In this study, various regression models were utilized to predict total sediment yield in tons, while their performance was evaluated for accuracy and reliability. The dataset utilized contains numerous predictors that have been standardized and processed through principal component analysis to improve model performance. Models evaluated here include linear regression, normalized linear regression, Principal Component Analysis (PCA), Pearson Correlation Coefficient (PCC) with generalized ridge regression, kernel ridge regression, multivariate regression, lasso regression approaches such as artificial neural network Cellular Automata-Artificial Neural Network (CA-ANN or ANN), and more. Results suggest that the ANN model achieved the lowest mean squared error (MSE), 113.641; this suggests superior predictive capability compared to other models. Although environmental data were complex and relationships complex, an ANN model showed less error, followed closely by CA-ANN with an MSE of 124.83. Traditional models such as linear or lasso regression revealed larger errors with negative squared values that indicated poor fits to data. This analysis highlights the effectiveness of advanced machine learning in environmental modeling, emphasizing the importance of selecting models suited to data and specific phenomena, aiding environmental planners in predicting and managing soil erosion and sediment transport.
5 June 2025 in Iranian Journal of Science and Technology, Transactions of Civil Engineering
This study offers a detailed case-based evaluation of a laboratory-scale vertical flow constructed wetland (VFCW) system treating aquaculture wastewater characterized by a controlled effluent matrix and carefully managed hydraulic retention times (HRTs). Unlike more generalized constructed wetlands (CWs) applications, our research integrates specific operational parameters including vertical flow design, a 360 L treatment capacity, and a hydraulic loading rate of 8.2 L/d/m2 to systematically assess the removal of critical pollutants such as nitrate (N), phosphate (P), hydrogen sulfide (H2S), biochemical oxygen demand (BOD), and potassium (K+). The significant pollutants reductions achieved N by 73.9%, P by 96.3%, and BOD by over 90% demonstrate the system's high efficacy and validate its practical potential as a climate-resilient, nature-based treatment solution tailored for sustainable small-holder aquaculture systems. Indeed, our work emphasizes the quantification of multi-parameter attenuation under five distinct HRTs conditions during a short-term experimental trial, offering valuable, actionable insights for optimizing VFCW configurations adapted to aquaculture wastewater characteristics. This focused approach, addressing a specific aquaculture effluent matrix with operational conditions tailored to laboratory-scale treatment systems, clearly differentiates our contribution as an incremental yet valuable advancement. Moreover, it provides practical, real-world context insights and design implications for climate-resilient, nature-based wastewater treatment solutions that are well-suited to support sustainable aquaculture practices. We thereby explicitly acknowledge the foundation established by extensive prior research while delineating our study's unique contextual and operational contributions within the constructed wetland domain.