Assessment of the impact of biomass burning on air quality in the Colombian Orinoco River Basin

The “Llanos” is a half a million square kilometers savanna ecosystem that occupies most of the Orinoco River Basin (ORIB) in Northern South America, from the Colombian Andes foothills almost to the Orinoco River delta at the Atlantic Ocean in Venezuela. This binational savanna ecosystem undergoes periodic and extensive biomass burning (BB). During the dry season (BB period) northeast trade winds are intensified over the Llanos region, which implies that BB emissions may be transported to Colombian Orinoco river basin cities located southwest of the Llanos near the Andes foothills. Colombian Llanos have a population of 1.7 million inhabitants, 73% of them in the urban areas, who can potentially be exposed to BB pollutants. Biomass burning (BB) is a significant source of trace gases and aerosols that affects human health, atmospheric chemistry and climate. Health effects are mainly caused by particulate matter (PM) and photochemically-produced ozone. Most of BB emitted particles are within the accumulation mode (particle diameter 1μm), which implies low deposition velocities, long residence times in the atmosphere and potential for long-range transport. BB-derived organic particles contain mutagenic and carcinogenic components that lead to genotoxic effects on human alveolar cells, even when ambient air PM10 concentrations are below the World Health Organization (WHO) standard. Previous research conducted in Venezuela in the 1980-1990’s showed significant air quality impacts of BB in the Venezuelan Llanos. Until recently, there were no studies on the influence of biomass burning in the Colombian Llanos. Furthermore, at the beginning of this research no Colombian ORIB city had air quality monitoring systems, so there was scarce information on particulate matter levels in the region. Only recently, one of the main Colombian Llanos cities has established an air quality network. The source attribution problem, specifically the one of disentangling the BB emissions contribution to air quality degradation is usually tackled using simulation and/or analytic tools (e.g. chemical composition for receptor modeling), both of which have its own challenges. One of the fundamental problems when simulating BB impacts is building emission inventories, because BB is a complex, non-steady state problem, of high spatiotemporal variability. Emission estimation is even more difficult in the Llanos savannas where available information shows the predominance of small fires with more than 75% of patches smaller than 115 ha, which is smaller than the minimum detectable burned size of available global burned area products. The other aspect is to determine if emissions and meteorological conditions result in regional or even synoptic scale transport of BB emissions. Also, measurement and sampling of PM is required for both approaches (simulation and analytical tools). This thesis investigated the impact of biomass burning emissions on air quality of Colombian Orinoco river basin. The study included the following steps: (1) the development of a BB emission inventory using a methodology for integrating different satellite products to detect small fires that are predominant in the region; (2) monitoring of aerosol column using sun photometry (exploratory AERONET site); (3) measurement campaigns focused on particulate matter measurements and sampling during periods of high and low biomass burning activity; (4) meteorological simulations with a state of science model (WRF), and (5) Lagrangian atmospheric simulations to understand the transport of biomass burning emissions and their contribution to the particulate matter levels at relevant locations. The measurements showed that air quality in the Colombian Llanos significantly deteriorates during biomass burning periods. Fine mode aerosol content in the total atmospheric column increased the Aerosol Optical Depth (AOD) from ~0.15 during periods with low fire activity up to ~1 during biomass burning periods. BB affected in a similar way rural and suburban locations close to the Andes foothills, causing PM10 to increase by 21 µg/m³ and 26 µg/m³ at Taluma (rural) and Libertad (suburban) sites, respectively, during a period of high fire activity in Colombia and Venezuelan Llanos. Urban locations closer to the Venezuelan border were more severely impacted. Concentrations at Yopal and Arauca increased by ~ 70 µg/m³ during a period with high fire activity in the Venezuelan Llanos. The levels of PM10 and BC reached during the biomass burning period in the Llanos region are similar, or even higher than the levels observed in some urban monitoring stations of Bogota, where dense industrial and mobile sources exist. PM10 enhancements were closely related enhancements of biomass burning (BC, K+) and dust tracers. Exploratory source apportionment techniques demonstrated that around 80% of the observed PM10 enhancements are due to biomass burning emissions, and 20% are caused by soil dust. One of the main contributions of this thesis was the development of an indicator of the contribution at the receptor of particles from fire emissions (CFER). This tool uses Fire Radiative Power (FRP) from MODIS active fire detection as a proxy of fire emissions and the footprint generated from backward stochastic Lagrangian modeling to estimate the influence of fire emissions on the daily average aerosol concentration at receptors. This indicator performed reasonably well for PM10 observations at Yopal, Arauca and Taluma and has the potential to become an operational tool combining near real-time MODIS fire detections and Lagrangian simulations. Estimated PM10 enhancements calculated with STILT significantly underestimated observed enhancements. The analysis revealed that the underestimation is rather insensitive to: (1) driving meteorology (GDAS or WRF wind fields); (2) time allocation of emissions, and (3) emissions from other sources. Results suggest that the underestimation is likely related to secondary aerosol formation, not included within the modeling framework. This thesis contributes to the understanding of the impact of biomass burning on air quality of Colombian Orinoco River basin, addressing a regional pollution problem not previously identified in Colombia. Results provide solid scientific evidence of the impact of Llanos biomass burning on the air quality of Colombian Llanos cities, provides for the first time hard evidence of transboundary transport of BB pollution from Venezuela, and also provides an indicator that could be used by the environmental authorities to forecast pollution episodes and take actions to reduce population exposure.

• Autores: Hernández Villamizar, Andrea Juliana
• Tipo de recurso: Doctoral thesis
• Fecha de publicación: 2019
• Institución: Universidad Nacional de Colombia
• Repositorio: Universidad Nacional de Colombia
• Idioma: spa
• OAI Identifier: oai:repositorio.unal.edu.co:unal/76979
• Acceso en línea: https://repositorio.unal.edu.co/handle/unal/76979; http://bdigital.unal.edu.co/74080/
• Palabra clave: Biomass burning; Lagrangian simulations; Particulate matter; Orinoco River Basin; Aerosols; Quema de biomasa; Simulaciones Lagrangianas; Material particulado; Orinoquia; Contaminación transfronteriza; Aerosoles
• Rights: openAccess
• License: Atribución-NoComercial 4.0 Internacional