Fuzzy Logic for Environmental Risk Assessment: Modelling Uncertainty in Multi-Parameter Conditions

Abstract

Environmental risk assessment requires the interpretation of complex, uncertain, and nonlinear data from multiple environmental parameters. Conventional threshold-based methods often fail to represent gradual transitions in environmental conditions, leading to oversimplified classifications and delayed early warnings. This study aims to develop a flexible and interpretable environmental risk assessment framework using fuzzy logic, specifically the Sugeno inference system, to address uncertainty and imprecision in environmental data. The proposed model incorporates methane concentration, temperature, humidity, and particulate matter (PM2.5) as linguistic input variables represented through fuzzy membership functions. A rule-based inference mechanism is applied to integrate and generate a numerical environmental risk index associated with low, moderate, or high risk levels. The model is evaluated in a controlled laboratory environment using synthetic sensor data that simulate varying environmental conditions. Experimental results demonstrate that the Sugeno-based fuzzy system produces smooth risk transitions, avoids rigid decision boundaries, and responds effectively to overlapping parameter scenarios. Visualization of membership functions and rule evaluation pathways further enhances system transparency and interpretability. The proposed approach provides a scalable and explainable alternative to conventional environmental risk assessment methods and shows strong potential for real-time early warning applications in landfill gas monitoring, industrial air quality control, and urban environmental management