Integrating Satellite and Sensor Measurements to Understand PM2.5 in Asunción, Paraguay

Exposure to high levels of ambient fine particulate matter (PM2.5) can increase risks for numerous adverse health effects including lung cancer, lower-respiratory infections, ischemic heart disease, adverse birth outcomes, and premature mortality. The World Health Organization recommends annual average PM2.5 levels not exceed 5 mg/m3 and 24-hr averages not exceed 15 mg/m3. Thus it is important that exposure levels be quantified through direct measurement. However, in many parts of the world including the Global South there are few or no monitors active to provide quantitative measurements of PM2.5 levels. In 2019 the Aire Paraguay PM2.5 sensor network was initiated in Asunción and other cities in Paraguay to address the lack of comprehensive data in the region. The network used Sensirion SPS30 laser-scattering PM2.5 sensors (Sensirion AG, Switzerland). These data were quality controlled and recorded as 5-min averages, and then distributed to the public via Twitter and a website. To date, these data come from the only ground-based air pollution network in Asunción. To augment these point source measurements and to better understand the spatial distribution of PM2.5, data were obtained from the Atmospheric Composition Analysis Group at Washington University in St. Louis. These data provided global high resolution (0.01° x 0.01°) PM2.5 concentrations at annual and monthly timescales derived from satellite Aerosol Optical Depth (AOD) retrievals combined with the GEOS-Chem chemical transport model and calibrated to global ground-based observations using a Geographically Weighted Regression (GWR). These two complementary data sets were used to characterize the spatial and temporal variation across the area as part of an effort to identify the representativeness of potential monitor locations. The sensor data showed clear diurnal variations by factors of 2 to 4 while the monthly-averaged satellite data showed spatial variations of a similar scale across the city. Both data sets showed significant areas and time periods where the PM2.5 concentration exceeded the WHO recommendations. These complementary data were valuable to create a spatially contiguous historical record of concentrations in addition to the real-time activity guidance available from the sensor network.