Stereo SCIDAR: Profiling atmospheric optical turbulence with improved altitude resolution

Transcription

1 Stereo SCIDAR: Profiling atmospheric optical turbulence with improved altitude resolution, Richard Wilson, Vik Dhillon, Remy Avila and Harry Shepherd

2 SCIDAR Triangulation method Peaks in crosscorrelation patterns determines height and strength of turbulent layer

3 Spatial fluctuations Z = 5 km Z = 10 km D=2.5 m r F = (λz) [z=10km à r F =0.07 m]

4 Spatial fluctuations Z = 5 km Z = 10 km D=2.5 m r F = (λz) [z=10km à r F =0.07 m]

5 Analysis Impulse function is altitude independent response of the system to a single turbulent layer Response function is wider for higher layers limits resolution Impulse function (Only showing 2km intervals for clarity)

6 Profile estimation Cn 2 (h) profile via fit of correlation function with impulse functions dh ~ 200 m dh ~ 1000 m 1D Correlation function

7 SCIDAR: SCIntillation Detection And Ranging Covariance High Layer Intensity Auto-Correlation (Time-Averaged) SCIDAR response functions

8 Generalised SCIDAR Conjugate below the ground Overlapping pupil images High Layer Low Layer Response from the ground G-SCIDAR response functions Intensity Auto-correlation (Time-Averaged)

9 Stereo SCIDAR Generalised SCIDAR but conjugate above the ground, with two cameras Separated pupils One set of correlation peaks Blind Spot High Layer Low Layer Response functions, c_alt=5000 m Intensity Cross-correlation (Time-Averaged)

10 Optical Conjugation

11 Conjugate SCIDAR Response functions c_alt=-2 km c_alt=5 km c_alt=7.5 km Any altitude Response is narrower near to conjugate altitude BUT, smaller amplitude too therefore very little signal from weak layers Blind spot We can probe any areas of interest Must scan to build up full profile

12 Comparison High Layer Low Layer Generalised SCIDAR: One camera, Auto correlation, 3 sets of peaks Low Layer High Layer Stereo SCIDAR: Two cameras, Cross correlation, 1 set of peaks

13 Simulation Results Conjugate altitude: -2 km 8 km Conjugate SCIDAR gives same result as Generalised SCIDAR

14 Simulation results Generalised SCIDAR Stereo SCIDAR Conjugating to 7.5 km allows us to resolve two closely separated layers, dh = 230 m at 10 km

15 Optomechanical design

16 On Sky results NOT data (25/03/13) conj. alt. = -2 km conj. alt. = -1 km If we conjugate the detector plane closer to the turbulence the local resolution is increased: conj. alt. = 0 km conj. alt. = 4 km dh (-2 km) = 420 m dh (4 km) = 290 m For a target altitude of 10 km.

17 On Sky results NOT data (25/03/13) conj. alt. = 7.5 km To go to higher altitudes the SNR of high layer becomes small due to scintillation from the ground (central peak and background noise). With a GLAO system would be able to conjugate much closer. dh = 207 m

18 On sky results JKT data (24/05/13) Conj. alt. = -2 km Something interesting?

19 On sky results (24/05/13) Conj. alt. =+1.5 km, dt = 2 mins Conj. alt. = 1.5 km Conj. alt. = -2 km

20 Wind Velocity Measurements

21

22

23

24

25

26

27 Wind Velocity Profile Wind profile 25/09/12, NOT 2.5 m, 9 Layers Cross correlations of sum of three time steps Cross correlations only contain one set of peak. Geometric wind profile algorithm rather than wavelet analysis Stationary central peak corresponds to dome seeing (dt +ve and ve to avoid confusion with other peaks)

28 Applications Stereo SCIDAR is two camera version of Generalised SCIDAR not necessarily conjugate below the ground Two cameras allow easy separation of conjugated pupil images negates correlation normalisation problem Can use larger magnitude difference Used to probe the atmosphere instead of full profile Might be useful for Pre-optomise tomographic reconstructors with resolution required for ELT scale (~200 m up to 20 km) Wind velocities for smart reconstructors (Gaetano Sivo, Friday 14:20) Validate performance Scheduling Validating and assisting Meteo models

29 Future work Validate with SLODAR at La Palma Support AO tomography (CANARY) Statistically significant data for turbulence characteristics Study short term variability of turbulent layers, required for AO tomography dependence on wind speed Could GLAO improve high altitude resolution? Access to JKT (1m a bit small) Test at other locations access to 2m telescopes?

30 Conclusions Stereo SCIDAR Generalised SCIDAR conjugate above the ground probe altitude of interest, with high resolution scan altitude ranges to build profile Two Cameras quick and easy re-conjugation normalise each pupil image separately increase magnitude difference tolerance (normally dm_max = 2) profile correction (from normalisation errors) not required easy wind velocity estimations 290 m resolution at 10 km demonstrated (c.f 420 m for one camera generalised SCIDAR) <200 m resolution at 20 km might be possible on 2m telescope with GLAO.