Wildfires expose populations to increased morbidity and mortality due to the increase of air pollutant emissions. This study assesses the impact of wildfire exposure in Portugal. In this work, we analyze the effects of the wildfire seasons (June-July-August-September-October) on monthly mortality by using data from atmospheric composition reanalysis, air quality stations, remote sensing, and mortality for exposure assessment, cluster analyses and regression models. Cluster analysis separated the months within fire seasons with extreme atmospheric conditions (months with more frequency of lower relative humidity and higher temperature, higher pollutant concentrations and higher wildfire activities), Cluster 1, from months with cleaner air and stable atmospheric conditions, Cluster 2. Linear regression showed statistically significant (p-value < 0.05) correlation (r) between Cluster 1 and cardiorespiratory mortalities due to Diseases of the Respiratory System (DRS), Pneumonia (PNEU), Chronic Obstructive Pulmonary Disease (COPD) and Diseases of the Circulatory System (DCS) (rDRS = 0.49; rPNEU = 0.42; rCOPD = 0.44; rDCS = 0.45). Cluster 2 presented no significant statistical correlation between atmospheric conditions and health outcomes. Results shows epidemiological evidence that heat stress combined with air pollution during wildfire season are contributing to increase disease burden. Besides that, we performed smoke forecasts over Portugal by using the Hybrid Single-Particle Lagrangian Integrated Trajectory (HYSPLIT) model with satellite‐based fire emission. Forecasted PM10 and PM2.5 concentration reproduced the behavior of the observations (NRMSEPM10 = 3.70, rPM10 = 0.75; NRMSEPM2.5 = 1.51, rPM2.5 = 0.46) during October 15-16th, 2017, fire episode. BC results matches with satellite observations.
Wildfires expose populations to increased morbidity and mortality due to increased air pollutant concentrations. Data included burned area, particulate matter (PM10, PM2.5), carbon monoxide (CO), nitrogen dioxide (NO2), ozone (O3), temperature, relative humidity, wind-speed, aerosol optical depth (AOD) and mortality rates due to Circulatory System Disease (CSD), Respiratory System Disease (RSD), Pneumonia (PNEU), Chronic Obstructive Pulmonary Disease (COPD), and Asthma (ASMA). Only the months of the 2011-2020 wildfire season (June-July-August-September-October) with burned area greater than 1000 ha were considered. Multivariate statistical methods were used to reduce the dimensionality of the data to create two fire-pollution-meteorology indices (PBI, API), which allow us to understand how the combination of these variables affect cardio-respiratory mortality. Cluster analysis applied to PBI-API-Mortality divided the data into two Clusters. Cluster 1 included the months with lower temperatures, higher relative humidity, and high PM10, PM2.5, and NO2 concentrations. Cluster 2 included the months with more extreme weather conditions such as higher temperatures, lower relative humidity, larger forest fires, high PM10, PM2.5, O3, and CO concentrations, and high AOD. The two clusters were subjected to linear regression analysis to better understand the relationship between mortality and the PBI and API indices. The results showed statistically significant (p-value < 0.05) correlation (r) in Cluster 1 between RSDxPBI (rRSD = 0.539), PNEUxPBI (rPNEU = 0.644). Cluster 2 showed statistically significant correlations between RSDxPBI (rRSD = 0.464), PNEUxPBI (rPNEU = 0.442), COPDxPBI (rCOPD = 0.456), CSDxAPI (rCSD = 0.705), RSDxAPI (rCSD = 0.716), PNEUxAPI (rPNEU = 0.493), COPDxAPI (rPNEU = 0.619).