Monitoring Climate Change Impacts on Lake Van Using Machine Learning, Statistical Methods, and Remote Sensing Data


Erkoç M. H.

he third workshop of the Inter-Commission Committee on Geodesy for Climate Research (ICCC) of the International Association of Geodesy (IAG), 24 - 25 Mart 2025, ss.1, (Özet Bildiri)

  • Yayın Türü: Bildiri / Özet Bildiri
  • Sayfa Sayıları: ss.1
  • Yıldız Teknik Üniversitesi Adresli: Evet

Özet

The current research paper deals with the impact of climate change on Lake Van-the largest lake in Turkey-

between 1995 and 2024. In this study, variations in lake water level, temperature, and precipitation trends, SPEI-

12 drought indicators, and changes in surface area by using a traditional approach, such as the classical linear

model and Mann-Kendall test, together with a machine learning algorithm like LS-SVM, were analyzed. The

annual lake level decline was -4.2 cm and -3.9 cm according to in-situ measurements and HydroWeb satellite

altimetry data, respectively. Meteorological station data confirmed the increasing temperature and decreasing

precipitation trends that have been intensifying drought conditions in the region. SPEI-12 analyses revealed that

from the last decade onward, the hydrological drought has been in moderate to severe conditions. Landsat

satellite imagery analysis presented a loss of 0.0477 km²/yr in lake surface area. The result clearly presents the

ecological imbalance and the mismanagement of water resources caused by the change in climate over Lake Van.

The presented work points out the relevance of the application of machine learning techniques combined with

classic approaches in identifying climate change indicators and contributes effectively to the literature. The

findings emphasize that sustainable management of the lake and the ecosystem requires the identification of

relevant strategies.