Seeing Measurements Differential Image Motion Monitor · ference XX on "Big Bang, Active Galac-tic...
Transcript of Seeing Measurements Differential Image Motion Monitor · ference XX on "Big Bang, Active Galac-tic...
597. F. Barone et al.: Gravitational WaveBackground from a Sampie of 330+4Pulsars. Astronomy and Astrophysics.
598. G. Contopoulos: Short and Long PeriodOrbits. Celestial Mechanics.
599. L. Milano et al.: Search for Contact Systems Among EB-Type Binaries. 11: ESLib and AR Boo. Astronomy and Astrophysics.
600. C. N. Tadhunter et al.: Very Extendedlonized Gas in Radio Galaxies: IV. PKS2152-69. Monthly Notices of the RoyalAstronomical Society.
601. D. Baade and O. Stahl: Rapid LineProfile Variability of the A-Type Shell-
and Possible Pre-Main Sequence StarHO 163296. Astronomy and Astrophysics.
602. D. Baade and O. Stahl: New Aspects ofthe Variability of the Probable Pre-MainSequence Star HR 5999. Astronomyand Astrophysics.
603. S. D'Odorico: Multiple Object Spectroscopy at ESO: Today's Facilities andFuture Prospects. Invited paper to aconference.
604. G. Setti: The Extragalactic X-RayBackground. Invited paper to appear inthe Proceedings of the YAMADA Conference XX on "Big Bang, Active Galac-
tic Nuclei and Supernovae, Tokyo,March 28-April 1, 1988.
605. S. Cristiani et al.: Quasars in the Field ofSA94. 111. A Colour Survey. Astronomyand Astrophysics.
606. F. Barone et al.: Search for Contact Systems among EB-Type Binaries. 111: UU Cnc and VZ Psc, ContactSystems Before the Common Envelope Phase? Astronomy and Astrophysics.
607. G. Contopoulos: Nonuniqueness ofFamilies of Periodic Solutions in a FourDimensional Mapping. CelestialMechanics.
Seeing Measurementswith a Differential Image Motion MonitorH. PEDERSEN, F. RIGAUT, M. SARAZIN, ESO
Concept of the DIMM
Seeing is possibly the most importantparameter describing a ground-basedastronomical observatory. Under conditions of good seeing, an aberration-freetelescope will produce sharp and brightimages. The astronomer can then explore the universe to greater depthsthan otherwise possible.
In recent years, a considerableamount of theoretical and experimentalseeing studies have been conducted.The action of the earth's atmosphere onthe quality of astronomical observationsis now understood in quite some detail,and it has also become possible to measure the prevailing seeing, without theuse of large and very expensive telescopes. This is obviously of great interest in the search for new observatorysites.
One particular instrument which cansimulate seeing conditions at larger telescopes is called the Differential ImageMotion Monitor, or DIMM. Its conceptgoes back at least to 1960, when it wasused for qualitative seeing studies [1].Later, F. Roddier [2] has shown its potential for quantitative measurements.This prompted ESO to use DIMM in thesearch for the site for the Very LargeTelescope.
The detector unit of DIMM houses anintensified CCD and is attached to analt-alt mounted, 350-mm apertureCassegrain telescope. All essentialfunctions are computer controlled, andtracking is assisted by an autoguider.The instrument is placed in open air on a5 m high tower. Typically, the telescopefollows a bright star for a couple ofhours, while the star crosses the meridian.
8
The full aperture of the instrument isused for self-calibration, while, in its regular mode of operation, the entrance isrestricted to two circular holes. Theseare 4 cm diameter, and spaced 20 cm,centre to centre. Under perfect conditions, light arrives as a plane wave,forming two images at fixed positions.The presence of turbulence in theearth's atmosphere causes the arrivaldirection to differ slightly between thetwo holes. The two spots on the detector will then shift relative to each other.Their time-averaged motion is proportional to the astronomical seeing. Inprinciple, the scaling factor is defined bythe system parameters, and does notdepend on any empirically determinedvalue. To demonstrate that this is reallythe case, we decided to compare theDIMM with standard seeing measurements at big telescopes. Such havebeen conducted on a regular basis forseveral years, using imaging CCDcameras [5]. In order to allow a comparison as realistic as possible, it wasnecessary to mount the DIMM inside adome. For the presently described tests,May 26 to 28, 1988, it was mounted onthe exterior of the 2.2-m telescope'smirror Gell.
Measurement in Parallel with the2.2-m Telescope
In spite of the fact that the instruments were observing in parallel, atleast two effects tend to complicate thecomparison. A relatively small effect isdue to turbulence within the 2.2-mmirror Gell. This is not measured by theDIMM, since it used its own optics. Thesize of the error is difficult to quantify.
We believe it is comparable to the measurement accuracy (0.1 "), or smaller.
A more severe effect is due to opticalaberrations in the 2.2-m. In order toquantify this, we refer to the last opticaltests, wh ich were conducted in December 1987. A set of Shack-Hartmannplates show that the main error is due toastigmatism, decentring coma beingnegligible and spherical aberration absent. At best focus, 80 % of the energyis concentrated within a diameter of0.45". For seeing - 1", this correspondsto a quadratic contribution of 0.35" interms of FWHM. To allow also for themirror-seeing problem, we have
Figure 1: The Differential Image MotionMonitor mounted on the 2.2-m telescope.
o 2 t1 6 8 10TIME (UT)
Figure 2: During the seeond night of observation, the seeing started around 0.8", and laterbeeame worse. In this presentation the 2.2-m data have been rebinned to mateh the DIMMtime slots, and eorreeted to a standard wavelength 560 nm.
relation (Figure 3) changes to S2.2m =
1.02 X SOIMM + 0.08 arcseconds.The difference from a one-to-one re
lation is more likely due to an undescribed problem in the 2.2-m data thanto a systematic underevaluation on partof DIMM. Therefore we believe that thepresent comparison provides sufficientreassurance for the use of DIMM as aquantitative seeing measurement tool.
DIMM in Operation
The first DIMM unit has been in regular use since April 1987 when it wasinstalled at Gerro Paranal, one of thecandidate sites for the VL1. Over 40,000individual seeing measurements areavailable from that place. A second system was recently built, and is nowmounted at another candidate site,Gerro Vizcachas, a few km south-east ofLa Silla.
The hardware of both systems is prepared for automatization. This will becompleted in the near future, whereafterseeing monitoring can be done on apermanent basis. As a by-product, weobtain quantitative data on the photometrie quality of the sky.
We note that seeing monitors are notonly of interest in the site-testing phase.Also remote controlled observations,and automated programme selectioncan benefit much from such information.
References1. M. Sarazin: "ESO-VLT Instrumentation far
site evaluation in Northern Chile", in Advaneed Teehnology Optieal Teleseopes111", SPIE, Vol. 628, 138-141 (1986).
2. J. Stock and G. Keller: "Astronomieal Seeing", in Stars and Stellar Systems Vol. 1,138-153 (1960).
3. F. Roddier: "The effeet of atmospherieturbulenee in optieal astronomy", in Progress in Optics XIX, editor E. Wolf,281-376 (1981).
4. M. Sarazin: "Site evaluation for the VLT: aStatus Report", The Messenger No. 49,37-39 (1987).
5. H. Pedersen: "Seeing at La Silla", InternalReport, ESO (1988).
The simultaneously acquired DIMMmeasurements lasted 1 minute each,and the statistical error was 6%. Theconversion from differential image motion to equivalent seeing was also madeat 560 nm. Note, however, that theDIMM results are independent of thedetector's chromatic characteristics,since image motion is not a function ofwavelength (3).
Figure 2 shows the results obtainedduring one of the test nights. The seeingranged from 0.8" in the beginning, tomore than 3". Periods without datacorrespond to change of star, combinedwith a careful refocussing of the 2.2-m.The 2.2-m data have been rebinned intime to correspond to the time-slots ofthe DIMM.
The direct comparison (using datafrom both nights) gives a best linear fit:S2.2m = 1.05 X SOIMM + 0.14 arcseconds,with a correlation coefficient of 0.965.Following correction for the knownaberrations of the 2.2-m telescope the
6"
88 May 27/28
4" 2.2m (offseL 2") ~::E
~ ~::r:~;;::
~
2" Seeing Mon.
assigned DA" as the intrinsic imagequality of the 2.2-m telescope.
During the nights of simultaneousseeing measurements, the 2.2-m GGDcamera employed adetector with pixelsize 0.365". At this resolution propersampling is ensured. The optical filterwas RG-9 and we have assumed thatthis corresponds to 800 nm effectivewavelength. The integration time was 50seconds, followed by 15 seconds deadtime. Standard software was used to fitGaussian profiles to star-images. Theseeing at full width, half maximum, wascalculated in two orthogonal directions,Sx and Sy, and transformed to 560 nmwavelength. The astigmatism wasnoticed through slow drifts of the focus,caused by temperature changes of thetelescope structure. To minimize theeffect, we used the average, S, in thetwo directions. The rms error of a singledetermination of S is smaller than 0.1",as shown by an analysis of S. -Sy.
4
.. ":0;, ,,' ::'
1 234F'WHM SEEING MONITOR (Arcsec)
Figure 3: Following a eorreetion for imageaberrations at the 2.2-m, the eorrelation between the two sets of data is aeeeptabie.
Visiting Astronomers(October 1, 1988-Apri11, 1989)
Observing time has now been alloeated forPeriod 42 (Oetober 1, 1988-April 1, 1989).As usual, the demand for telescope time wasmuch greater than the time actually available.
The following list gives the names of thevisiting astronomers, by telescope and inchronological order. The complete list, withdates, equipment and programme titles, isavailable from ESO-Garching.
3.6-m Telescope
Oetober 1988: Wampler, Guzzo/CollinslHeydon-Dumbleton, Soucail/FortlTysonlTurner, Mellier/MatheziSoucail, Maccagni/GioiaiMaccaearoNettolani, lovino/ShaverlCristiani/Clowes, Danziger/Gilmozzi, Moorwood/Oliva, Danziger/Moorwood/Oliva, Danziger/Fosbury/LucylWampler/Bouchet.
November 1988: Moeller/K. Rasmussen,Danziger/Cristiani/Guzzo, Barbieri/Clowesl
9