This is a collection of comments on the Keck operation from the European astronomers I know to have used the telescope. I collected them because of their potential relevance to VLT. I have not troubled to edit or order them. Bob Fosbury, 21 June 1996 Dear Sperello, Marijn and Rolf, Given the imminent arrival of the VLT, I thought it would be useful to jot down a few comments about the Keck operation to see what we can learn. I was only there for two nights using one instrument and so it would be a great help to broaden this experience with your own. If you are happy to add your comments and thoughts, I will make this note available to Riccardo, Alvio and Jim. Keck is clearly a very different operation from what is envisaged for VLT - we all take this for granted because of the fundamentally different user community but I wonder if we might go too far and make some mistakes on the way. I give a few of my comments below just to start this exercise off. Do you know of any other European Keck observers who could be contacted? Regards, Bob... May 1996, 2 nights with LRIS + polarimeter ------------------------------------------ 1) The telescope operation: The night assistant (OA) we had (Terry) worked also as part of the day crew (electronics) and was very competent in telescope operation. He knew when to start the various sequences, telescope focussing etc. in order to be ready at the start of the night without time wasting. The philosopy of using day crew as OAs is being pushed by Peter Gillingham from his AAT experience but I gather that the idea is not universally accepted within the organisation. I was impressed by the low overhead associated with telescope adjustment and control. The focus is done infrequently (once only during the run in our case) by tilting the segments by fixed amounts inside and outside of expected focus and then using measurements of the hexagonal image array to predict focus position. The servo loop is then left to get on with it while the astronomers observe. There were no telescope or instrument problems during our 2-night run and so the on-target efficiency was high. Of special note is the excellence of the TV acquisition and guiding system and the ease and rapidity of selecting and locking a guide star. CARA have just hired Bob Goodrich as (the first) instrument support scientist. He is very competent - especially with LRIS and the polarimeter (which he helped design). It is not yet clear to him whether he gets any research time. I spent one night at the summit and one night in Waimea doing remote control. In addition to a high speed data link - one actually logs-on remotely to the summit computers - there is a big-screen PictureTel videoconference link with panning/zooming cameras at each end. The result is that you actually feel you are in the control room but with a sea-level oxygen supply. Papers and charts can be held in front of the TV camera and do not always need to be faxed. Judy Cohen is just about to do experiments remote from California using a collaboration with NASA with one of their high capacity communication satellites. Although I am not in favour of remote control as an end in itself, there is no doubt that if you have high enough bandwidth it works well. There is a real problem assessing the weather - even if you are at the summit. Detecting cirrus during dark time is hard but it is really necessary. Modern control rooms are well-lit and going outside, especially at high altitude, does not really tell you what you want to know. It is essential that we have a sensitive and reliable method of determining the relevant weather parameters AND have a way of recording these with the data. 2) Instrument operation: The Keck telescopes have few instruments so far: HIRES, LRIS, NIC and LWS. Only HIRES is at Nasmyth - so there are plenty of spare foci. LRIS suffers from grating flexure but I don't know about the other Cass instruments. LRIS has to be dismounted to install the polarimeter module. This is a pain but it seemed to work without problems. LRIS is an excellent polarimeter and it has made faint object polarimetry into a regular precision science - a far cry from the 4m struggles we had with photons as scarce as rocking-horse shit (pardon the lapse into Australian). I append a FITS header below to give an idea what goes into the data. This is from an acquisition camera frame, not an LRIS data file. SIMPLE = T / file does conform to FITS standard BITPIX = 16 NAXIS = 2 / number of data axes NAXIS1 = 384 / length of data axis 1 NAXIS2 = 288 / length of data axis 2 DATE = '19/05/96' / FITS file creation date (dd/mm/yy) WINDOW = '0,1,1,384,288' / chip #, xstart, ystart, xsize, ysize BINNING = '1,1 ' / xbin, ybin ELAPTIME= 1.000000E+01 / total elapsed time (s) TTIME = 1.000000E+01 / total exposed time (s) UTC = '10:59:23.00' / coordinated universal time (h) LST = '16:27:03.35' / local apparent sidereal time (h) SIMULATE= 'false ' / simulating dcs SIMUTC = '04:16:08.56' / simulating utc (h) DCSLOCK = 'false ' / drive and control locked DCSHOST = 'cara ' / drive and control host DCSPORT = '5001 ' / drive and control port DCSVERS = '9.22 ' / drive and control version TELESCOP= 'Keck I ' / telescope name FOCALSTN= 'cass (cassegrain)' / focal station RA = 2.029734293246E+02 / telescope right ascension (13:31:53.62 h) DEC = -1.143459330397E+00 / telescope declination (-01:08:36.4 deg) RADECSYS= 'FK4 ' / ra dec coordinate system EQUINOX = 1.950000000000E+03 / telescope equinox AIRMASS = 1.475644847195E+00 / air mass HA = 4.318861278767E+01 / telescope hour angle (02:52:45.26 h) AZ = 2.486285445002E+02 / telescope azimuth (deg) EL = 4.266182297654E+01 / telescope elevation (deg) CALOCAL = 8.888888888889E-03 / collimation azimuth local (32.0 arcsec) CELOCAL = -9.722222222222E-03 / collimation elevation local (-35.0 arcsec) DATE-OBS= '19/05/96' / universal date of observation MJD-OBS = 5.022245790509E+04 / modified julian date of observation DCSSTAT = 'running ' / drive and control status AXESTAT = 'tracking' / axes control status DOMESTAT= 'tracking' / dome status SECFOCUS= 1.180570000000E-02 / secondary mirror focus raw (11.81 mm) TELFOCUS= 1.242151909232E-02 / telescope focus compensated (12.42 mm) TUBETEMP= 2.280092672094E+00 / tube temperature (2.28 deg C) SECOFFT = -4.104166809769E-04 / secondary temp offset (-0.41 mm) SECOFFE = -2.167678153066E-04 / secondary el offset (-0.22 mm) SECTEMP = T / secondary auto temp comp (enabled) SECEL = T / secondary auto el comp (enabled) ROTMODE = 'position angle' / rotator tracking mode ROTPOSN = -3.300054327325E+01 / rotator user angle (deg) ROTPPOSN= -9.429803471512E+01 / rotator physical angle (deg) PARANG = -6.129749144186E+01 / telescope parallactic angle (deg) TARGNAME= '1331-01 ' / target name TARGRA = 2.029737500000E+02 / target right ascension (13:31:53.70 h) TARGDEC = -1.141388888889E+00 / target declination (-01:08:29.0 deg) TARGEPOC= 1.950000000000E+03 / target epoch TARGEQUI= 1.950000000000E+03 / target equinox TARGFRAM= 'FK4 ' / target frame PONAME1 = 'POL ' / pointing origin name 1 POSTAT1 = 'nam,def,sel' / pointing origin status 1 POXPOS1 = 4.800000000000E-04 / pointing origin xposition 1 (0.48 mm) POYPOS1 = -2.753500000000E-01 / pointing origin yposition 1 (-275.35 mm) PONAME2 = 'REFM260 ' / pointing origin name 2 POSTAT2 = 'nam,def ' / pointing origin status 2 POXPOS2 = -2.435000000000E-01 / pointing origin xposition 2 (-243.50 mm) POYPOS2 = -2.480000000000E-01 / pointing origin yposition 2 (-248.00 mm) PONAME3 = 'LRIS ' / pointing origin name 3 POSTAT3 = 'nam,def ' / pointing origin status 3 POXPOS3 = 1.320000000000E-02 / pointing origin xposition 3 (13.20 mm) POYPOS3 = -3.070000000000E-01 / pointing origin yposition 3 (-307.00 mm) OBJECT = '1331-01 ' DISK = 'FITS ' DATATYPE= 'INTEGER*2 ' BZERO = 3.0859999389630E+03 BSCALE = 7.9653309732353E-02 END 3) On-line analysis, calibration and archiving: Not much to say here - it is up to the observer. This is a very trimmed-down operation and there is no attempt to make data available for posterity or to anyone but the observer. There is no on-line tape recording (at least for LRIS), the data are simply written to disk for later transfer to exabyte or ftp. The OA does, however, keep rather comprehensive statistical data about the night with a detailed breakdown of the use of time, efficiency on target, time spent on calibrations, technical down-time etc. This is all written to a computer file at the end of the night.I think that Jim Crocker has a copy of the statistics given by Peter Gillingham at the ESO/SPIE conference. I looked at these in Peter's office and was struck by the large fraction of time lost to weather - especially in the winter. Hawaii seems to be going through a bad phase at the moment. 4) Conclusions: The idea is to react fast, be there first, skim the cream and publish. Although the total data flow from the telescopes may not be as high as it could be, the rate per eligible astronomer is very large. There is no mechanism in place to make a global matching of observations to conditions. Date: Tue, 4 Jun 1996 12:15:43 +0200 From: Sperello di Serego Alighieri To: rfosbury@eso.org Subject: Re: Keck operation Dear Bob, I think your idea to see what we can learn from our Keck experience for the VLT operation is fine, particularly considering that the 3 persons, to which you would like to make our comments available, have very little, if any, personal experience with ground based telescopes. Therefore I answer gladly to your call, although what I can add to what you said is limited, given that I also just had 2 nights, we used the same instrument and I have already told you about my experience. You might also contact Andrea Cimatti and Alec Boksenberg (you said he was there). April 1996, 2 nights with LRIS + polarimeter ------------------------------------------ 1) The telescope operation: I was also impressed by the competence of the OA. We had some pointing problems with the telescope: a few times per night the telescope went to a wrong position. It was enough to start the pointing sequence again to get around the problem, which was known, but not its cause. By the time we had realized that the pointing was wrong (which took a bit longer from Waimea) and we had pointed again we had lost around 5-10 minutes each time. Except for that, no problem. I also thought that the remote control from Waimea was nearly as efficient as from the summit, although the video link was not yet installed. I also felt the problem with assessing the weather and tried to convince them to keep the lights down in the control room, so that going outside it would not take long to adjust to the dark. I would be very much in favour with a real time and wide angle measurement of the relevant weather parameters (any idea how to realize that?). 2) Instrument operation: No problems here, except that I found that the link between the instrument work station and the telescope control was very poor. For example we had to insert the name of the object in the instrument workstation, although it was already in the list given to the telescope for pointing. We sometime forgot which resulted into a wrong header. Furthermore it is fine that the OA keeps a log of the on target time, but I found unbelievable that this log is compiled at the end of the night, when everybody is tired, by asking the astronomer to read times off his hand written log sheets. 3) On-line analysis, calibration and archiving: Indeed everything is left to the observers. Marshall and Pat had very good VISTA procedures which gave polarization images (i.e. contour maps with polarization vectors overlayed on) very quickly, which helped a lot in optimizing what we would do in the remaining time or even in the following night. This is clearly thanks to our presence and would be very difficult to duplicate with service observing. I leave conclusions to you and would be happy to help more if I can. Cheers, Sperello Sperello di Serego Alighieri Osservatorio Astrofisico di Arcetri Largo E. Fermi 5 I-50125 Firenze Italy Tel.: +39-55-2752-311 Fax: +39-55-220039 E-mail: sperello@arcetri.astro.it (Internet) Date: Fri, 7 Jun 1996 09:24:35 +0200 (MET DST) From: Rolf Peter Kudritzki To: Bob Fosbury Subject: Re: Keck operation Mime-Version: 1.0 Dear Bob, I am just back from Potsdam where I spent the beginning of the week. I just read the beginning of your message. You know that I had several nights at Keck last August. My conclusions may be exactly the same as yours. Best regards Rolf +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ !Rolf Peter Kudritzki ! !Uni.-Sternwarte Muenchen Internet: KUDRITZKI@USM.Uni-Muenchen.DE! !Scheinerstr. 1 Voice: +49-89-92209422 ! !81679 Muenchen FRG FAX: +49-89-92209427 ! +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ Date: Wed, 5 Jun 1996 09:28:03 -0700 (PDT) From: cimatti@igpp.llnl.gov (A. Cimatti) To: rfosbury@eso.org Subject: KECK Hi Bob, it's a good idea to write down our experiences ! However I found your experience very similar to mine, so I will try to add more (I have a run on June 16th), can you wait till after my run ? All the best Andrea To: cimatti@igpp.llnl.gov cc: rfosbury@ns3.hq.eso.org Subject: keck ops Date: Wed, 05 Jun 1996 22:17:01 +0200 From: Robert Fosbury Dear Andrea, Thanks for the reply, yes I can wait until after your next run - you can write comments on the fly.... What are you doing this time? I see that marshall told ou what happened during our run - sorry I was travelling when your message arrived. I hope the weather is good. To see my photos, try http://ecf.hq.eso.org/~rfosbury/keck/ I look forward to your news... Regards, Bob... Date: Wed, 5 Jun 1996 10:22:55 +0200 From: Marijn Franx To: rfosbury@eso.org Cc: franx@cosmos.astro.rug.nl Hi Bob, it's may be not such a bad idea to describe some of our experiences - I will try to do so over the coming week and a half. My experience is dramatically different - things worked out well for the first time in our last run (the first in a long row) ; but it was interesting to note that many of the improvements were user driven, and would not have occurred if observations had been done in "service mode". more later, Marijn Date: Thu, 20 Jun 1996 14:09:07 -0700 (PDT) From: cimatti@igpp.llnl.gov (A. Cimatti) To: rfosbury@eso.org Subject: Keck Dear Bob, I send my short comments about my KeckI experience. KeckII will start observing regularly in October 96, and "normal" observing time has been already allocated. I would say that probably one mistake of ESO was not to adopt a very competitive policy : I mean that it would have been a good idea to decide to start with 1 VLT as soon as possible, equipped with the most simple instrument possible (e.g. an imaging camera and/or a spectrograph), in order to be in a real competition with the Keck(s). My concern is that when the 1st VLT will start effectively the observations, KeckI and KeckII will be operating since a long and considerable time respectively...., skimming the cream ... as you said... But this is also pretty obvious.... All the best Andrea ---------------- KECK EXPERIENCE Andrea Cimatti ---------------- My experience at KeckI is based on three different runs from 1995 to 1996, for a total of 4 nights. I used the low resolution imaging-spectrograph (LRIS) in polarimetric mode in order to observe faint (V~21-23) radio galaxies at high redshift (Program in collaboration with Wil van Breugel, Hy Spinrad, Ski Antonucci and Arjun Dey). Please note that KeckI can presently work only with the red arm (lambda>4000 AA), but they told me that very soon also the blue arm will be available. We have never had a single technical failure or instrument problem, and we had 3 perfect nights with sub-arcsec seeing (1995) and 1 partly cloudy night (1996). The "efficiency" of our observing runs, namely the time spent actually on the targets, has been extremely high. I can say that I was impressed by the high efficiency and competence of the night and technical assistants, who always allowed us to start the observations with no delays, and always helped us with courtesy and efficiency. I always took part to the observations going physically to the top of the mountain, but remote control is possible from Waimea too, or part of the observing team can stay in Waimea to follow the observations thanks to a very efficient "video-conference" communication system. For more specific and detailed comments, my experience has been very similar to that of Bob Fosbury, so please refer to his extended report. Andrea