PartnerObserving Log for 2007 Mar 19/20

Observers: Rick Pogge, Eichii Egami
Support Scientist: none
TO: Aaron Ceranski

Mostly clear during the afternoon, with high clouds around us. Satellite looking much more promising for later this afternoon and tonight.

All times below are UTC 2007 March 20

0045 - Afternoon Calibrations: 25bias and dark900. Lot of high, thin stuff around obvious at sunset. conditions are non-photometric

0138 - Sunset, opening the enclosure and getting sky flats at 6h/+27.
  • SkyFlatU_5: 1x - 014956 - 015350, levels 36K-13K
  • SkyFlatG_5: 1x - 015743 - 020045, levesl 39K-18K
  • SkyFlatR_5: 10x - 020417 - 020750, levels 25K-13K

Still lots of thin high stuff more apparent at sunset, and some stuff towards the west/southwest that is dissipating.

0209 - Moving to field and initializing the primary mirror using Rfastextray. Made inefficient by the archive system being very slow to transfer frames at times, and sometimes not doing it at all until I take another image. We've IssueTrak'd this problem repeatedly. The mirror is being a wee slow to converge tonight, mostly big swings in Z11 and Z7 that finally settled down.

0250 - GRB/grb070311 - last observation to finish this one off. 5x200s dithers, twice.
  • Image1: 025055 - more archive stall, didn't see this image for a while, Satellite streak across Chips 2 & 3, 0.8" FWHM
  • Image2: 025452 - GUI stalled, revived by changing pages, no sign of the image in the Repository for a while, <0.8" and round
  • Image3: 025851 - images a little lopsided, fwhm~0.8"
  • Image4: 030248 - 0.75" FWHM
  • Image5: 030702 - 0.7", reasonably round
  • Image6: 031120 - 0.8" fwhm, fairly round
  • Image7: 031517 - 0.8+" fwhm, pretty round (seeing getting fuzzier than before) - we didn't update mirror between OBs.
  • Image8: 031918 - 0.9" fwhm
  • Image9: 032315 - 0.9-1" fwhm, slipping out steadily
  • Image10: 032725 - 1" fwhm.
Tweaking up mirror and one more shot (since we started early)
  • Image11: 033513 - 0.9", fairly round.
  • Image12: 033915 - ~1"(-), satellite streak across chips 1, 2, and 4, missed GRB
  • Image13: 034319 - ~1"
  • Image14: 034724 - guiding jumped, images badly lopsided, exposure ruined
  • Image15: 035134 - seeing ~1.1" we're done.

0356 - Kochanek/M81V - fits nicely into time until the next target is above its airmass limit. Tweaking primary (offset field Bfastextra), converged quickly
  • Image1: 040653 - 1" seeing
  • Image2: 040830 - 1.1" seeing
  • Image3: 041008 - 1" seeing
  • Image4: 041146 - 1" seeing, stars nice and round

0413 - COSMOS/Ublock1 - tweaking primary with Bfastextra, will do this between each OB in this set (a-d). Will try again to see if we can do better than on the previous night when variable seeing was playing hell with getting guide stars. Starting with OB SDTUspec_1a.
  • Image1: 042609 - did not find a guide star, though there are stars on the technical chips. This one is going open-loop. FWHM 4.5pix and slightly elongated
  • Image2: 043310 - found guide star, appears to be guiding stably, lost lock half way through exposure for no apparent reason, then recovered. Maybe 30-40s open loop, dropped out again a few times in, again for no apparent reason as the star was stable and a sharp image. The tolerances for "good guide star" must be set way too tight, there's no way it should not have had an easy time guiding on this star. Images are 0.9" FWHM and reasonably round.
  • Image3: 044015 - 32s gs exposures, lost lock about 200s/360s briefly, then again at 300s - Images are 0.9" FWHM and very round.
  • Image4: 044707 - 16s gs exposures, kept losing lock, running mostly open loop like Image1 above. Images are elongated, 1"FWHM
  • Image5: 045415 - 16s gs exposures, got a nice guide star, help lock whole exposure. Much better, 0.84" FWHM and very round.

Note: 1st and 2nd images didn't show up in the repository until nearly 5 minutes into the 3rd image in the sequence. 3rd was out quicker, as was the 5th (4th was slow, but not nearly as slow as 1&2).

0501 - Tweaking primary with Bfastextra, then running SDTUspec_1b. Only a small update required.
  • Image1: 050506 - appears to have found a guide star, then lost lock and never recovered, ran open loop for most of the exposure. Tech chip images show fine guide stars. Images is 0.9" FWHM but a little elongated
  • Image2: 051204 - got a gs and started. Occasionally lost lock briefly. 0.8" FWHM and very round.
  • Image3: 051905 - got a gs in a 32s loop, keeping tight lock for entire integration, very round 0.8" (or less) images.
  • Image4: 052557 - got a gs in a 32s loop, held for entire integration, 0.8" and very round
  • Image5: 053301 - don't know gs history because of GUI lockup just after start of exposure, was guiding when GUI came back (but guiding was in/out all the time). Images are 0.72" and very round.

Note: GUI locked up during Image 5, could not even change pages, could not ping the lbccontrol computer or login, then it came back on its own. This must be telling someone something.

0540 - Tweaking primary with Bfastextra before proceeding on with SDTUspec_1c.
  • Image1: 054444 - started out OK, then lost gs lock very early and ran open loop - images elongated about 2:1 (effectively doubled), exposure ruined, but a good engineering image for looking at LBT open-loop performance at this position in 360s.
  • Image2: 055141 - got gs in 8s loop, held lock entire integration - 0.8" FWHM and very round
  • Image3: 055845 - got gs in 16s loop, lost lock after first minute & never recovered - exposure ruined , effectively doubled.
  • Image4: 060528 - ramped up to 32s but never found a gs, running open loop, given previous bad images, I'm aborting this exposure and the OB before it makes more of a mess.
OB aborted during 4th image, something about this particular region and the present gs algorithm make it unobservable.

Last chance...

0608 - Tweaking primary with Bfastextra, then trying to execute SDTUspec_1d.
  • Image1: 061320 - got a nice 8s gs, then lost it - why can't it ramp up the exptime when that happens? - never recovered, running open loop. Letting it go for its possible diagnostic utility, otherwise I'm aborting the OB after this integration. There were more than sufficient stars visible in the technical chips, and it had a nice one at the outset.

NOTE it is clear that the autoguiding algorithm is not up to the job in high-latitude fields at U. We do not wish to waste more subarcsecond seeing on this.

0620 - Setting up for g imaging, running Gfastextra, then running COSMOS/gSloan_4. CSQ subsystem went down around 0622 when I tried running lbcfpia, Aaron bringing it back online w/o problems.
  • Image1: started out with a gs in 8s loop then dropped it 46s in, tech chips show lots of guide stars. - aborted exposure

This is very frustrating, we're going to observe some other project fields now

0638 - Zaritsky/NGC3184, tweaking primary with Vfastextra, then executing NGC_3184_V.ob (10x164s). Seeing is probably 0.8 judging from the tech chip images previously
  • Image01: 064811 - <0.8" seeing, reasonably round
  • Image02: 065153 - seeing 0.7" FWHM (or better?) on the galaxy, but is variable across Chip 2 (top-to-bottom stuff again)
  • Image03: 065549 - 0.7" or better in places.
  • Image04: 065928 - 0.6" seeing and very round
  • Image05: 070326 - 0.6" on galaxy and reasonably round (wind is 10m/s with gusts)
  • Image06: 070708 - 0.6" (or better?) on galaxy.
  • Image07: 071052 - 0.5" seeing - wow, we weren't expecting that...
  • Image08: 071450 - 0.6" seeing and round
  • Image09: 071842 - 0.6" (or better?) on galaxy and round
  • Image10: 072223 - 0.6" or better on galaxy.

Another GUI lockup around 072325, prevented me from getting on top of the primary tweak for the next OB... Cost us a couple of minutes.

0728 - Tweaking primary with Bfastextra, then executing NGC_3184_U.ob (10x164s).
  • Image01: 073459 - 1" FWHM and elongated, guiding wasn't very solid (16s loop).
  • Image02: 073843 - lost 8s gs lock, taking most of the 164s open loop, images are all doubled (separation ~3") exposure ruined
  • Image03: 074229 - got 8s gs lock (always makes me nervous...) - maintained lock for entire integration, -0.8" FWHM and round
  • Image04: 074603 - got 16s loop gs, probably held (GUI lockup in last 60s), ~0.8" FWHM and round
  • Image05: 074946 - got 8s loop gs, holding lock, <0.8" FWHM and a little ratty in roundness
  • Image06: 075328 - got 8s loop gs, holding lock, <0.8" FWHM and very round
  • Image07: 075709 - got 8s loop gs, holding..., still 3.3pix FWHM (<0.8") and slightly out-of-round.
  • Image08: 080102 - got 8s loop gs, holding..., ~0.8" FWHM and a little squashed roughly along the x-axis.
  • Image09: 080452 - got 8s loop gs, holding..., ~0.8" FWHM and reasonably round (more so than previous 2 images)
  • Image10: 080832 - got 8s loop gs, holding..., 3.2pix FWHM (0.72") and reasonably round.

Image transfer to the Repository was very constipated after image 2. At one point we saw no data-transfer action for 7 minutes, despite taking guider images every 8-16sec! A couple of GUI locks cleared by toggling between pages.

0812 - Weaklens/MS1358 - hoping the seeing will also be good enough there to execute it in g/r (must be <0.7"). Tweaking primary with Gfastextra, then running OB MS1358_1.ob and if the seeing is within the program's limits, we'll proceed...
  • Image1: g 300s - 082114, got 8s loop gs, FWHM=0.8",
  • Image2: g 300s - 082659, got 8s loop gs, FWHM=0.8" and a little elongated (guider not locking well?)
  • Image3: g 300s - 083241, got 8s loop gs, FWHM=0.9" and images are out of round on the diagonal (roughly along a NE/SW direction)
  • Image4: g 300s - 083818, got 8s loop gs, FHWM=0.9" and out of round on the diagonal (in same way, lopsided to the NE)
  • Image5: r 300s - 084400, got 8s loop gs, FWHM=0.9", and out of round same as the others.

0850 - Tweaking primary with Gfastextra - a bit of Y astigmatism that might be consistent with the image distortion, iterating, Took a 60s snap, seeing much better now, as is roundness, but still not 100%. Wind 11.2m/s from 140deg, current azimuth 6deg, proceeding...
  • Image1: g 300s - 090545 - all 3.8pix FWHM (~0.9") and still some elongation (???) aborting OB and abandoning observation as outside program spec

Seeing has deteriorated some, we can no longer do this project. Instead executing some more poor-seeing-tolerant programs

0920 - Dwarf/fuobj8_g, a completion program. Tweaking primary with Gfastextra, then running OB fuobj8_g_2 to complete earlier observations and close out this particular OB.
  • Image1: 092956 - 8s loop gs, satellite streak across chips 1-3, images are elongated and kind of bent. ~1.3" FWHM, elongation is SE/NW
  • Image2: 093547 - 8s loop gs, (GUI lock, usual trick worked), another trail, similar trajectory! Pair of satellites? 1.3" FWHM, similar to first image.
  • Image3: 094135 - 8s loop gs, (another GUI lockup), no streak this time, similar to the others. Images are more circular in chip 3, but not much more
  • Image4: 094723 - 8s loop gs, same deal as the others. Also ~1.3" seeing.

Images not so good, seeing has definitely degraded, and now image quality (regardless of FWHM) is not so good, but nothing obviously out of whack in the Zernikes that lbcfpia is returning.

1000 - Going to Quasar/Point4 (blew the pupil last time, we know more now), setting scope with Gfastextra. Had to increase exptime 2x to get pupils, then running OB sdt_fan1_gr_4:
  • Dither1/Image1: r - 101352 - nice and round (better), seeing ~1.2"
  • Dither1/Image2: g - 101632 - 1" FWHM, nice and round.
  • Dither2/Image1: r - 101930 - 1.1+", reasonably round
  • Dither2/Image2: g - 102211 - <1", nice and round
  • Dither3/Image1: r - 102454 - 0.9" and very round (guider is behaving)
  • Dither3/Image2: g - 102734 - <1" and very round
  • Dither4/Image1: r - 103033 - <1" and very round
  • Dither4/Image2: g - 103312 - 0.8" and very round (last one was best - seeing is settling down again)

I realized only after the fact that I had accidentally left the 2x scaling for the Gfastextra set, and so the exptimes of the above were all doubled. Oops. Back in January when I was here, loading a new OB would reset that, this time it does not (along with all the broken check boxes). I hope this gets fixed, too.

1035 - Tweaking primary with Bfastextra, then executing sdt_fan1_U_4 -- using Gfastextra, no pupils in Bfastextra...
  • Image1: 105038 - 16s loop gs, 16s failed, aborted

1053 - Doing Quasar/Point6 (5 done previous night), gr only since U is pretty much guaranteed to have guider failure with few stars. Humidity is coming up a little, now about 50% - tweaking mirror with Gfastextra, then doing OB sdt_fan1_gr_6 Couple of screw-ups (late-nite typing) delayed things, as did the problem of data getting stuck in transfer to the Repository every other images or so. Took a while to converge

  • Dither1/Image1: r - 111022 - ~1" seeing and nicely round
  • Dither1/Image2: g - 111159 - 0.8" seeing - nice
  • Dither2/Image1: r - 111351 - ~1" seeing, kind of ratty
  • Dither2/Image2: g - 111528 - 1.1" (5pix), still under 1.2
  • Dither3/Image1: r - 111710 - 1.2" and a little ratty
  • Dither3/Image2: g - 111847 - 0.9" and round
  • Dither4/Image1: r - 112034 - 1.1" and a little ratty
  • Dither4/Image2: g - 112213 - 0.8" and round

Repository transfers stalled again around UTC11:13, didn't resume until 1120, after the arrival of image 111847 , which sort-of unstuck things, though they dribbled in more or less at random.

1129 - Last target, Milne/SN2005cf, since it is late and this executes quickly. Adjusting primary using Vfastextra, executing OB SN2005cf
  • Image1: V - 113345 - 1.1" (or so) and reasonably round
  • Image2: V - 113929 - 1.4" and still round (getting worse)
  • Image3: B - 114514 - seeing ~2", guide star was a bit swimmy
  • Image4: B - 115052 - seeing ~2" - kind of looks like the telescope is going out of focus, but also the guide star is swimmy, so may just be really bad seeing.
  • Image5: B - 115629 - seeing ~2"
  • Image6: B - 120206 -
*sequence aborted when it became clear it was starting 8x300s in r, which will never survive the inevitable shift in the primary, the deteriorating seeing, or remaining time. Why was this OB not broken into components? The queue tool indicated this was a 0.3h project. That's not what this OB is. Please check.

1211 - driving the telescope to 17h/+27 for twilight flats.
  • SkyFlatV_5: 5x, 124841-125146, levels 15K-32K
  • SkyFlatVrot180_5: 1x, 125416 - 125653, levels 15K-25K

Night Summary

Gooding seeing and not-so-good seeing (sometimes downright awful). Transparency was iffy because of cirrus blowing through and dissipating, so we declared it non-photometric based on thin high-cloud visible at sunset, ...

Programs Done:
  • GRBs/GRB07311 - deep follow-up to get a limit on its brightness and close the loop. Seeing started out 0.8", degraded to 1.1". Done, never observe again

  • Kochanek/M81V - a good quick queue filler in 1" seeing

  • COSMOS/Ublock1 - Did our best, but the guider kept losing lock at bad times. When it worked it was pretty good, around 0.8" seeing, but it finally got to be too difficult to sustain and we opted to go after more favorable targets. Also tried gSloan_4, but it failed to get guide stars, too. High-latitude fields are hard with this guider system in its present configuration. Copious notes above.

  • Zaritsky/NGC3184 - Dennis et al. win the seeing lottery. We figured based on previous data that it was around 0.8-0.9" so we went to one of this program's high-priority fields. As we settled in on the OB cycle, the seeing got great, at times we think we were scratching against 0.5" at V. In the U sequence the first couple were problematic, then it settled down, though never as good as V it stayed around 0.8" or better at times. Overall a very nice data set for this team to look at - a very pretty face-on, late-type spiral. A good example of what LBT+LBC can do when everything works well (tonight's pretty-picture winner).

  • Weaklens/MS1358 - inspired by the great seeing, we were below the threshold for weak lensing, but "...sometimes the magic works, sometimes it does not." Image shape as well as FWHM was never as good as it should have been, with elongated images. We finally had to give up as conditions no longer permitted us to continue with this observation.

  • Dwarf/fuobj8 - A completion program, shot the 2nd g filter OB to complete the observations of this Dwarf field. Seeing and image shape also weren't great, though.

  • Quasar/Point4 and Quasar/Point6 were shot in g+r, U tested and found problematic vis-a-vis the guider again. Seeing and image quality were much better, around 0.9"-1.1" but with nice round images.

  • Milne/SN2005cf - In the last 30-odd minutes available we decided to give it a shot given the seeing. Two problems, one was rapid deterioration of the seeing, the other was the realization as the OB executed that it was going to take a lot longer than the 0.3h indicated in the queue scheduler. Something broke down in communication, because the OB as written tries to take 4200s of exposure time (2x300s @ V, 4x300s @ B, and 8x300s @ r). The OB hadn't been vetted, and it should have been broken into components.

Engineering Info:
  • We noticed that on some occasions when the guider would lose it and run open loop, the images were doubled. See notes above, we hadn't seen the LBT jump like that - hard to say if the guider was involved or not, a couple times it was clearly not (unless it was trying in the background as part of a recovery process, but there was no indication of that on the GUI). Wind was never much over 10m/s, and we had the ventilator doors closed. The times give you an idea of where to look

  • We tried to note for the software teams when the GUI jammed (which it did a lot when we weren't touching it), when the archive stopped up (which it also did a lot, sometimes for 8-10 minutes at a time), and when the guider failed (...). We did our best to record the relevant parameters at each episode.

Notes added later (Olga Kuhn):
  • Jump/Double images: I checked 4 images where double stars were noted:
    • 034724 - more like a tail extending to the upper left than a double. Parallactic angle = 45, PA=0. Angle of elongtion(tail) is about 45deg E of N and corresponds to elevation axis.
    • 054444 - double images at an angle of about 60 deg E of N (+Y). Parallactic angle = 5.75, PA=0. Angle between image centers doesn't correspond to elevation or azimuth axis.
    • 055845 - double images at an angle of about 80 deg E of N (+Y). Parallactic angle = 11.4, PA=0. Angles between image centers doesn't correspond to elevation or azimuth.
    • 073843 - double images at an angle of about 120 deg E of N (+Y). Parallactic angle = 120, PA=0 - corresponds to jump in elevation. * From just a quick look at these images so far, it seems like 2 of these are tied to elevation, though only one shows clear jump, the other just a tail. The other 2 which do show clear jumps don't seem to be jumps along either axis.


That's all, we're beat. Forecast for tomorrow night is looking bad, increasing clouds with a chance of snow on Wednesday late. We can already see stuff coming in from the Pacific on the IR satellite loop.

-- RichardPogge - 19 Mar 2007
Topic revision: r5 - 21 Jul 2009, NormCushing
This site is powered by FoswikiCopyright © by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
Ideas, requests, problems regarding Foswiki? Send feedback