In this article, we present the results of astronomical seeing measurements conducted at the Maidanak observatory and the influence of meteorological parameters on astronomical seeing. A differential image motion monitor (DIMM) was used to measure astronomical seeing. Astronomical seeing observations were made between 2018–2022, mainly in the summer and autumn months. A total of 266 night observations were conducted between 2018 and 2022. For the whole time frame, the median seeing value was 0.69 arc seconds. With a median value of 0.65″, 2019 is the year with the best seeing, and 2021 has the worst seeing (0.71″). Analysis of long-term variations of the amount of clear time revealed a small trend of decrease in the amount. The correlation between the meteorological parameters and seeing was analyzed. It was determined that the meteorological parameters with the greatest impact on the seeing are wind speed and nighttime temperature variation. The correlation coefficient was equal to 13% and 21% for both quantities. It was found that wind direction and temperature have a relatively weak correlation with astronomical seeing.

Knowledge of the turbulence spectra is of interest for describing atmospheric conditions as applied to astronomical observations. This article discusses the deformations of the turbulence spectra with heights in a wide range of scales at the sites of the Maidanak and Suffa observatories. It is shown that the energy of baroclinic instability is high at the sites of these observatories and should be taken into account in the calculations of the refractive index structure constant $C_n^2$.

In the present article, we study the possibilities of machine learning for the estimation of seeing at the Maidanak Astronomical Observatory (38${}^{\circ }$40${}^{\prime }$24${}^{″}$ N, 66${}^{\circ }$53${}^{\prime }$47${}^{″}$ E) using only Era-5 reanalysis data. Seeing is usually associated with the integral of the turbulence strength $C_n^2\left(z\right)$ over the height z. Based on the seeing measurements accumulated over 13 years, we created ensemble models of multi-layer neural networks under the machine learning framework, including training and validation. For the first time in the world, we have simulated optical turbulence (seeing variations) during night-time with deep neural networks trained on a 13-year database of astronomical seeing. A set of neural networks for simulations of night-time seeing variations was obtained. For these neural networks, the linear correlation coefficient ranges from 0.48 to 0.68. We show that modeled seeing with neural networks is well-described through meteorological parameters, which include wind-speed components, air temperature, humidity, and turbulent surface stresses. One of the fundamental new results is that the structure of small-scale (optical) turbulence over the Maidanak Astronomical Observatory does not depend or depends negligibly on the large-scale vortex component of atmospheric flows.

Results of night-time surface layer turbulence measurements using ESO LuSci at the Maidanak observatory are presented. Turbulence in the surface layer was evaluated at the Maidanak observatory using LuSci during the period of 2021–2022. The overall median DIMM-seeing was 0.74 arcseconds during this period. It was determined that the seeing in the surface layer up to 256 m equals 0.44 arcseconds. This corresponds to 42% of the contribution to the integral seeing in the entire atmosphere. A telescope installed at 6 m above the ground will be affected by 33% of the integral turbulence and about 79% of the surface layer one. Taking into account that the free atmosphere contributes around 30%, we conclude that around 30% of the overall seeing is contributed by the boundary layer above the Maidanak observatory.

In the present article, we study the possibilities of machine learning for estimation of seeing at the Maidanak Astronomical Observatory (38∘40′24′′N, 66∘53′47′′E) using only Era-5 re-analysis data. Seeing is usually associated with the integral of the turbulence strength Cn2(z) over the height z. Based on the seeing measurements accumulated over 13 years we created ensemble models of multi-layer neural networks under the machine learning framework, including training and validation. For the first time in the world, we have simulated optical turbulence (seeing variations) during night-time with deep neural networks trained on a 13-year database of astronomical seeing. A set of neural networks for simulations of night-time seeing variations have been obtained. For these neural networks the linear correlation coefficient ranges from 0.48 to 0.68. We show that modeled seeing with neural networks is well described through meteorological parameters, which include wind speed components, air temperature, humidity and turbulent surface stresses. One of the fundamental new results is that the structure of small-scale (optical) turbulence over the Maidanak Astronomical Observatory does not depend or depends negligibly on the vortex component of atmospheric flows.