. Type "q" to quit. Returns seeing in arcseconds.

Opening new iraf: Rt click, open xgterm. cd, cl, cd tonight, mscred.

Checking airmass: skycalc, "t" for tololo, "r 00 00 00 d 00 00 00 h" (with the right coordinates) then type an object name gives you the hourly airmass for the object.

Checking rsync: ps aux | grep rsync

Computers, users and useful files
Log onto ctiokq as sm to start up vncviewers. (Note vncviewers have different password.)
Log into ctioa8 as mosaic to observe.
"cd smsn" (to our home observing directory)
"cd sm05****" (to get to that night's directory)
/ua85/mosaic/smsn/ = where to find the observations.
/ua85/mosaic/smsn/sm05****/sm.ocs and walpha.ocs or wbeta.ocs = the night's observing files.
/ua85/mosaic/smsn/smsn05obslog.sum = summary log of each night's observations.
/ua85/mosaic/smsn/sm05crap/ = directory in which to move bad images that should not go to the pipeline.
/ukc3/sm/data/mscv6.0/SMSN/mscraw (accessed by typing cdmscraw, as user sm on ctiokc) = second location of images (second place bad images need to be dealt with). Printer In Alfredo's office, npv3.

Contacts

Chirs Smith: x270(Hm), x214(Office), 9-2005-09-1427391(Cell)
Armin Rest: x207(Office)