Simulation of Qinghai Lake Water Level Fluctuations Using Machine Learning

Expand
  • Chengdu University of Information TechnologyChengdu 610225SichuanChina

Online published: 2025-09-23

Abstract

A comprehensive analysis was conducted to examine the processes and impacts of water level variations in Qinghai Lake under changing climatic conditions. The study utilized monthly mean water level data from 1959 to 2017sourced from the Qinghai Lake Basinin conjunction with meteorological and climate variables derived from the ERA5 reanalysis dataset developed by the European Centre for Medium-Range Weather Fore‐ castsECMWF. Several large-scale atmospheric circulation indices were also incorporated to investigate their influence on lake dynamics. This integrated dataset enabled a systematic assessment of the dominant climatic drivers and facilitated the development of predictive models to simulate future water level changes. To identify the most relevant influencing factorsthe Random ForestRFalgorithm was employed to perform feature selection and importance ranking. This process allowed for an evaluation of the relationship between feature relevance and model performance. Subsequentlya comparative analysis was undertaken using five machine learning modelsRFSupport Vector MachineSVM),Multi-layer PerceptronMLP),Long Short-Term MemoryLSTMnetworksand Multiple Linear RegressionMLR. The models were trained and validated to simulate monthly water level fluctuations and to assess the influence of model complexity and temporal learning ability on predictive accuracy. The analysis revealed that key drivers of Qinghai Lake water levels include the North Atlantic OscillationNAO),Atlantic Multidecadal OscillationAMO),Niño 3. 4 indexrelative humidity at 400 hPa450 hPaand 100 hPaRH400RH450RH100),precipitationtemperature at 1000 hPaT1000),vertical wind velocity at 1000 hPaW1000),and longwave radiationLW. Among the models testedthe LSTM network exhibited superior performance due to its ability to capture complex nonlinear and sequential dependencies in the data. When the ten most significant features were selectedthe LSTM model achieved a Pearson correlation coefficientRof 0. 95Nash-Sutcliffe EfficiencyNSEof 0. 96Normalized Root Mean Square ErrorNRMSEof 0. 14and Kling-Gupta EfficiencyKGEof 0. 87. The MLP model demonstrated the second-best performancewhile RF and SVM yielded comparable but slightly lower results. MLR performed the worstreflecting its limitations in modeling nonlinear and temporal relationships. Projections based on the LSTM model indicate that the water level of Qinghai Lake is likely to rise by approximately 2. 55 m between 2017 and 2030. This anticipated increase reflects the continuing influence of climate change and underscores the importance of adaptive water resource management strategies in plateau lake regions. The findings offer a reliable methodological frame‐ work for modeling and forecasting hydrological changes in alpine lake systems under future climate scenarios.

Cite this article

HUANG Jiawen, LONG Yinping, MA Qimin, XU Weixin, BIAN Yuxia, CHEN Hui, TAN Xiwen, LI Suowu . Simulation of Qinghai Lake Water Level Fluctuations Using Machine Learning[J]. Plateau Meteorology, 0 : 1 . DOI: 10.7522/j.issn.1000-0534.2025.00078

Outlines

/