SCIDAR stands for SCIntillation Detection and Ranging
Purpose of the SCIDAR
The SCIDAR technique is aimed to measure the strength of the optical turbulence (
) and the
displacement velocity (v(h)) of the turbulent
layers as a function of height, 
.
History of the SCIDAR
The concept of the SCIDAR technique was proposed by Vernin & Roddier in 1973, followed by interesting developments during several years: Rocca, Roddier & Vernin 1974, Vernin & Roddier 1975, Azouit et al. 1978, Vernin et al. 1979, Azouit & Vernin 1980; Vernin & Azouit 1983, Caccia et al. 1987, and more recently by Fuchs, Tallon & Vernin (1994) who settled the basis for the Generalized SCIDAR, which was developed and tested by Avila, Vernin & Masciadri in 1997 and finally exploited by Avila, Vernin & Cuevas in 1998 and Klueckers et al. 1998. The first monitoring of the velocity profiles using a Generalized Scidar was published in 2001 by Avila, Vernin & Sánchez. A complete list of references related to Scidar can be found in References.The Classical SCIDAR concept
Let us suppose, for the sake of simplicity, a unique turbulent layer a hight
(Fig. 1a). A binary star (Star A, Star B) with angular separation 
projects on the pupil plane two scintillation patterns (intensity distributions)
separated by a distance 
.
and 
respectively.
The expression of the autocorrelation is:
Where
 |
 |
 |
with 
being the difference of
the stellar magnitudes. 
is the single-star autocorrelation that can be calculated theoretically. Notice
that one of the lateral peaks is enough for retrieving of
..
The profile
is retrieved from the autocorrelation using an inversion algorithm based on
maximum entropy.
The classical SCIDAR is insensitive to telescope aberrations.
The Generalized SCIDAR concept
The variance of the scintillation produced by a layer at altitude
is proportional to
.
For that reason, if images are taken at the pupil level, a layer at ground level
will not be detected. To overcome this limitation, in the Generalized SCIDAR
the plane of the detector is made the conjugate of a plane (analysis plane)
at a distance 
,
of the order of a few kilometers, below the telescope pupil. In this case the
distance relevant for scintillation produced by a turbulent layer at an altitude
is 
,
which makes the turbulence at ground level detectable (including that of the
telescope dome). The separation of the scintillation patterns projected on the
analysis plane by a double star is 
(Fig 1c), which is also the separation of the lateral peaks in the autocorrelation
(Fig 1d). The expression of the autocorrelation becomes in this case:
The image processing is identical to that applied in the Classical
SCIDAR case. The inversion algorithm delivers 
,
from which the actual profile is deduced.
Determination of the velocity displacement of the turbulent layers
Imagine again a single turbulent layer at an altitude ,
which is moving at a horizontal velocity 
.
Scintillation images, produced by a double star, are taken every 
,
in the Generalized configuration. The speckle patterns move a distance
from one image to the following. The mean cross-correlation of consecutive images
will be constituted of the same triplet as explained above, but displaced a
distance
from the correlation center (Fig. 3). As
a known parameter, the layer velocity is deduced straightforwardly.
In the realistic case of multiple layers, the position of each triplet gives the velocity of the corresponding layer. Sometimes the triplets are superimposed which dificults the data reduction.
