Long-term water level variations in Lake Qinghai and their climatic and human activities driving mechanisms

Located in the northeastern part of the Tibetan Plateau, Lake Qinghai is China's largest lake, saltwater lake and inland lake. Since 2004, the water level of Lake Qinghai increased at an annual rate of 0.21 m, which has led to a rapid expansion of the lake's area. This study aims to investigate the hydrological variability process of Lake Qinghai Basin by integrating meteorological, hydrological, anthropogenic, and remote sensing data collected between 1960 and 2020. Statistical analysis methods and a hybrid model of long short-term memory (LSTM) – autoregressive integrated moving average (ARIMA) were used to identify the key drivers affecting lake water level fluctuations and to predict trends in lake water level changes and inundation areas in the coming decades. The research found that the water level of Lake Qinghai maintained the trend of decrease and then increase from 1956 to 2020. In particular, precipitation is the major cause of the water level change of Lake Qinghai, while the influence of human activities on the water level change of Lake Qinghai is relatively small. From 2004–2020, the trend of increase in the water level of Lake Qinghai was significant, and this period experienced significant warming and humidification, with a sharp increase in runoff and a drastic decrease in evapotranspiration. Using the LSTM – ARIMA hybrid prediction model, it was expected that the water level of Lake Qinghai would continue to rise in the next 30 years, and might exceed 3200 m by 2055. However, the rapid expansion of the lake area could potentially threaten the neighboring railroads, highways, and settlements, and trigger a series of ecological and environmental effects. Enhanced monitoring and research are essential to protect the ecological functions and socio-economic values of Lake Qinghai. • The interrelationships and trends of climate, hydrology and human activities in the Lake Qinghai Basin have been analyzed. • Factors affecting the water level change of Lake Qinghai have been revealed. • A hybrid LSTM–ARIMA model was applied to predict water level trends over the next 30 years.