Graph-Based Deep Learning Model for Forecasting Chloride Concentration in Urban Streams to Protect Salt-Vulnerable Areas

Abstract

In cold-climate regions, road salt is used as a deicer for winter road maintenance. The applied road salt melts ice and snow on roads and can be washed off through storm sewer systems into nearby urban streams, harming the freshwater ecosystem. Therefore, aiming to develop a precise and accurate model to determine future chloride concentration in the Credit River in Ontario, Canada, the present work makes use of a “Graph Neural Network”–“Sample and Aggregate” (GNN-SAGE). The proposed GNN-SAGE is compared to other models, including a Deep Neural Network-based transformer (DNN-Transformer) and a benchmarking persistence model for a 6 h forecasting horizon. The proposed GNN-SAGE surpassed both the benchmarking persistence model and the DNN-Transformer model, achieving RMSE and R2 values of 51.16 ppb and 0.88, respectively. Additionally, a SHAP analysis provides insight into the variables that influence the model’s forecasting, showing the impact of the spatiotemporal neighboring data from the network and the seasonality variables on the model’s result. The GNN-SAGE model shows potential for use in the real-time forecasting of water quality in urban streams, aiding in the development of regulatory policies to protect vulnerable freshwater ecosystems in urban areas.

Uncertainty in Environmental Micropollutant Modeling

“…al. (2024) used GNN to predict pollution sources in a water distribution system and Santos et al (2023) use GNNs to predict chloride concentration in urban streams. To our knowledge, GNNs have not yet been applied on MP prediction, but they could be used in similar applications as LSTMs if the data is structured in a graph like format, like in rivers.…”

Chlorophyll-a Estimation in 149 Tropical Semi-Arid Reservoirs Using Remote Sensing Data and Six Machine Learning Methods

“…When used as inputs in a predictive model, highly correlated variables may introduce noise into the dataset, thereby increasing the model's variance and reducing its accuracy [86,87]. However, discarding variables solely based on a high or low correlation may also be detrimental, as they may still carry relevant spatiotemporal information that can improve forecasting performance [14].…”

“…Estimations of Chla concentrations from remote sensing data can be achieved using machine learning (ML). The ML approach retrieves complex non-linear relationships within satellite data by capturing the underlying structure connecting the satellite data and the desired target variable [14,15]. Combining ML architectures with remote sensing data has been used to successfully monitor Chla in inland and ocean waters.…”

Chlorophyll-a Estimation in 149 Tropical Semi-Arid Reservoirs Using Remote Sensing and Machine Learning

“…Highly correlated variables, when used as inputs on a predictive model, may add noise to the dataset, increasing the model's variance and thus reducing its performance [96,97]. However, discarding highly and low correlated variables altogether may also harm the model's performance since they may still carry relevant spatiotemporal information relating to the inputs' and outputs' attributes, leading to improved forecasting performance [30].…”

“…The ML approach retrieves complex non-linear relationships within satellite data, capturing the underlying structure bonding the satellite data and the desired target variable [30,31]. Combining ML architectures with remote sensing data has been able to provide top-notch results in a plethora of scientific fields, such as solar irradiance forecasting [32], mapping of mineral extraction sites [33], forest fire mapping [34], and crop water stress evaluation [35].…”