УДК 520.16

ESTIMATION OF THE ALTITUDE PROFILE OF ATMOSPHERIC OPTICAL TURBULENCE INTENSITY USING GRADIENT BOOSTING Estimation of the altitude profile of atmospheric optical turbulence intensity using gradient boosting

Опубликовано в Solar-Terrestrial Physics · Том 12, Номер 3 · Страницы 102–108 · Рубрика: Results of current research
DOI: https://doi.org/10.12737/stp-123202612 · EDN: PPYMVW
Получено: 13.11.2025 Одобрено: 19.03.2026 Опубликовано: 19.09.2026 Язык публикаций: ENG
Wavefront distortions caused by atmospheric turbulence significantly reduce the quality of astronomical observations. A key characteristic of atmospheric turbulence intensity is the structural constant of the refractive index Cₙ². Direct measurements of its altitude profile are complex and labor-intensive, which stimulates the development of forecasting methods. In this work, we present a method for estimating the altitude profile of Cₙ² at 12 levels (from 0.5 to 22.6 km) using machine learning. The gradient boosting model was trained based on ground measurements of turbulence characteristics, local meteorological data, and vertical profiles of meteorological variables extracted from the ERA-5 reanalysis. The performance of the resulting model was compared with that of a random forest model. The comparison showed that gradient boosting outperforms the random forest method at most heights, starting from 0.71 km. The greatest improvement in accuracy (up to 7.4 % according to Pearson's correlation coefficient) was achieved at heights above 5 km. Analysis of the importance of features for the best model revealed that the most significant parameters are temperature and wind speed at isobaric surfaces. The results demonstrate the effectiveness of gradient boosting for predicting the Cₙ² profile, which simplifies and speeds up the evaluation of astronomical sites where optical telescopes are located.
atmospheric optical turbulence, refractive index structure constant, machine learning, gradient boosting, random forest, astronomical sites
Финансирование
The research was financially supported by the Russian Science Foundation (Grant No. 24-72-10043 [https://rscf.ru/project/24-72-10043/])
Текст Список литературы
Текст (EN) (PDF)
Читать Скачать

1. Baki H., Basu S. Estimating high-resolution profiles of wind speeds from a global reanalysis dataset using TabNet. Environmental Data Science. 2024, vol. 3, p. e32. https://doi.org/10.1017/eds.2024.41.

2. Cai J., Xu K., Zhu Y., et al. Prediction and analysis of net ecosystem carbon exchange based on gradient boosting regression and random forest. Applied Energy. 2020, vol. 262, 114566. https://doi.org/10.1016/j.apenergy.2020.114566.

3. Cuadros E.T., Almeida A.P., Carlesso F., et al. Assessment of daytime sky quality and atmospheric refractive index (Cn2) analysis at Pico dos Dias Observatory. Astrophys. J. Suppl. Ser. 2025, vol. 281, 24. https://doi.org/10.3847/1538-4365/ae0a2a.

4. Jellen C., Nelson C., Burkhardt J., et al. Hybrid optical turbulence models using machine-learning and local measurements. Applied Optics. 2023, vol. 62, no. 18, pp. 4880–4890. https://doi.org/10.1364/AO.487280.

5. Kornilov V., Tokovinin A., Shatsky N., et al. Combined MASS-DIMM instrument for atmospheric turbulence studies. Mon. Not. Roy. Astron. Soc. 2007, vol. 382, pp. 1268–1278. https://doi.org/10.1111/j.1365-2966.2007.12467.x.

6. Natekin A., Knoll A. Gradient boosting machines, a tutorial. Front. Neurorobot. 2013, vol. 7, 21. https://doi.org/10.3389/fnbot.2013.00021.

7. Pierzyna M., Basu S., Saathof R. OTCliM: Generating a near-surface climatology of optical turbulence strength (Cn2) using gradient boosting. Artificial Intelligence for the Earth Systems. 2025, vol. 4, no. 2, e240076. https://doi.org/10.1175/AIES-D-24-0076.1.

8. Roddier F.V The effects of atmospheric turbulence in optical astronomy. Progress in Optics / Ed. E. Wolf. Amsterdam, Elsevier, 1981, vol. 19, pp. 281–376. https://doi.org/10.1016/S0079-6638(08)70204-X.

9. Shikhovtsev A.Yu., Kovadlo P.G., Kiselev A.V., et al. Application of neural networks to estimation and prediction of seeing at the Large Solar Telescope site. Publ. Astron. Soc. Pacific. 2023, vol. 135, no. 1043, 014503. https://doi.org/10.1088/1538-3873/acb384.

10. Shikhovtsev A.Yu., Driga M.B., Kiselev A.V., Kovadlo P.G., Eselevich M.V. Neural network models for estimating the atmospheric angular resolution of ground-based telescopes. Proc. All-Russian Conf. with Int. Particip. “Magnetism and Solar Activity – 2025”. Nauchnyi, Crimea, 2025a, p. 1. (In Russian).

11. Shikhovtsev A.Yu., Kiselev A.V., Kovadlo P.G. Physically informed neural networks for assessing atmospheric optical turbulence intensity. Izvestiya vuzov. Fizika. 2025b, vol. 68, iss. 5, pp. 66–73. (In Russian). https://doi.org/10.17223/00213411/68/5/8.

12. URL: https://rscf.ru/project/24-72-10043/ (accessed June 22, 2026).

13. URL: https://ckp-rf.ru/catalog/usu/200615/ (accessed June 22, 2026).

Войти или Создать
* Забыли пароль?