Journal of Environmental Engineering Technology•2026•DOI: 10.13205/j.hjgc.202604001
This study systematically investigated the occurrence, spatial distribution, sources, and ecological risks of 160 pesticides in Dianchi Lake, a typical plateau lake impacted by agricultural activities. A total of 37 pesticides were detected in the water, with total concentrations ranging from 64.2 to 1132.8 ng/L (average 610.0 ng/L). Fungicides, including boscalid (BOS), fluopicolide (FPC), and dimethomorph (DMM), were dominant, contributing up to 65.0% of the total concentration. Spatially, the southern lake region exhibited significantly higher concentrations (672.5 ng/L) than the north, attributed to intensive facility agriculture. Highly hydrophobic pesticides, such as penconazole (PEN), showed a tendency to enrich in bottom layers. Source apportionment identified inflowing rivers and wastewater treatment plant effluents as primary input sources, with average concentrations 7 and 9 times higher than lake water, respectively. Ecological risk assessment revealed that pesticides posed the highest risk to algae, followed by daphnia and fish. Prometryn (PMT) was identified as a high-risk factor for algae, while profenofos (PFF) and carbendazim (CBD) posed potential threats to higher trophic levels. These findings provide fundamental data and technical support for understanding pesticide pollution in plateau lake ecosystems.
Journal of Environmental Engineering Technology•2026•DOI: 10.13205/j.hjgc.202607021
The continuous expansion of urban sewage treatment capacity has led to a sustained increase in sludge generation, making efficient treatment, disposal, and resource recovery critical in environmental engineering. Machine learning (ML) offers substantial potential for prediction and optimization in sludge treatment by extracting non-linear features from complex operational data. This review systematically examines the application of ML across typical sludge treatment processes, including dewatering, resource recovery (anaerobic digestion), and terminal disposal (incineration and landfill). The general modeling workflow is summarized across three dimensions: dataset preparation, algorithm selection, and model evaluation. A comparative analysis evaluates the applicability and limitations of support vector machines (SVM), random forests (RF), artificial neural networks (ANN), and other deep learning models. SVMs demonstrate greater stability with small-to-medium sample sizes and high-dimensional data, while RFs exhibit strong generalization and provide variable importance insights. ANNs and deep learning models excel in large-scale data and time-series or image tasks but require high data quality. Key findings from the literature include ANN achieving R²=0.99 and RMSE=0.02 in dewatering prediction, and R²=0.86 with NRMSE=0.31 in anaerobic digestion, while gradient boosting reached R²=0.90 and RMSE=0.33. Future directions emphasize multi-source data fusion, model interpretability (e.g., SHAP), and coupling ML with mechanistic models to enhance predictive accuracy and generalization, supporting intelligent and refined sludge treatment management.
Journal of Environmental Engineering Technology•2026•DOI: 10.13205/j.hjgc.202607022
Anaerobic sludge digestion is the core process for achieving energy recovery and sludge reduction in wastewater treatment plants. However, its complex biological reaction mechanisms and multivariable coupling characteristics pose persistent challenges to process optimization and stable control. Traditional mechanistic models, while theoretically clear, suffer from parameter calibration difficulties and insufficient adaptability under dynamic and nonlinear conditions. Machine learning (ML) has gained attention for its powerful data modeling capabilities. This review systematically examines ML applications in sludge anaerobic digestion, focusing on biogas production prediction, process monitoring and early warning, and process parameter optimization. For gas production, hybrid models and deep learning achieve high-precision methane yield predictions. Soft-sensing models using easy-to-measure parameters enable real-time estimation of volatile fatty acids and total ammonia nitrogen. At the optimization level, coupling surrogate models with optimization algorithms provides dynamic regulation strategies for co-digestion ratios and pretreatment conditions. Interpretable methods address the 'black-box' issue, enhancing engineering acceptability. Deep integration of these methods with dynamic optimization supports an intelligent decision-making framework. However, translation from laboratory to engineering faces constraints including data quality, model generalization, and implementation. This paper provides an analytical framework combining predictive capability with engineering reliability for sludge treatment optimization.