Abstract:Deep learning methods, which have strong capabilities for mapping highly nonlinear relationships with acceptable calculation speed, have been increasingly applied for water quality prediction in recent studies. However, it is argued that the practicality of deep learning methods is limited due to the lack of physical mechanics to explain the prediction results of water quality changes. A knowledge gap exists in rationalizing the deep learning results for water quality predictions. To address this gap, an interpretable deep learning framework was established to predict the spatiotemporal variations of water quality parameters in a large spatial region. Mereological, land-use, and socioeconomic variables were adopted to predict the daily variations of stream water quality parameters across 138 sub-catchments in a total of over 575,250 km2 in southern China. The coefficients of determination of chemical oxygen demand (COD), total phosphorus (TP), and total nitrogen (TN) predictions were over 0.80, suggesting a satisfactory prediction performance. The model performance in terms of prediction accuracy could be improved by involving land-use and socioeconomic predictors in addition to hydrological variables. The SHapley Additive exPlanations method used in this study was demonstrated to be effective for interpreting the prediction results by identifying the significant variables and reasoning their influencing directions on the variation of each water quality parameter. The air temperature, proportion of forest area, grain production, population density, and proportion of urban area in each sub-catchment as well as the accumulated rainfall within the previous 3 days were identified as the most significant variables affecting the variations of dissolved oxygen, COD, ammoniacal nitrogen(NH3-N), TN, TP, and turbidity in the stream water in the case area, respectively.

DLUP: A Deep Learning Utility Prediction Scheme for Solid-State Fermentation Services in IIoT

IUP: An Intelligent Utility Prediction Scheme for Solid-State Fermentation in 5G IoT

Daily water supply prediction method based on integrated learning and deep learning

Uncertainty Quantification and Interval Prediction of Equipment Remaining Useful Life Based on Semisupervised Learning

Lithium-ion Battery Remaining Useful Life Prediction: a Federated Learning-Based Approach

Data-Driven Soft Sensor Model Based on Deep Learning for Quality Prediction of Industrial Processes

Machine learning regression-based prediction model for the autonomous control of coagulant dosing in smart water purification plants

Deep-Learning-Based Approach for Prediction of Algal Blooms

Profitability Related Industrial-Scale Batch Processes Monitoring via Deep Learning based Soft Sensor Development

Integrating water quality and streamflow into prediction of chemical dosage in a drinking water treatment plant using machine learning algorithms

Prediction Method of Biological Fermentation Data Based on Deep Neural Network

Electronic nose signals-based deep learning models to realize high-precision monitoring of simultaneous saccharification and fermentation of cassava

A Water Quality Prediction Method Based on the Deep LSTM Network Considering Correlation in Smart Mariculture

Probabilistic Bayesian Deep Learning Approach for Online Forecasting of Fed-Batch Fermentation

Prediction of intensified ethanol fermentation of sugarcane using a deep learning soft sensor and process analytical technology

Enhancing Fermentation Process Monitoring through Data-Driven Modeling and Synthetic Time Series Generation

LSTM Autoencoder-based Deep Neural Networks for Barley Genotype-to-Phenotype Prediction

A hybrid spatial-temporal deep learning prediction model of industrial methanol-to-olefins process

Smart Soft Sensor Design with Hierarchical Sampling Strategy of Ensemble Gaussian Process Regression for Fermentation Processes

Large-Scale Prediction of Stream Water Quality Using an Interpretable Deep Learning Approach

Intelligent Aeration Amount Prediction Control for Wastewater Treatment Process Based on Recurrent Neural Network