Abstract:Depletion of dissolved oxygen (DO) is a significant incentive for biological catastrophic events in freshwater lakes. Although predicting the DO concentrations in lakes with high-frequency real-time data to prevent hypoxic events is effective, few related experimental studies were made. In this study, a short-term predicting model was developed for DO concentrations in three problematic areas in China's Chaohu Lake. To predict the DO concentrations at these representative sites, which coincide with biological abnormal death areas, water quality indicators at the three sampling sites and hydrometeorological features were adopted as input variables. The monitoring data were collected every 4 h between 2020 and 2023 and applied separately to train and test the model at a ratio of 8:2. A new AC-BiLSTM coupling model of the convolution neural network (CNN) and the bidirectional long short-term memory (BiLSTM) with the attention mechanism (AM) was proposed to tackle characteristics of discontinuous dynamic change of DO concentrations in long time series. Compared with the BiLSTM and CNN-BiLSTM models, the AC-BiLSTM showed better performance in the evaluation criteria of MSE, MAE, and R2 and a stronger ability to capture global dependency relationships. Although the prediction accuracy of hypoxic events was slightly worse, the general time series characteristics of abrupt DO depletion were captured. Water temperature regularly affects DO concentrations due to its periodic variations. The high correlation and the universal importance of total nitrogen (TN) and total phosphorus (TP) with DO reveals that point source pollution are critical cause of DO depletion in the freshwater lake. The importance of NTU at the Zhong Miao Station indicates the self-purification capacity of the lake is affected by the flow rate changes brought by the tributaries. Calculating linear correlations of variables in conjunction with a permutation variable importance analysis enhanced the interpretability of the proposed model results. This study demonstrates that the AC-BiLSTM model can complete the task of short-term prediction of DO concentration of lakes and reveal its response features of timing and magnitude of abrupt DO depletion.

Enhanced Prediction of Dissolved Oxygen Concentration using a Hybrid Deep Learning Approach with Sinusoidal Geometric Mode Decomposition

Enhanced predictive modeling of dissolved oxygen concentrations in riverine systems using novel hybrid temporal pattern attention deep neural networks

A deep learning interpretable model for river dissolved oxygen multi-step and interval prediction based on multi-source data fusion

Temporal prediction of dissolved oxygen based on CEEMDAN and multi-strategy LSTM hybrid model

From Hydrometeorology to River Water Quality: Can a Deep Learning Model Predict Dissolved Oxygen at the Continental Scale?

Prediction of riverine daily minimum dissolved oxygen concentrations using hybrid deep learning and routine hydrometeorological data

LSTM-TCN: dissolved oxygen prediction in aquaculture, based on combined model of long short-term memory network and temporal convolutional network

A hybrid prediction model of dissolved oxygen concentration based on secondary decomposition and bidirectional gate recurrent unit

Satellite-Based Global Sea Surface Oxygen Mapping and Interpretation with Spatiotemporal Machine Learning.

A hybrid intelligence model for predicting dissolved oxygen in aquaculture water

Multi-step ahead dissolved oxygen concentration prediction based on knowledge guided ensemble learning and explainable artificial intelligence

Predicting abrupt depletion of dissolved oxygen in Chaohu lake using CNN-BiLSTM with improved attention mechanism

Spatiotemporal-aware machine learning approaches for dissolved oxygen prediction in coastal waters

An ensembled method for predicting dissolved oxygen level in aquaculture environment

A hybrid XGBoost-ISSA-LSTM model for accurate short-term and long-term dissolved oxygen prediction in ponds

A hybrid water quality prediction model based on variational mode decomposition and bidirectional gated recursive unit

Comparison of Forecasting Models for Real-Time Monitoring of Water Quality Parameters Based on Hybrid Deep Learning Neural Networks

Dissolved oxygen concentration inversion based on Himawari-8 data and deep learning: a case study of lake Taihu

Enhancing Dissolved Oxygen Concentrations Prediction in Water Bodies: A Temporal Transformer Approach with Multi-Site Meteorological Data Graph Embedding

Dissolved Oxygen Prediction Based on PCA-LSTM

Water Quality Prediction Based on Hybrid Deep Learning Algorithm