OPTICAL TURBULENCE MEASUREMENTS - SITE TESTING

At present many instruments conceived to measure one or a few astroclimatic parameters exist. These instruments are based on different physical principles and can be divided in four big families:

(1) Vertical profilers: Generalized Scidar (Fuchs et al. 1998, Avila et al. 1997), Multi Aperture Scintillation Sensor - MASS  (Kornilov et al. 2003, Tokovinin et al. 2003), radiosoundings with macrothermal sensors for CN2 measurements (Azouit & Vernin 2005), SLope Detection and Renging - SLODAR (Wilson, 2002)

(2) Integrated-based instruments: Differential Image Motion Monitor - DIMM (Sarazin & Roddier 1990), Generalized Seeing Monitor - GSM, (Martin et al. 2000)

(3) Instruments dedicated to monitoring turbulence developed in the boundary layer: SHABAR (Beckers, 1993 , Tokovinin et al. 2010), MAST equipped with microthermal sensors
for CN2 measurements (Azouit & Vernin 2005), HVR-Generalized Scidar (Egner & Masciadri 2007), Low Layer Scidar - LOLAS (Avila et al. 2008), SLope Detection and Ranging - SLODAR at high resolution (Osborn et al. 2010)

(4)
Instruments detecting the height of the turbulent surface layer: SOnic Detection and Ranging - SODAR, SNODAR (Bonner et al. 2009).


All these instruments have some specificities and can be suitable to be used for different goals.


An extended site testing campaign (43 nights) has been recently completed above Mt. Graham (Masciadri et al. 2010) using two instruments: the Generalized Scidar (GS) and the High Vertical Resolution Generalized Scidar (HVR-GS), a recent method we recently proposed (Egner & Masciadri, 2007).

The GS
is an optical instrument that measures the scintillation maps produced by binary stars on the pupil of the telescope (at least 1.5 m size). In this case the GS has been run at the focus of the Vatican Adavnced Technology Telescope (VATT). The optical turbulence vertical distribution (CN2 profiles) is retrieved from the auto-correlation of the scintillation maps. The binary stars have the following properties: a separation θ within the (3" - 14") range;   m1, m2 ≤ 5.8 mag; Δ m ≤ 1.5 mag. Wind speed vertical profiles can be retrieved from the cross-correlation of scintillation maps taken with a time lag of typically 20 - 40 msec.

Using the GS it has been possible to retrieve the vertical distribution of the optical turbulence with a resolution of ~ 1 km all along the 20 km. Using the HVR-GS it has been possible to retrieve the turbulence spatial distribution in the first kilometer from the ground with a vertical resolution of 25-30 m.

The astronomical site of Mt. Graham is located in Arizona (USA). A few moments of the ForOT group activity on the summit can be find here.

Generalized Scidar runs
1. [28/5/2007 - 4/6/2007]: 8 nights
2. [15/10/2007 - 28/10/2007]: 15 nights
3. [21/2/2008 - 3/3/2008]: 11 nights
4. [9/11/2008 - 20/11/2008]: 11 nights


 
Fig 1-Left: Generalized Scidar  mounted to the focus of the Vatican Advanced Technology Telescope (D = 2m) at Mt. Graham.
Fig 1-Right: Temporal evolution of the  CN2 extended on the 20 km (whole atmosphere) during a night . The turbulence evolves in intermittent way in the low as well as high atmosphere producing turbulence bumps at different heights in different instants during the night.                                                      



INTEGRATED ASTROCLIMATIC RESULTS: STATISTICAL RESULTS

 
Fig 2:
(from Masciadri et al. 2010): Cumulative distribution (43 nights) of four integrated astroclimatic parameters. Top left: the seeing in the total atmosphere (including the dome contribution). Top-right: the seeing in the free atmosphere (h > 1 km). Bottom-left: the isoplanatic angle. Bottom right: the wavefront coherence time. Thick lines: the whole sample. Dotted lines: summer time. Thin line: winter time.




Fig 3:
(from Masciadri et al. 2010): Cumulative distribution of the dome seeing for all the 43 nights (thick line), the summer (dotted line) and the winter (thin line) period.


Table 1:
(from Masciadri et al. 2010): Median, first and throd quartiles values of the main integrated astroclimatic parameters above Mt. Graham (43 nights): seeing in the total atmosphere (included the dome seeing), isoplantaic angle, wavefront coherence time, integrated equivalent wind speed.

Knowing that the median seeing in the whole atmosphere (included the dome seeing) is ε = 0.95" and that the median dome seeing is εd = 0.52", it follows that the median seeing related to the whole atmosphere without the dome contribution for the richest statistic we collected so far (43 nights) is ε = 0.72".


OPTICAL TURBULENCE VERTICAL DISTRIBUTION

The optical turbulence distribution has been estimated with a Generalized Scidar above Mt. Graham with a statistical sample of 43 nights.


Fig.4:
(from Masciadri et al. 2010): Median CN2 profile obtained with a sample of 43 nights and a Generalized Scidar


OPTICAL TURBULENCE VERTICAL DISTRIBUTION AT HIGH RESOLUTION IN THE FIRST KILOMETER FROM THE GROUND


The optical turbulence vertical distribution has been estimated above Mt. Graham in the first kilometer with a vertical resolution of 25-30 m with a HVR-GS. The statistical sample is constituted by 43 nights.

   

Fig. 5:
(from Masciadri et al. 2010): Red line:  optical turbulence vertical distribution (J profile, 50 % case) above Mt. Graham in the first kilometer above the ground.  J is the integral of the CN2 on the thickness equal to h. For a precise definition of J (see Eq.9, Masciadri et al. 2010). The vertical resolution is 25-30 m (HVR-GS). Black line: optical  turbulence vertical distribution above Mauna Kea (Chun et al. 2009) in the first 600 m above the ground. The vertical resolution is in the range (15 m - 80 m). The higher resolution is obtained with LOLAS (Avila et al. 2008) in the first 60 m. The lower resolution is obtained with a SLODAR (Wilson et al. 2009). At ~ 45 m from the ground, it is visible a local minimum above Mauna Kea. This height corresponds, more or less, to the abrupt detection break due to the sensitivity threshold from LOLAS (Table 3, Chun et al., 2009). The evident huge vacuum zone between 600 m and 1 km above Mauna Kea simply means that the turbulence is not measured in this vertical slab. Conclusions: the optical turbulence vertical distribution appears very similar above Mt. Graham and Mauna Kea.

Using the HVR-GS it has been observed that the turbulence decreases, above astronomical sites in nightly stable conditions, in a much sharper way than what has been believed and predicted so far by models (such as the Hufnagel model). We discovered that 50% of the turbulence develops typically in the first 80 m +/- 15 m from the ground. The spatial structure of the turbulence as monitored by the HVR-GS appears concentrated in thin layers. It appears evident  that the higher the vertical resolution of the instrument, the sharper is the decreasing of the optical turbulence distribution with the height from the ground.


Fig. 6: (from Masciadri et al. 2010): Percentage of turbulence developed between the ground and the height h with respect to the turbulence developed in the whole atmosphere (~ 20 km) as retrieved from the HVR-GS measurements and extended to the first kilometre.

Acknowledgments: This work is funded by the Marie Curie Excellence Grant ForOT - MEXT-CT-2005-023878

E.Masciadri, 3/2010