Air Quality Assessment and Source Speciation of PM2.5 in Florida’s Everglades Agricultural Area

Abstract Fine particulate matter (PM2.5) poses a significant environmental and public health challenge, especially in regions impacted by agricultural biomass burning. This detailed study investigates the spatial and temporal variations, chemical composition, and source attribution of PM2.5 in Florida’s Everglades Agricultural Area (EAA) during the 2024 preharvest sugarcane burning season. Using high-resolution monitoring data from six strategically placed sites, along with Scanning Electron Microscopy with Energy Dispersive X-ray Spectroscopy (SEM-EDS) for particle analysis and HYSPLIT trajectory modeling to track pollutant movement, the primary sources of PM2.5 were identified. Results indicated a dominant presence of Saharan dust from across borders, comprising approximately 85% of PM2.5, while sugarcane burning ash made up about 5%. Temporal analysis showed a significant peak in PM2.5 levels on May 14, associated with specific weather patterns and long-range pollutant transport from distant sources. Advanced machine learning techniques, particularly Gradient Boosting Regression, demonstrated excellent predictive performance with an R² of 0.93. These findings underscore the complex behavior of PM2.5 and reveal that regional and transboundary influences have a greater impact on air quality than local agricultural burning. The study emphasizes the need for integrated monitoring systems and advanced modeling approaches to enhance air quality management and inform policy decisions in agricultural regions. Graphical Abstract A Comprehensive study delves into the effects of sugarcane burning and atmospheric transport on the concerning PM2.5 concentrations within Florida's Everglades Agricultural Area. Utilizing high-resolution sensors, researchers identified a notable and alarming spike in PM2.5 levels on May 14, 2024, a surge that can primarily be attributed to both the controlled practice of sugarcane burning and prevailing weather conditions during that period. A meticulous source apportionment analysis painted a compelling picture of the origins of these fine particulate matter emissions, revealing that a staggering 85% of PM2.5 was traced back to Saharan dust, a natural phenomenon that can travel vast distances across the globe. In stark contrast, the contribution from sugarcane ash was minimal, accounting for less than 5% of the pollution. Despite these occasional spikes in particulate matter, the overall Air Quality Index remained firmly in the "good" category, which signifies generally healthy air conditions for the wider population. However, it is crucial to note that sensitive groups, such as individuals with pre-existing respiratory issues, faced heightened risks from short-term exposure during these events. Further enhancing the analysis, a Gradient Boosting Regression model demonstrated exceptional predictive accuracy, achieving an impressive R² value of 0.93. This result underscores the promising role of machine learning techniques in assessing and better understanding air quality dynamics.