A low-cost approach to ambient fine particulate matter monitoring - the case of Grenada

Abstract Ambient fine particulate matter (PM2.5), despite its association with significant morbidity and mortality, is not robustly monitored in the small island developing states (SIDS) of the Caribbean region. This is due in part to a lack of local and regional financial and technical capacity. To establish baseline parameters and assess historical trends in Grenada, an island frequently impacted by Saharan dust storms, a historical analysis was conducted using comprehensive atmospheric reanalysis data from NASA's Global Modeling and Assimilation Office. A time series decomposition analysis, along with a multivariate linear regression model, was used to study the trends and interactions between PM2.5 and meteorological correlates. The analysis revealed that, despite stable mean trends, PM2.5 concentrations increased consistently and significantly from May to August, with seasonal effect indices ranging from 0.198 to 6.799 units above the average trend. These pronounced peaks were linked to recorded dust storms and other natural phenomena, including volcanic activity. A correlation analysis with results from a pilot study from 2020 showed a moderate association between satellite-based estimates and low-cost ground monitors (ρ = 0.56). Given the link between PM2.5 exposure and various adverse health effects, the findings from this study can serve as a basis for establishing more comprehensive monitoring networks. This can inform air quality management decisions tailored to the region and protect public health. Additionally, the lessons learned from this study can enhance research capabilities and support the implementation of air quality management initiatives in Grenada and other similar SIDS. Key messages • Grenada and the southern Caribbean region air quality is primarily affected by Sahara dust storms. • In absence of reference ground monitors, a robust network can be established using satellite estimates and low-cost monitoring.