22- Application of aggregation operators for forecasting PM10 fluctuations : From available Caribbean data sites to unequipped ones [Plocoste et al (2024)]

Air pollution is a major issue for public health. Predicting levels of airborne particles, particularly those of natural origin like African sand mists, enables more effective warnings for affected populations. Currently, only three islands continuously measure PM10 concentrations in the Caribbean area. Thus, there is a significant issue in forecasting these particulate pollutants for unequipped sites. In this article, a new approach based on the use of aggregation operators in a non-fuzzy framework allows for predicting the quantity of PM10 for certain sites in this geographical area. It is the characteristics of non-linearity and downward reinforcement of the fuzzy operators used (triple 𝑃𝑖 and 𝑡-norms) that allow modeling the evolution of a phenomenon as complex as sand mists. With time series normalized, tests were conducted on two of the three sites (Martinique and Guadeloupe) to predict values for the third site (Puerto Rico) and evaluate prediction quality. The performance metrics show that this approach yields significant results. The seasonality of desert dust has a significant impact on these performance measures.