SURFACE LAYER TURBULENCE PROFILING WITH THE SL- SLODAR AND LUSCI AT ESO PARANAL OBSERVATORY

Transcription

1 Florence, Italy. Adaptive May 2013 Optics for Extremely Large Telescopes III ISBN: DOI: /AO4ELT SURFACE LAYER TURBULENCE PROFILING WITH THE SL- SLODAR AND LUSCI AT ESO PARANAL OBSERVATORY G. Lombardi 1,a, M. Sarazin 2, F. Char 3, C. González Ávila 4, J. Navarrete 1, A. Tokovinin 5, R.W. Wilson 6, and T. Butterley 6 1 European Southern Observatory, Casilla 19001, Santiago, Chile 2 European Southern Observatory, Karl-Schwarzschild-Straße 2, Garching bei Muenchen, Germany 3 Universidad de Antofagasta, Unidad de Astronomia, Av.. U. de Antofagasta Antofagasta, Chile 4 Las Campanas Obs., Carnegie Institution of Washington, Colina El Pino Casilla 601, La Serena, Chile 5 AURA/CTIO/NOAO, Colina El Pino Casilla 603, La Serena, Chile 6 University of Durham, Department of Physics CfAI, South Road, Durham, UK Abstract. In the context of the Surface Layer investigation at ESO Paranal Observatory, a Surface Layer Slope Detection And Ranging (SL-SLODAR) instrument prototype has been used at Paranal during 2012, while Lunar Scintillometer (LuSci) measurements campaigns are being carried out since Simultaneous Surface Layer profiling data from the two instruments are analysed in order to compare the two instruments to enforce their reliability and finely characterize the Paranal Surface Layer profile. Instruments, data acquisition and analysis, and results are discussed. 1. Introduction This paper describes the comparison of Surface Layer (SL) measurements carried out at ESO Paranal Observatory in 2012 by a Surface Layer Slope Detection and Range (SL-SLODAR) prototype instrument and a Lunar Scintillometer (LuSci) instrument in the context of the SL investigation at the observatory. 2. Instruments 2.1. The Surface Layer SLODAR The SL-SLODAR uses an optical triangulation method observing wide separation binaries (>100 ). The profile is determined from the spatial covariance of the slope of the wavefront phase aberration at the ground for the two different paths through the atmosphere defined by a double star target [1]. The SL-SLODAR gives the Cn 2 (h) for 8 layers starting from the telescope pupil with a resolution (Δh) that varies between ~6 and ~16 m depending on the binary separation (θ) and its zenithal distance. A reflective wedge diverts the light for the two target stars into separate wavefront sensors with independent detectors [2]. Figure 1 shows the SL-SLODAR at Paranal and a simple scheme of the SLODAR technique. a glombard@eso.org

2 Third AO4ELT Conference - Adaptive Optics for Extremely Large Telescopes Fig. 1. (left) the SL-SLODAR at Paranal; (right) a simple scheme of the SLODAR technique. Fig. 2. (left) LuSci concept; (middle) LuSci at Paranal; (right) the LuSci instrument 2.2. The Lunar Scintillometer LuSci consists of a linear array of photo-detectors pointed at the Moon. Fast fluctuations of moonlight are detected by six photodiodes (PD) in linear configuration, digitized and recorded in the computer. At large distances where the light cones from detector pairs overlap, turbulence produces correlated signals, while at shorter distances there is no correlation (see Figure 2). The optical turbulence profile is determined by using models of the turbulence spectrum and of the Lunar shape [3]. The Moon illumination must be >80%, for this reason LuSci can only be used 10 days around Full Moon. 3. Databases and analyses description The SL-SLODAR database has been collected with observations carried out between March 23rd, 2011 and July 8th, The LuSci database and the Paranal Meteo databases are much larger, nevertheless for this analysis they have been considered between the mentioned dates only (see Table 1). The SL-SLODAR entrance pupil (primary mirror of the Ritchey-Chrétien telescope) is considered 2 m above ground on average. The LuSci height above ground is considered in the barycenter of the 6 PD and is 1 m on average. Being J the integral of the Cn2 for each instrument!=!!"#!!! (ℎ)!ℎ!!"# (1) we want to fix the integration limits for the two instruments in order to exactly integrate the same range of altitudes above ground. This means that when taken separately, the two instruments have different integration limits, but when compared relatively to the height above ground, those limits correspond to the same exact altitudes above the ground.

3 Table 1. Databases description and altitudes integration range Data set between 23/03/2011 and 08/07/2012 N. of Points Height above ground SL-SLODAR data set m (OTA entrance pupil) LuSci data set m (6 PD barycenter) METEO data set (wind speed and wind direction) m and 30 m Simultaneous SL-SLODAR and LuSci data points 9748 Simultaneous SL-SLODAR and Meteo data points Simultaneous SL-SLODAR, LuSci and Meteo data points 9657 Being Dh the width of the SL-SLODAR layers, the integration limits for each instrument have been fixed as follows: h MIN corresponds to the LuSci minimum sensed altitude that is 3 m above the instrument (h MIN = 3 m), that is 4 m above ground and 2 m above the SL-SLODAR pupil (h MIN = 2 m); remembering that the SL-SLODAR entrance pupil is 2 m above ground and that LuSci is 1 m above ground, for LuSci h MAX corresponds to the SL-SLODAR maximum sensed altitude with the sum of the difference in altitude between the two instruments (h MAX = [7.5 Dh] + 1), while for the SL- SLODAR h MAX is simply 7.5Dh; when comparing the first 20 m above ground (Surface Layer, see [4]), for LuSci h MAX = 20 1 = 19 m, while for the SL-SLODAR h MAX = 20 2 = 18 m. Fig. 3. Comparison of the J values for SL-SLODAR and LuSci in the first 20 m above ground (SL) vs BETA.

4 Fig. 4. Comparison of ΔJ vs BETA for the entire SL-SLODAR integral above h MIN. β (or BETA) is the slope of the turbulence power spectrum delivered by the SL-SLODAR. It is intended purely as a diagnostic tool to indicate whether the Cn 2 profile can be trusted. When BETA is significantly less than (Kolmogorov law value) this generally indicates that the wind speed is low and the data sets are too short to fully sample the low frequency components of the turbulence. The use of this parameter in this paper as to be intended as a comparison tool with the limitations and the precautions deriving from its estimation. We define ΔJ as the difference between simultaneous SL- SLODAR and LuSci integrals in the same range of altitudes, several analyses have been carried out. Figure 3 shows the comparison of the J values for SL-SLODAR and LuSci in the first 20 m above ground (SL) vs BETA for any wind direction, for northerly winds only and for southerly winds only. SL-SLODAR data points have been acquired using northern or southern targets in the sky. Figure 4 and Figure 5 show the comparison of ΔJ vs BETA, having the SL-SLODAR observing a northern or a southern target. The analysis is developed considering both the entire SL-SLODAR sensed altitude (above h MIN ; Figure 4) and the first 20 m above ground only (SL; Figure 5). In the SL, J LUSCI looks more stable than J SL-SLODAR for different BETA. Nevertheless, a slight increasing trend with BETA is present, and values are more similar to those from SL-SLODAR when BETA is close to Interestingly, for southerly winds, the two instruments look pretty identical. Around the Kolmogorov value, the integrals form the SL-SLODAR and LuSci are pretty much the same. This is valid also in the first 20 m above ground only (SL). It is interesting to notice that the difference between J SL-SLODAR and J LUSCI is more significant in connection with northerly winds on the Paranal platform.

5 Fig. 5. Comparison of ΔJ vs BETA for the SL (20 m above ground). In Figure 6 we plot the wind-roses in connection with BETA measurements delivered by the SL- SLODAR. As shown in the figure, there is not a clear connection between BETA and the wind direction on the Paranal platform. In Figure 7 ΔJ for the entire SL-SLODAR sensed altitude (having the SL-SLODAR observing a northern or a southern target) is plotted against simultaneous wind speed measurements at 10 m above ground. From the figure it clearly appears that ΔJ is more significant when the wind speed is <3 m s 1. The SL-SLODAR and LuSci agree pretty well when the wind speed is >3 m s 1. Finally, in Figure 8 we show the linear regression passing by the origin of J SL-SLODAR vs J LUSCI in connection with wind speed lower or higher than 3 m s 1. The plot on the left shows the linear regression calculated for data concerning the entire SL-SLODAR sensed altitude. For wind speed lower or higher than 3 m s 1, the slope of the fit is 0.92 against 0.72 calculated for any wind speed. The plot on the right concerns the first 20 m above ground only (SL). For wind speed lower or higher than 3 m s 1, the slope of the fit is 0.84 against 0.59 calculated for any wind speed. Results in Figure 8 suggest that wind speed higher than 3 m s 1 allow to have more reliable turbulence profile measurements from both instruments.

6 Third AO4ELT Conference - Adaptive Optics for Extremely Large Telescopes Fig. 6. BETA wind roses. Fig. 7. ΔJ vs wind speed.

7 Fig. 8. J SL-SLODAR vs J LUSCI for different wind speed range. (left) entire SL-SLODAR sensed altitude; (right) 20 m above ground (Surface Layer only) 4. Conclusions We have compared the SL-SLODAR and LuSci instruments using databases of simultaneous data acquired at ESO Paranal Observatory during Around the Kolmogorov value of the slope of the turbulence power spectrum delivered by the SL- SLODAR (3.667), the integrals form the SL-SLODAR and LuSci are pretty much the same. There is not a clear connection between BETA and the wind direction on the Paranal platform. Both instruments agree very well when the wind speed on the Paranal platform is >3 m s 1. This last result suggests that wind speed higher than 3 m s 1 allow to have more reliable turbulence profile measurements from both instruments for further analyses of the Surface Layer. Furthermore, the disagreement of the two instruments in connection with wind speed <3 m s 1 also suggests that the wind speed is a critical parameter to be taken into account before the treatment of the data. 5. References 1. T. Butterley, R.W. Wilson, and M. Sarazin, MNRAS, 369, pp. 835 (2006) 2. J. Osborn, R.W. Wilson, T. Butterley, H. Shepherd, M. Sarazin, MNRAS, 406, pp. 1405O (2010) 3. A. Tokovinin, E. Bustos, and A. Berdja, MNRAS, 404, pp (2010) 4. G. Lombardi, J. Melnick, R.H. Hinojosa Goñi, J. Navarrete, M. Sarazin, A. Berdja, A. Tokovinin, R.W. Wilson, J. Osborn, T. Butterley, and H. Shepherd, Proc. SPIE, 7733, pp (2010)