NEURAL NETWORK METHODS FOR PROCESSING SATELLITE IMAGERY FOR GEO-ENVIRONMENTAL MONITORING AND NATURAL HAZARD ASSESSMENT

Authors

  • Kanetova D.E., Madalieva Z.Zh., Aydarlieva A.I. Jalal-Abad Scientific Center of the National Academy of Sciences of the Kyrgyz Republic

Keywords:

neural networks, satellite imagery, geo-environmental monitoring, natural hazards, deep learning, landslides, floods

Abstract

The paper provides a review of modern neural network methods for processing satellite imagery applied to geo-environmental monitoring and natural hazard assessment. Deep learning architectures are considered, including convolutional neural networks, segmentation models (U-Net, DeepLab), recurrent networks, and transformers. Particular attention is paid to the application of these methods for landslide mapping, soil erosion assessment, flood monitoring, and water quality analysis. International experience is analyzed, and the prospects for applying neural network technologies in the mountainous regions of Kyrgyzstan are highlighted. The advantages and limitations of the approaches are outlined, and directions for further research are suggested in the context of developing a national system for environmental monitoring and disaster risk reduction.

 

Author Biography

  • Kanetova D.E., Madalieva Z.Zh., Aydarlieva A.I., Jalal-Abad Scientific Center of the National Academy of Sciences of the Kyrgyz Republic

     

     

References

Khasanov S.; Juliev M.; Uzbekov U.; др. Landslides in Central Asia: a review... DOI: 10.2478/geosc-2021-0009.

Wang X.; Li X.; Ozturk U.; др. Atmospheric triggering conditions... DOI: 10.5194/nhess-21-2125-2021.

Saidaliyeva Z.; др. Adaptation to climate change... DOI: 10.1002/wcc.891.

Caleca F.; др. Regional-scale landslide risk assessment... DOI: 10.5194/nhess-24-13-2024.

Kattenborn T.; Leitloff J.; Schiefer F.; Hinz S. Review on CNN... DOI: 10.1016/j.isprsjprs.2020.12.010.

Li J.; др. A review of remote sensing image segmentation... DOI:10.1080/17538947.2023.2278277.

Wang R.; др. Transformers for Remote Sensing... DOI: 10.3390/s24113495.

Zhang Q.; др. Deep Learning for Exploring Landslides... DOI: 10.3390/rs16081344.

Bentivoglio R.; Isufi E.; Jonkman S. N.; Taormina R. Deep learning methods for flood mapping.DOI: 10.5194/hess-26-4345-2022.

Toma A.; др. Flooded area detection... DOI: 10.1080/22797254.2024.2351234.

Yang Z.; др. Exploring deep learning for landslide mapping... DOI: 10.13203/j.whugis20230401.

Aleissaee A. A.; др. Transformers in Remote Sensing... DOI: 10.3390/rs15071860.

Márquez-Grajales A.; др. Characterizing drought prediction... DOI: 10.1088/2752-5295/ad3f2c.

Poggi V.; др. Development of a regional probabilistic seismic hazard model... DOI: 10.5194/nhess-25-817-2025.

Scaini C.; др. A new regionally consistent exposure database... DOI: 10.5194/nhess-24-929-2024.

Chen H.; др. A landslide extraction method with channel attention... DOI: 10.1080/17538947.2022.2158720.

Gallear J. W.; др. Evaluation of ML approaches for drought monitoring... DOI: 10.5194/nhess-25-1521-2025.

Downloads

Published

2026-02-07

How to Cite

NEURAL NETWORK METHODS FOR PROCESSING SATELLITE IMAGERY FOR GEO-ENVIRONMENTAL MONITORING AND NATURAL HAZARD ASSESSMENT. (2026). PROCEEDINGS OF THE NATIONAL ACADEMY OF SCIENCES OF THE KYRGYZ REPUBLIC, 9, 85-89. https://journal.uia.gov.kg/index.php/main/article/view/1003