Документ взят из кэша поисковой машины. Адрес оригинального документа : http://site2010.sai.msu.ru/static/doc/ATokovinin_site2010.pdf Дата изменения: Mon Oct 18 12:28:22 2010 Дата индексирования: Mon Oct 1 20:09:52 2012 Кодировка: Поисковые слова: р п р п р п р п р п р п р п р п р п р п р п р п р п р п р п

Surface-layer turbulence measurements with lunar scintillometer Idea of the method
Hardware & software Campaigns & results New horizons

Andrei Tokovinin NOAO/CTIO
Kislovodsk Oct. 2010 1


Turbulence measurement from scintillation of extended sources
Moon's angular diameter =0.5 deg Correlated scintillation for b > z


z=10m b=0.1m z=100m b=1m z=10km b=100m

BI(b) = W(b,z) Cn2(z)dz

Weighting functions zW(z) for b=0...0.4m and d=1cm
2

Kislovodsk Oct. 2010


The principle

I/ ~104

Kislovodsk Oct. 2010

2010 MNRAS, 404, 1186

3


Restoration of turbulence profile
Cn2 is nonlocal by definition. No "thin layers"! Fit covariance BI(b) to a smooth function Cn2(z) using W(b,z) Use 5 "pivot points", powerlaw segments Calculate seeing etc. from the model Old method: linear combination of data turbulence integrals in "layers"

Kislovodsk Oct. 2010

4


Example (1)

Kislovodsk Oct. 2010

5


Example (2)

Kislovodsk Oct. 2010

6


Pros and cons
Simple instrument Absolute calibration Solid theory (no
saturation, achromatic)

Does not measure high turbulence Sensitive to outer scale Wind bias NonKolmogorov?

Kislovodsk Oct. 2010

7


J.Beckers: SHABAR (ATST site campaign)

Various implementations

P.Hickson: lunar SHABAR (CTIO, Arctic & Mauna Kea) CTIO: lunar scintillometer, LuSci (several prototypes) ESO: several LuSci instruments Las Campanas: MooSci

Kislovodsk Oct. 2010

8


The CTIO instrument

Kislovodsk Oct. 2010

9


Instrument control software
Remote observing "Cold start" possible Correct pointing by webcam Automatic acquisition Periodic sky meas.

Software written by E.Bustos, for Windows
Kislovodsk Oct. 2010 10


What can go wrong? Pointing!

Flux vs. time plots
Kislovodsk Oct. 2010 11


Data processing (IDL)
Filter the data for each night Calculate covariances .cov file Calculate W(b,z) and fit OTP .tp file Use the OTP to compute SL seeing etc.
The code is available at http://www.ctio.noao.edu/~atokovin/profiler/code2.tar.gz
Kislovodsk Oct. 2010 12


Paranal: October 14, 2008

Kislovodsk Oct. 2010

13


Instrument comparison (ESO)

LuSci1

LuSci2

Kislovodsk Oct. 2010

14


Cerro Tololo, March 2009
Hickson's 12channel scintillometer (UBC) CTIO LuSci (LC) ESO LuSci3 (L3) MASSDIMM site monitor

UBC Mar 14/15

LC Mar 14/15

Kislovodsk Oct. 2010

15


Scintillometer comparison
log(Cn2): av. <0.12 rms 0.2 +30% Cn

2

Can SLODAR measure 1015 m2/3 ??
Kislovodsk Oct. 2010 16


Comparison with SLSLODAR
Turbulence integral JSL =Cn2(h) dh, in 1013 m1/3 limits (hmin,h
max

)

Paranal, 2009 (J. Osborn, A. Berdja) Osborn et al. 2010, MNRAS
Kislovodsk Oct. 2010 17


Comparison with microthermals

Penyon, Nov. 26, 2009, h=10m (J. Sebag)
Kislovodsk Oct. 2010 18


Comparison with DIMMMASS
Paranal Feb. 2009 Tololo Mar 2009

Penyon Nov 2009

Kislovodsk Oct. 2010

19


Where is the surface-layer turbulence?
Total seeing (DIMM) Free atm. (MASS) Surface layer (LuSci)

Kislovodsk Oct. 2010

20


Average profiles in the SL

Cn2(h) 1015 (h/30m)

1

J(6,500) = 1.3 (0.37")
Kislovodsk Oct. 2010

SL thickness and strength are meaningless without (hmin,hmax) SL integral is dominated by h
m in 21


Is the SL turbulence really weak?
Paranal, 1998: J(3,21)=0.4 (microthermal), no
difference between GSM (2m) and DIMM (6m)

Paranal, SLSLODAR: J(6,65)=0.6 Racine (2005): J(6,100)=0.5 Lunar scintillometers at Paranal, Tololo, Penyon, Mauna Kea: J(6,200) ~ 0.5
Turbulence integral JSL(hmin,hmax) in 1013 m DIMMMASS: JSL = 2...3 typically
Kislovodsk Oct. 2010 22
1/3


Can a DIMM be wrong?...
Northern wind

Els et al. 2009, PASP, 121, 922

... yes, sometimes!
Kislovodsk Oct. 2010

JGL(CTIO DIMM)/JGL(TMT DIMM)
23


The "SL discrepancy"
The assumption that all site monitors have zero local turbulence is questionable. We do not know how strong Jloc is. DIMMs give an upper limit on seeing, it can in fact be better! Local turbulence is detected in all telescopes by SCIDARs, SLODARs, etc. It matters under good seeing only
Kislovodsk Oct. 2010 24


Site seeing depends on the elevation of the site monitor (cf. R. Racine) Systematic bias by local turbulence A modeldependent concept (real distortions are not stationary, the Kolmogorov model is approximate; gravity waves?) Telescopes have internal seeing, too (1990s: ~2" now: 0.5" or 0"?)

Elusiveness of "site seeing"

Kislovodsk Oct. 2010

Maidanak, 1975

25


A typical situation?

Seeing 0.84"/0.62" with/without Jloc
Kislovodsk Oct. 2010 26


A new exciting challenge
If JSL=0.5 (0.2" seeing), can telescopes get images that good? A new round of dome/mirror environment study and optimization? Active optics widefield GLAO?
IMAKA (Chun et al. 2010): CFHT, 1o FoV, NGS Sitetesting and telescope optimization is not over yet!
Kislovodsk Oct. 2010 27


Acknowledgement
This presentation makes use of data obtained by many people at many sites. Thanks for sharing the data and for the collaboration!
CTIO ­ E. Bustos, A. Berdja ESO ­ M. Sarazin, G. Lombardi UBC ­ P. Hickson U. Durham ­ R. Wilson, J. Osborn LSST ­ J. Sebaq LCO ­ J. ThomasOsip http://www.ctio.noao.edu/~atokovin/papers/sl.pdf
Kislovodsk Oct. 2010 28