What is the Seeing at LBT?

The frequently asked question: "What is the seeing at LBT?" can refer to either "What is the seeing right now?" or "What is the historical distribution of seeing?" Of course, you don't really want to know the site seeing. You, the observer, actually want to know the delivered image quality of the telescope plus atmosphere plus instrument (which we all imprecisely call "seeing").

What is the seeing right now?

We measure the "seeing" (delivered image quality of the telescope) every night at LBTO in a number of ways, but almost all of them have problems. The AGw1/LUCI1 guider is the only device we have (beyond LUCI1 itself) that can accurately measure seeing better than 0.7 arcsec. Here's my (JMH) summary of the various seeing measurements and their limitations: a) AGw1 guider for LUCI1: 2 sec r-band images; minimum value 0.25 arcsec FWHM (depends on binning of the guider)

b) AGw2 guider for LUCI2: 2 sec r-band images; minimum value 0.46 arcsec FWHM (guider camera is slightly defocussed)

c) MODS1 guider: 2 sec r-band images; minimum value 0.65 arcsec FWHM (guider camera is significantly defocussed, plus vignetting problems of the active optics, and MODS often guides with CLEAR filter so guider image shows atmospheric dispersion. The defocus and vignetting of MODS1 guider/wfs were corrected in October 2016.)

d) MODS2 guider: 2 sec r-band images; minimum value 0.45 arcsec FWHM (guider camera slightly defocussed, and MODS often guides with CLEAR filter so guider image shows atmospheric dispersion)

e) AGw7 guider for PFU-SX: filter and exposure times vary

f) AGw8 guider for PFU-DX: filter and exposure times vary

g) S-H wfs (on all 6 AGws): 30 sec r-band S-H spots; minimum value 0.55-0.6 arcsec FWHM (limited by pixel sampling on wfs).
AGW8 wfs can get down to 0.43 arcsec FWHM mean spot size.
MODS1 wfs can get down to 0.51 arcsec FWHM mean spot size.

h) on-telescope DIMM: The mininum measured value is between 0.5 - 0.58 arcsec. The minimum should be ~0.4 arcsec based on r0 spacing of the two apertures. I presently suspect that there is noise in the centroiding process, AND that we don't have a consistent focussing procedure. (more information can be found in Issue 2897 - the focusser was replaced in Summer 2015) See the histogram below that shows the DIMM measurements corrected by removing a 0.5 arcsec floor. Correlation plots below with the guider fwhm suggest that a better correction is simply subtracting 0.2 arcsec from the DIMM estimate.

All of the above are available as real-time (~1 minute delay) plots [LBTplot] while you are observing (depending on the focal stations being used). They are also available in the TCS Telemetry system (S-H spots are just being added to telemetry in Spring 2015, but are available from TCS logs).

Additional potential measures of the seeing are the wavefront rms residuals from the AO systems. You can see these numbers on the W-sensor GUIs, but they are not yet available in plots or telemetry.

What is the historical seeing distribution at LBT?

The simple answer to the question is "no" - we do not have any long-term, nicely-conditioned, seeing statistic for LBT. The best data is the SCIDAR data taken at VATT ~8 years ago around the start of the ARGOS activities. See Masciadri, Stoesz, Hagelin, Lascaux (2010) "Optical turbulence distribution with standard and high resolution at Mt.Graham" MNRAS, 404, 144 http://adsabs.harvard.edu/abs/2010MNRAS.404..144M [see also Sebastian E. Egner (2006) 115g001a, and Egner, Masciadri, McKenna (2007), and Egner & Masciadri (2007)]

My (JMH) impression is that the site median seeing is consistent with 0.60-0.65 FWHM measured in the ancient studies (pre-1990) [see Cromwell, Haemmerle & Woolf (1988) in VLT Instrumentation, and Cromwell, Haemmerle & Woolf (1990) in SPIE 1236 pp520-529]. We also have some residual dome seeing, so I would say that the median seeing at LBT zenith today is somewhere around 0.75 arcsec FWHM. What the historical studies don't show well is that there are are long tails in the distribution (at the 5 percentile level) of really good seeing (<0.25 arcsec) and horribly bad seeing (up to 5 arcsec).

The other thing I believe I know (without hard evidence) is that the outer scale of the turbulence is small (less than 30 m) most of the time.

Seeing Distribtutions

Seeing Distribution Estimated from Guider Images 2014

  • Histogram of Guider Images for All of 2014 (by JHill):
    • This histogram includes all SX guider images where a star was detected. Thus it includes uncollimated data and a few images which are enhanced or destroyed by adaptive optics. Saturated images and fake guide stars have been removed.
    • Most of the LUCI1 images are 2-second exposures in r', but exposure times may vary. Many of the MODS1 images are with an open filter and may frequently show elongation from differential refraction.
    • As mentioned above, the MODS1 guider is significantly defocussed as shown by this histogram.
    • This histogram represents the actual guider image size and is not corrected for zenith angle or wavelength. gcs.sx.guihist.20141231lbcobs.jpg

Seeing Distribution Estimated from Guider Images 2018

The solid curves are histograms of the observed guider FWHM, and the dashed curves represent those same images corrected to estimate the FWHM at zenith (assuming that the FWHM is dominated by atmospheric seeing). The full distribution has been separately plotted for MODS and LUCI. Adaptive Optics and ESM corrected images have been removed before making the histogram. Guiding from instruments other than MODS/LUCI has not been included.

  • Histogram of SX guider images for all of 2018:
    gcs.sx.guihist.20181231_fullyear.jpg

  • Histogram of DX guider images for all of 2018:
    gcs.dx.guihist.20181231_fullyear.jpg

Seeing Distribution Estimated from DIMM for 2014-2015

  • 2014_FWHM_hist.tiff: Histogram of Seeing Measured by the DIMM (by JChristou)
    • This histogram uses DIMM data from May 2013 through Feb 2015.
    • See the comments above relating to how the LBT DIMM is not sensitive to seeing below about 0.6 arcsec.
    • This histogram is corrected to the zenith regardless of the zenith distance at which the observation was made.
  • Histogram of Seeing Measured by the DIMM (by JChristou):
    2014_FWHM_hist.jpg

  • Histogram of DIMM seeing estimates for all of 2015:
    • This histogram uses DIMM telemetry data from Jan 2015 through Dec 2015.
    • The black curve is the histogram of the raw seeing measurements.
    • The blue curve is the histogram of the dimm measurements corrected to zenith (same as the 2014 plot above).
    • See the comments above relating to how the LBT DIMM is not sensitive to seeing below about 0.6 arcsec. The green curve has subtracted in quadrature a measurement floor of 0.5 arcsec from the raw measurements and then corrected the result to zenith.

dhist.all2015jhill.jpg

Histogram of DIMM seeing estimates for all of 2016

dhist.all2016jhill.jpg

Be aware that some DIMM parameters related to star selection and minimum number of samples were changed in Fall 2016, so the 2016 histogram may not be statistically comparable to the 2015 histogram.

Histogram of DIMM seeing estimates for all of 2017

dhist.all2017.jpg

Histogram of DIMM seeing estimates for all of 2018

dhist.all2018.jpg

Histogram of DIMM seeing estimates for all of 2019

dhist.all2019.jpg

Comparing Guider FWHM to DIMM Measurements

Here is a correlation plot (on the right) of AGw2 (LUCI2 on RFBG, DX) guider FWHM (corrected from r' to a wavelength of 0.5 micron, not corrected for a small defocus) versus DIMM seeing (corrected to match the elevation of the guider measurement). The left plot is the same except with 0.5 arcsec of unspecified "instrumental noise" (estimated from histograms above without physical justification) subtracted in quadrature from the DIMM seeing. As you can see, a better fit to a straight line could be had by subtracting linearly 0.2 arcsec from the DIMM seeing (matching the yellowish line rather than the purplish one).

gcs.dx.guidevd.20160228airmass.jpg

Comparison of AGw2 guider image vs DIMM for Sep-Dec 2018

Three months of AGw2 (LUCI2 on RFBG, DX) guider data compared to DIMM measurements for Sep-Dec 2018. The DIMM measurements have been corrected to the elevation of the guider images, and the guider images have been scaled from r' to V (500 nm). No correction for the small focus offset of the guider has been applied. Plotted with: ./guidevdimm_hdfextract.py -d 20181231 -e 13.9 -b 2400 -s DX

gcs.dx.guidevd.20181231jhill.jpg

Comparison of AGw1 guider image vs DIMM for Sep-Dec 2019

Three months of AGw1 (LUCI1 on LFBG, SX) guider data compared to DIMM measurements for Sep-Dec 2019. The DIMM measurements have been corrected to the elevation of the guider images, and the guider images have been scaled from r' to V (500 nm). No correction for the small focus offset of the guider has been applied. Plotted with: ./guidevdimm_hdfextract.py -d 20191231 -e 13.9 -b 2400 -s DX

gcs.sx.guidevd.20191231observer.jpg

Seeing Predictions

Predicted Skew-T distributions for each night on Mt. Graham are available from UA Atmospheric Sciences (through 2019). See JHill's xplot web page. [The full discussion of seeing prediction is beyond the scope of this wiki page.] We also have detailed predictions of seeing and other weather factors from the ALTA project. See the nightly predictions at http://alta.arcetri.astro.it. See ALTA publications by Turchi etal 2017, Turchi etal 2019, and Masciadri etal 2020 (all in MNRAS).

Click here: TrendsInDomeSeeing for a look at a summer 2020 student research project on Trends in Dome Seeing During the Night.

-- %USERSIG{JohnHill - 2015-03-07}%

Topic attachments
I Attachment Action Size Date Who Comment
2014_FWHM_hist.jpgjpg 2014_FWHM_hist.jpg manage 84 K 07 Mar 2015 - 04:46 JohnHill Histogram of Seeing Measured by the DIMM
2014_FWHM_hist.tifftiff 2014_FWHM_hist.tiff manage 59 K 07 Mar 2015 - 04:34 JohnHill Histogram of Seeing Measured by the DIMM
dhist.all2015jhill.jpgjpg dhist.all2015jhill.jpg manage 247 K 09 Jan 2016 - 01:48 JohnHill Histogram of DIMM seeing estimates for all of 2015
dhist.all2016jhill.jpgjpg dhist.all2016jhill.jpg manage 333 K 08 Mar 2017 - 03:33 JohnHill Histogram of DIMM seeing estimates for all of 2016
dhist.all2017.jpgjpg dhist.all2017.jpg manage 328 K 19 Feb 2019 - 23:20 JohnHill Histogram of DIMM seeing estimates for all of 2017
dhist.all2018.jpgjpg dhist.all2018.jpg manage 332 K 19 Feb 2019 - 23:21 JohnHill Histogram of DIMM seeing estimates for all of 2018
dhist.all2019.jpgjpg dhist.all2019.jpg manage 326 K 13 Feb 2020 - 23:49 JohnHill Histogram of DIMM seeing estimates for all of 2019
gcs.dx.guidevd.20160228airmass.jpgjpg gcs.dx.guidevd.20160228airmass.jpg manage 546 K 18 May 2016 - 01:27 JohnHill plot comparing guider fwhm to dimm measurements
gcs.dx.guidevd.20181231jhill.jpgjpg gcs.dx.guidevd.20181231jhill.jpg manage 1 MB 04 Mar 2019 - 23:01 JohnHill Comparison of AGw2 guider image vs DIMM for Sep-Dec 2018
gcs.dx.guihist.20181231_fullyear.jpgjpg gcs.dx.guihist.20181231_fullyear.jpg manage 464 K 04 Mar 2019 - 20:19 JohnHill Histogram of DX guider images for all of 2018
gcs.sx.guidevd.20191231observer.jpgjpg gcs.sx.guidevd.20191231observer.jpg manage 1 MB 17 Feb 2020 - 18:24 JohnHill Comparison of AGw1 guider image vs DIMM for Sep-Dec 2019
gcs.sx.guihist.20141231lbcobs.jpgjpg gcs.sx.guihist.20141231lbcobs.jpg manage 273 K 07 Mar 2015 - 04:21 JohnHill Histogram of Guider Images for All of 2014
gcs.sx.guihist.20181231_fullyear.jpgjpg gcs.sx.guihist.20181231_fullyear.jpg manage 474 K 04 Mar 2019 - 20:18 JohnHill Histogram of SX guider images for all of 2018
Topic revision: r23 - 22 Oct 2020, JohnHill
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