Zhengshan Luo

This study addressed the limitations of traditional methods in predicting air pollution dispersion, which include restrictions in handling spatiotemporal dynamics, unbalanced feature importance, and data scarcity. To overcome these challenges, this research introduces a novel deep learning-based model, SAResNet-TCN, which integrates the strengths of a Residual Neural Network (ResNet) and a Temporal Convolutional Network (TCN). This fusion is designed to effectively capture the spatiotemporal characteristics and temporal correlations within pollutant dispersion data. The incorporation of a sparse attention (SA) mechanism further refines the model’s focus on critical information, thereby improving efficiency. Furthermore, this study employed a Time-Series Generative Adversarial Network (TimeGAN) to augment the dataset, thereby improving the generalisability of the model. In rigorous ablation and comparison experiments, the SAResNet-TCN model demonstrated significant advances in predicting pollutant dispersion patterns, including accurate predictions of concentration peaks and trends. These results were enhanced by a global sensitivity analysis (GSA) and an additive-by-addition approach, which identified the optimal combination of input variables for different scenarios by examining their impact on the model’s performance. This study also included visual representations of the maximum downwind hazardous distance (MDH-distance) for pollutants, validated against the Prairie Grass Project Release 31, with the Protective Action Criteria (PAC) and Immediately Dangerous to Life or Health (IDLH) levels serving as hazard thresholds. This comprehensive approach to contaminant dispersion prediction aims to provide an innovative and practical solution for environmental hazard prediction and management.

Some natural gases are toxic because they contain hydrogen sulfide (H₂S). The solubility pattern of elemental sulfur (S) in toxic natural gas needs to be studied for environmental protection and life safety. Some methods (e.g., experiments) may pose safety risks. Measuring sulfur solubility using a machine learning (ML) method is fast and accurate. Considering the limited experimental data on sulfur solubility, this study used consensus nested cross-validation (cnCV) to obtain more information. The global search capability and learning efficiency of random forest (RF) and weighted least squares support vector machine (WLSSVM) models were enhanced via a whale optimization–genetic algorithm (WOA-GA). Hence, the WOA-GA-RF and WOA-GA-WLSSVM models were developed to accurately predict the solubility of sulfur and reveal its variation pattern. WOA-GA-RF outperformed six other similar models (e.g., RF model) and six other published studies (e.g., the model designed by Roberts et al.). Using the generic positional oligomer importance matrix (gPOIM), this study visualized the contribution of variables affecting sulfur solubility. The results show that temperature, pressure, and H₂S content all have positive effects on sulfur solubility. Sulfur solubility significantly increases when the H₂S content exceeds 10%, and other conditions (temperature, pressure) remain the same.