Evolution trends of water quality in inflow rivers of Miyun Reservoir from 2000 to 2020

Abstract

[Objective] High water quality ensures the survival and reproduction of aquatic organisms and maintains biodiversity. Variations in the water quality of inflow rivers of the Miyun Reservoir are closely related to the ecological protection of its upstream watershed. [Methods] Water quality monitoring data from 2000 to 2020 at Daguan Bridge Station on the Bai River and Xinzhuang Bridge Station on the Chao River located upstream of the Miyun Reservoir(Beijing's drinking water source) were used. The Seasonal Mann-Kendall(S-MK) trend analysis, Spearman correlation analysis, and Long Short-Term Memory(LSTM) prediction model were applied to reveal the characteristics and evolution trends of water quality variations in the inflow rivers of the Miyun Reservoir. [Results] The result indicated that total nitrogen at Daguan Bridge increased year by year from a minimum of 1.85 mg/L to 2.09 mg/L, and at Xinzhuang Bridge from 4.89 mg/L to 5.93 mg/L. From 2000 to 2020, total nitrogen failed to meet Class Ⅱ standards and was the main influencing factor of the water quality variations in the inflow rivers of the Miyun Reservoir. Specifically, population growth and GDP contributed over 90% to water quality variations, while precipitation, temperature, and evapotranspiration contributed approximately 9%, indicating human factors as the primary drivers of water quality. LSTM predictions showed that in the next ten years, the water quality of inflow rivers of the Miyun Reservoir would generally remain within Class Ⅱ-Ⅲ standards. Total nitrogen and nitrate nitrogen could reach a maximum value of 8.36 mg/L in certain months, showing significant fluctuations. Seasonal variations in water quality required particular attention. [Conclusion] By analyzing water quality variations, exploring driving factors, and predicting future water quality trends, the findings provide references for water resource management in the watershed of Miyun Reservoir and for water quality protection research in rivers in northern China.

Keywords

water quality / trend analysis / correlation analysis / driving factors / LSTM prediction model / Miyun Reservoir