Investigation of recent level changes in Lake Van using water balance, LSTM and ANN approaches

Abstract

Abstract Lake Van, the greatest soda water lake in the world located in the east of Turkey, has always attracted the attention of researchers due to its significant water level changes. Identifying the causes for these level changes is very important with respect to the scientific world and the sustainability of the ecological balance. Although extensive research has been carried out on the water level changes in Lake Van in the past years, no any study exists which analyzes the recent level changes. In this study, recent water level changes in Van Lake were analyzed using two different methods, and the results were evaluated. First, the water level changes during the period 2010–2020 were examined through the meteorological and hydrological data collected by the water budget method. Second, the water level changes for 2000–2020 were estimated using the LSTM and NAR neural networks. In the light of the results, reasons for the recent level changes in Lake Van were discussed. It was concluded that the changes, especially those occurred after 2014, may be hydrometeorological. However, the unusual level changes between 2010 and 2013 cannot be explained by water balance, may be tectonic or volcanic origin. The findings of this study indicate that these changes may continue in the future, and therefore, further research with different disciplines is needed on this subject.

Keywords

• Lake Van
• Level change
• Water budget
• LSTM-RNN
• NAR
• ANN