5 Relative contribution of the different layers to the seeing at Maidanak

5.1 Surface layer

During night time the air near ground is subjected to a very strong radiative cooling because of the low humidity (Maidanak is located in a semi desert area). The high positive temperature gradient causes large temperature fluctuations in the first few meters above the ground.

Micro-thermal measurements in the surface layer at Maidanak were performed over several years (see summary in Gur92). The altitude dependence of $C_{\rm T}^2$ was found to be approximately as h^-2. However, the overall contribution of the surface layer to the total seeing degradation was found to be small.

As explained in Sect. 3, the configuration of the GSM at Mt. Maidanak when module 4 was installed at a lower height of 1.9 m and modules 1-3 at 3.4 m allowed us to measure directly the effect of the surface layer. The turbulence integral in the surface layer $J_{\rm SL}$,

$$J_{\rm SL} = \int_{1.9~{\rm m}}^{3.4~{\rm m}} C_{\rm n}^2(h) {\rm d}h,$$ (2)

was estimated from the difference of seeing measured by identical modules 3 and 4 with a typical precision of $1\ 10^{-13}~{\rm m}^{1/3}$. The relative contribution of the surface layer to the total $C_{\rm n}^2$ integral was typically less than 20%.

Using the $C_{\rm T}^{2} \propto h^{-2}$ model of Gur92, we have estimated the difference between seeing measured by DIMM and GSM due to their different heights above the ground. If 1.9-3.4 m layer contributes about 20% to the total seeing, then a 3.4-6 m layer contribution should be 10%. Hence, the DIMM-GSM systematic difference must be 0.9^3/5 = 0.94. It is exactly the ratio of the median seeing measured by both instruments (see Sect. 3.5).

However, at a site with good seeing like Maidanak, in some special circumstances, the local turbulence may become important. Depending on wind direction and velocity, surface layer may significantly contribute to seeing degradation, especially in a location like DIMM's which is not on the very top of the mountain, or in the vicinity of buildings. Local seeing effects were also observed at Paranal Martin et al.2000.

The most dramatic surface layer effect was found on July 16, 1998 (Fig. 17, top). Surface layer alone caused seeing degradation up to 1.4 $^{\prime\prime}$. This corresponded to the period of the worst seeing measured throughout this campaign. A more typical example of two nights with good seeing is shown in the lower plot of Fig. 17. Usually the surface layer contribution is barely apparent, its net effect being on the level of $1\ 10^{-13}\ {\rm m}^{1/3}$ (0.3 $^{\prime\prime}$ seeing) or less.

5.2 Boundary layer

The joint contribution of surface and boundary layers (up to 0.5 km) to the total $C_{\rm n}^2$ integral was estimated in Gur92 to be typically 12%. Now we know that in that paper the total $C_{\rm n}^2$integral was overestimated by a factor of 1.4-1.6, because of the bias in the optical seeing monitor used. This factor follows from comparison of our seeing data with the previous data Ilyasov et al.1999 and was already suspected in Gur92. It means that the boundary layer contribution must be increased to about 20%.

5.3 Free atmosphere

The $C_{\rm n}^2$ integral in the free atmosphere (roughly above 2 km) have been recently estimated in Kornilov et al.2000,Kornilov & Tokovinin2000 from the measurements of stellar scintillation with apertures of different diameters. These measurements were taken in 1998-99 on a total of 42 nights. The median free atmosphere seeing was estimated as 0.39 $^{\prime\prime}$. Comparing it to a median seeing of 0.69 $^{\prime\prime}$, free atmosphere contribution is about (0.39/0.69)^5/3=0.39. It was found that the degradation of the free atmosphere seeing was related to the appearance of strong turbulent layers at altitudes of $\approx 3$ km above the summit.

Looking at the wind profile (Fig. 15), we remark that above 2.2 km a wind shear appears, corresponding to the transition between boundary layer and the western planetary circulation in the free atmosphere. The Pamir mountain system to the east of Maidanak, about 6 km a.s.l. (3.4 km above summit) must be a factor defining this transition. So, the true boundary layer at Maidanak is up to 3 km, and not up to 0.5 km as assumed in Gur92. It means that the contribution of a boundary layer is somewhat higher, partly explaining the apparent discrepancy between Gur92 and Kornilov et al.2000.

Summarizing all available information, we suggest the following typical contribution of different layers to the total seeing: surface layer together with a boundary layer up to 3 km above summit provides some 70%, the remaining 30% being due to the free atmosphere.

These very preliminary considerations show the importance of the turbulence profile measurements for further characterization of Maidanak atmospheric physics. This provides a strong motivation for future site testing campaigns.