No description available.

Increasing levels of air pollution are prompting researchers to develop more reliable air pollution modeling approaches in order to protect the public and the environment from toxic contaminants and airborne pathogens. Although land use regression has long been used to assess exposure to air pollution, researchers are increasingly using machine learning algorithms to quantify the concentration of harmful pollutants—for this study black carbon (BC) and particle number (PN). Additionally, researchers are moving away from using fixed-site data in favor of using mobile monitoring data in a variety of locations to develop hourly empirical models of particulate air pollution. This study uses secondary data describing BC and PN pollutant levels, which are obtained from roads that bikers share in the more rural location of Blacksburg (VA). Machine learning (ML) algorithms are then built to develop accurate and reliable short-term empirical prediction models. Different pre-processing methods for the mobile monitoring data and various input variables are tested to assess how ML can be used effectively in this process. Three types of time-average models are developed (daytime, hourly average, and one second models). Various combinations of spatial and temporal input variables are used in the short-term models. The impact of adding more spatiotemporal variables (e.g., emissions) to machine learning models to improve model performance is assessed in the short-term models. Incorporating spatial and temporal autocorrelation is intended to develop more sophisticated validation approaches for identifying ML performance patterns—the goal of which is to predict concentration levels more accurately in comparison to using raw data without data reprocessing. The results show that the model developed using refined disaggregated data is able to detect the spatial distribution of the pollutant concentration at equivalent levels as the smoothed data models, although the latter display fewer errors. The performance of the short-term model including all variables is equivalent to the model omitting emissions. The ML results are compared to earlier stepwise regression model results, suggesting that ML has the ability to improve both long-term and short-term model accuracy. Our findings indicate that ML demonstrates higher predictive capacity in comparison to stepwise regression. The results from this study may be useful in enhancing the performance of ML through the incorporation of different data preprocessing tasks, as well as showing how different input variables contribute to the ML modeling process. The findings from this study could be used toward the development of environmental/eco-friendly routes that would decrease the risk for exposure to harmful vehicle-related emissions. / Doctor of Philosophy / Air pollution is a major environmental threat to human health, claiming the lives of millions of people each year, primarily as a result of fine particulate matter entering the respiratory system. As such, it is important to develop reliable and accurate air pollution modeling approaches in order to protect the public and the environment from toxic contaminants and pathogens in the air. Although an approach known as land use regression has long been used to assess exposure to air pollution, researchers are increasingly using machine learning (ML) algorithms to quantify the concentration of harmful pollutants—for this study black carbon and particle number, which is a generic assessment that captures a number of known airborne hazards. Additionally, researchers are moving away from using fixed-site data in favor of using mobile monitoring data in a variety of locations to develop hourly empirical models of particulate air pollution. In this study, machine learning algorithms are developed using secondary data collected from roads that bikers share, which are representative of pollution levels of particle number and black carbon in the more rural location of Blacksburg (VA), in order to develop accurate and reliable short-term empirical prediction models. Different pre-processing methods of the mobile monitoring data and various input variables are tested to assess how machine learning can be efficiently used in this process. Our findings indicate that machine learning demonstrates higher predictive capacity in comparison to stepwise regression. The results from this study are expected to be useful in enhancing the performance of machine learning through the incorporation of different data preprocessing tasks, as well as how different input variables contribute to the machine learning modeling process. The findings from this study could assist transportation planners and other stakeholders better assess pollution risks for bike riders and pedestrians. As such, this study's findings could be used toward the development of environmental/eco-friendly routes that would decrease the risk for exposure to harmful vehicle-related emissions.

Este trabalho teve como objetivo avaliar o impacto da mudança dos fatores de emissão veicular na formação de ozônio troposférico na Região Metropolitana de São Paulo (RMSP), Brasil, comparando os anos de 2004 e 2011. Essa avaliação foi realizada por meio de modelagem numérica com o modelo de qualidade do ar Weather Research and Forecasting with Chemistry (WRF-Chem), configurado sobre um domínio com 4 grades aninhadas: 27, 9, 3 e 1 km; e com o mecanismo químico considerado, o RADM2, acionado apenas nas grades de 3 e 1 km. Os inventários de emissão foram baseados em informações estimadas de número de veículos, fatores de emissão de compostos químicos e número médio de quilômetros percorridos por dia pelos distintos tipos de veículos considerados. Os períodos de estudo 06 a 09 de setembro de 2004 e 12 a 15 de novembro de 2011 foram selecionados, principalmente, por apresentarem episódios de altas concentrações de ozônio e pela existência de dados experimentais de determinação de fatores de emissão. Para avaliar a performance das simulações foram utilizados dados observados de temperatura, umidade relativa, vento, concentração de ozônio, óxidos de nitrogênio e monóxido de carbono das estações de qualidade do ar da rede de monitoramento da Companhia de Tecnologia de Saneamento Ambiental (CETESB). Os resultados mostraram que as configurações físicas e químicas estabelecidas no modelo conseguiram representar adequadamente os ciclos de formação do ozônio, no entanto, com defasagens e intensidades máximas menores, características relacionadas com o cálculo das distribuições espaço/temporal das emissões como também pelo transporte não só do ozônio, mas também de seus precursores desde regiões externas à RMSP. Por outro lado, dadas as características de emissão em 2004 e 2011, encontrou-se que a atmosfera da RMSP estava caracterizada por um regime COV-limitante naqueles anos, principalmente no ano 2004. Finalmente, o cenário de emissão para 2011 considerando os mesmos fatores de emissão de 2004 mostrou um grande impacto na formação do ozônio troposférico, considerando a grade de 1 km de espaçamento. Esse resultado foi 8 produto do estabelecimento de uma razão média COV/NOx mais eficiente em termos de formação de ozônio, apesar das emissões dos seus precursores ter diminuído em relação a 2004. / The objective of this work was to evaluate the impact of vehicular emission factors change on tropospheric ozone formation in the Metropolitan Area of São Paulo (MASP), Brazil for 2004 and 2011. Weather Research and Forecasting with Chemistry (WRF-Chem) model was used as the photochemical modeling tool. WRF-Chem was set up to run with four nested grid cells: 27, 9, 3 and 1 km. The Regional Acid Deposition Model (RADM) version 2 was the chemical mechanism considered, turned it on barely in 3 and 1 km grid cells. The emission inventories were based on the estimated information of number of vehicles, emission factors and average driving kilometers for vehicle per day. The study periods September 06-09, 2004 and November 12-15, 2011 were chosen due to the high ozone concentrations observed, and the experimental data available for these periods. Observation of temperature, relative humidity, winds, and O3, CO and NOx concentrations from the air quality monitoring network of Companhia de Tecnologia de Saneamento Ambiental do Estado de São Paulo (CETESB) were used to evaluate WRF-Chem simulations performance. The results showed that the chemical and physics configurations in WRF-Chem represented adequately the ozone formation cycles; however, with mismatches and maximum concentrations lower than observations, characteristics related with the estimative of the spatial and temporal distributions of the emissions as well as by the transportation of ozone and its precursors from external regions to MASP. On the other hand, given the characteristics of emission in 2004 and 2011, we found that MASP was under a strong COV-limited regime, principally in 2004. Finally, the scenario considering the emission factor of 2004 valid for 2011 in MASP showed a great impact in tropospheric ozone formation with high concentrations calculated for 2011 considering 1 km grid cell (with the most realistic vehicular activity), as result of a more efficient COV/NOx ratio calculation in terms of ozone formation.

As emissões globais de gases do efeito estufa (GEE) devido a ações do ser humano tem levado a um aumento na temperatura média da superfície terrestre de 0,55oC, e mudanças climáticas como aumento de eventos climáticos extremos, elevação dos níveis dos oceanos, e mudanças nos regimes pluviométricos são alguns exemplos das possíveis implicações deste aquecimento. O carbono orgânico do solo (COS) é o principal reservatório terrestre de C, contendo mais que o dobro do C da atmosfera, portanto, dependendo do manejo os solos podem se transformar em importantes fontes ou drenos de C atmosférico, influenciando significativamente os efeitos do aquecimento global. O objetivo desta pesquisa foi estimar as mudanças nos estoques do COS devido às mudanças no uso da terra e sistemas de manejo nos estados de Rondônia e Mato Grosso entre 1970 e 1985 e 1985 a 2002 utilizando dados específicos da região; e realizar a análise de incerteza destas estimativas mediante o método de Monte Carlo. Para alcançar o objetivo principal, a presente pesquisa foi composta das seguintes etapas: i) cálculo dos estoques do COS sob vegetação nativa (carbono de referência); ii) obtenção dos dados (áreas) das principais categorias de uso da terra nos estados de Rondônia e Mato Grosso para os anos de 1970, 1985 e 2002, a partir da combinação de técnicas de sensoriamento remoto, dados dos censos agropecuários, e informações de especialistas do setor agropecuário; iii) desenvolvimento dos fatores de emissão específicos para os principais sistemas de manejo da região de estudo utilizando um modelo linear misto; iv) e a etapa final que consistiu em combinar as etapas anteriores para se estimar as mudanças nos estoques de COS, e realizar a análise das incertezas associadas. Sucintamente, foram derivados fatores de emissão para as pastagens degradadas (0,91 ± 0,14), típicas em Latossolos (0,99 ± 0,08), típicas nos demais tipos de solos (1,24 ± 0,07), e pastagens melhoradas em Latossolos (1,19 ± 0,07), todos os fatores representam a comparação entre as pastagens manejadas e a vegetação nativa. Nos sistemas agrícolas foi possível derivar fatores de emissão para sistemas de plantio direto (PD) em áreas de Cerrado (1,08 ± 0,06), PD em áreas de floresta Amazônica e Cerradão (1,01 ± 0,17), cultivo convencional (CC) (0,94 ± 0,04) e culturas perenes (0,98 ± 0,14), sendo que o fator para o CC foi comparado aos dados de PD, enquanto que os demais fatores foram obtidos a partir da comparação com os estoques sob vegetações nativas. Quanto às emissões de COS, foi encontrado que usando o método de Monte Carlo com 20000 simulações no período de 1970 a 1985, os solos minerais apresentaram uma perda de C com fluxos anuais de 4,28 e 1,14 Tg C ano-1, para Mato Grosso e Rondônia, respectivamente, e com 95% de intervalo de confiança as incertezas foram de ± 41,5 e 21,9%, respectivamente. No segundo período, as emissões foram de 2,86 e 0,91 Tg C ano-1, com incertezas de ± 40,1 e 33,8%, respectivamente, para Mato Grosso e Rondônia. Quanto às fontes de incerteza, o carbono de referência, a opinião dos especialistas sobre as condições das pastagens e os fatores de emissão para pastagens típicas e degradadas foram às variáveis responsáveis por mais de 90% das incertezas das estimativas das emissões de C do solo. / Global emissions of greenhouse gases (GHG) due to human beings actions have led to an increase in average temperature of the earth of 0,55oC, and climate changes such as increases of the extreme weather events, sea level rise and precipitation changes are some examples of the possible implications of this global warming. Soil organic carbon (SOC) is the largest terrestrial organic carbon pool, containing more than the double of the atmospheric C; therefore, depending on the management the soils can became a source or a sink for the atmospheric C, influencing the effects of global warming. The objective of this research was to estimate the changes in SOC stocks due to the land-use and management systems in the states of Rondônia and Mato Grosso from 1970 to 1985 and from 1985 to 2002 using regional specific data, and perform the uncertainty analysis of these estimates through the Monte Carlo method. To achieve the main objective, this research was composed by the following steps: i) the estimate of the SOC stocks under native vegetation (reference carbon), ii) obtain data (areas) of the main land-use categories in the states of Rondônia and Mato Grosso for the years 1970, 1985 and 2002, from a combination of remote sensing, agricultural census data, and information from experts of the agricultural sector, iii) to derive the emission factors specific to the major management systems in the region of study using a linear-mixed model; iv) the final step was to combine the above steps to estimate the changes in SOC stocks, and carry out the uncertainty analysis associated with them. Briefly, emission factors were derived to the degraded grasslands (0.91 ± 0.14), typical in Oxisols (0.99 ± 0.08), typical in other soil types (1.24 ± 0.07), and improved grasslands in Oxisols (1.19 ± 0.07), all factors represent the comparison between managed pastures and native vegetation. In agricultural systems could be derived emission factors for no tillage (NT) systems in the Cerrado areas (1.08 ± 0.06), NT in Amazon Forest and Cerradão areas (1.01 ± 0.17), conventional tillage (CT) (0.94 ± 0.04), and perennial crops (0.98 ± 0.14). However, the CT factor was obtained from the comparison with NT data, while the other factors were compared to SOC stocks under the native vegetation. Using the Monte Carlo approach with 20000 simulations it was estimated the changes in the SOC stocks and the uncertainties associated to them. In the period from 1970 to 1985, mineral soils had a loss of C with annual fluxes of 4.28 and 1.14 Tg C yr-1, in Mato Grosso and Rondônia, respectively, and with a 95% confidence interval the uncertainties were of ± 41.5 and 21.9% respectively. In the second period, emissions were of 2.86 and 0.91 Tg C yr-1, with uncertainty of ± 40.1 and 33.8%, respectively, in Mato Grosso and Rondônia. In terms of the sources of uncertainty, the reference carbon, the experts opinions about the grasslands, and the emission factors for typical and degraded grassland were the variables responsible for more than 90% of the uncertainties in the SOC emissions estimates.

Les incendies de végétation sont caractérisés par de nombreuses échelles de temps et d’espace. Une approche multi-physique et multi-échelle est donc nécessaire pour aborder la complexité de ces phénomènes. Ce travail de thèse est une contribution expérimentale à l’étude des effets d’échelle et du système sur la dynamique de combustion des végétaux et les fumées émises. Notre objectif principal a été de déterminer quels protocoles expérimentaux et plus particulièrement quelles échelles permettent de caractériser au mieux la combustion des végétaux en laboratoire. Nous avons ainsi étudié le comportement au feu des aiguilles de deux espèces de pin représentatives de l’écosystème méditerranéen, Pinus pinaster et Pinus laricio. Les litières d’aiguilles de pin représentent un enjeu important car elles entretiennent la propagation des incendies et elles participent à la transition d’un feu de surface à un feu total. Trois configurations expérimentales ont été étudiées au cours de cette thèse, l’échelle du cône calorimètre, l’échelle du grand calorimètre ou LSHR (permettant la combustion statique de masses plus importantes que le cône) et enfin, la propagation dans le LSHR permettant d’étudier l’effet du système sur la dynamique de la combustion et sur la production des fumées. Pour comparer ces trois configurations les protocoles expérimentaux ont été adaptés tout en maintenant la charge de combustible. Différents paramètres ont été étudiés pour analyser la dynamique de combustion tels que l’efficacité de la combustion, l’énergie dégagée ou encore la vitesse de perte de masse. La production des fumées a été étudiée par la mesure du coefficient d’extinction qui caractérise leur opacité et permet d’obtenir le facteur d’émission des suies. Les facteurs d’émissions des principaux composés émis lors de la combustion de ces deux types d’aiguilles de pin ont été mesurés en continu à partir d’un analyseur Infrarouge à Transformée de Fourier et d’un analyseur Infrarouge Non Dispersif. Des analyses par chromatographie en phase gazeuse couplée à un détecteur à ionisation de flamme et un spectromètre de masse nous ont permis de compléter ces mesures. Un bilan massique de carbone a également été réalisé afin de quantifier le carbone total mesuré dans nos analyses. Les principales contributions de notre travail sont les suivantes : l’étude du comportement au feu des aiguilles de P. pinaster a révélé des différences importantes pour la puissance dégagée aux échelles du cône calorimètre et du LSHR. En revanche, le système de combustion (propagation) n’influence pas cette grandeur. L’efficacité de la combustion est apparue peu dépendante de l’échelle et du système de combustion. Nous avons observé une influence de l’échelle de combustion sur la production totale des fumées (RSR) et sur le facteur d’émission des suies. Nous avons également montré que le système de combustion (la propagation) influence la dynamique et la valeur des facteurs d’émission de dioxyde et de monoxyde de carbone, composés majoritairement émis par ces combustions. Une influence de l’échelle et du système est également à noter sur les facteurs d’émissions des composés azotés et des COV émis pour les trois configurations expérimentales. Nous avons attribué les différences observées aux valeurs de températures des fumées. Enfin, une influence de la géométrie des particules a été mise en évidence par comparaison des combustions réalisées avec le cône calorimètre et le LSHR pour les deux types d’aiguilles de pin. La dynamique de combustion des aiguilles de Pinus laricio est moins affectée par le changement d’échelle que celle des aiguilles de Pinus pinaster (plus faible variation de la puissance de combustion). Nous avons également observé que les aiguilles de Pinus laricio, thermiquement plus fines que les aiguilles de Pinus pinaster présentent une valeur plus faible pour le facteur d’émission des suies. / Wildfires are characterized by a lot of scales of time and space. A multi-physics and multi-scale approach is required to consider the complexity of these phenomena. This thesis is an experimental contribution to the study of the scale effects and the effects of the system on the combustion dynamics of forest fuels and smoke emission. The aim of this work was to determine which experimental protocols and specifically which scales can be used to characterize the combustion of vegetal fuels in the laboratory. The reaction to fire of pine needles species representative of the Mediterranean ecosystem, (Pinus pinaster and Pinus laricio) has been studied. Litters of pine needles are an important issue because they sustain fire spread and are involved in the transition from a surface fire to a total fire.Three experimental configurations were studied in this thesis: the cone calorimeter scale; the large scale calorimeter or LSHR (allowing static combustion with larger masses than used with the cone); a fire spread in the LSHR, allowing to study the effect of the system on the dynamics of combustion and release of smoke. To compare these configurations, experimental protocols have been elaborated while keeping the same fuel load. Different parameters were studied to analyze the combustion dynamics such as combustion efficiency, heat released rate and mass loss rate. Smoke production was studied by measuring the coefficient of extinction to characterize their opacity and an estimation of the soot emission factor was derived. The emission factors of the main compounds emitted during the combustion of these two pines needles were measured with a Fourier Transform Infrared analyzer and a Non-dispersive infrared analyzer. Analysis by gas chromatography coupled with a flame ionization detector and a mass spectrometer allowed us to complete these measurements. A mass balance of carbon was also performed to quantify the total carbon measured through our analyzes.The main contributions of our work can be summarized as follow: the study of the burning of Pinus pinaster needles reveals significant differences for heat release rate (HRR) at both cone calorimeter and LSHR scales. However, the combustion system (fire spread) does not influence the HRR value at the LSHR scale. The combustion efficiency appeared to be independent with regard to the scale and the system. We observed a wide influence on the rate of smoke release and the emission factor of soot. We also shown that the combustion system (fire spread) influences the dynamics and value of emissions factors of carbon dioxide and carbon monoxide (major compound emitted for these combustions). An influence of the three experimental configurations on the emission factors of nitrogen compounds and VOC emissions was also noted. This difference was attributed to the level of smoke temperature. Finally, an effect of particles geometry was also pointed out by the comparison between the burnings performed with the cone calorimeter and the LSHR for both types of pine needles. The combustion dynamics of Pinus laricio needles was slighlty affected by changing scale in comparison to needles of Pinus pinaster (weak variation of HRR). We also observed that Pinus laricio’s needles, which are thermally thiner than Pinus pinaster ones have the lowest soot emission factor.

Este trabalho teve como objetivo avaliar o impacto da mudança dos fatores de emissão veicular na formação de ozônio troposférico na Região Metropolitana de São Paulo (RMSP), Brasil, comparando os anos de 2004 e 2011. Essa avaliação foi realizada por meio de modelagem numérica com o modelo de qualidade do ar Weather Research and Forecasting with Chemistry (WRF-Chem), configurado sobre um domínio com 4 grades aninhadas: 27, 9, 3 e 1 km; e com o mecanismo químico considerado, o RADM2, acionado apenas nas grades de 3 e 1 km. Os inventários de emissão foram baseados em informações estimadas de número de veículos, fatores de emissão de compostos químicos e número médio de quilômetros percorridos por dia pelos distintos tipos de veículos considerados. Os períodos de estudo 06 a 09 de setembro de 2004 e 12 a 15 de novembro de 2011 foram selecionados, principalmente, por apresentarem episódios de altas concentrações de ozônio e pela existência de dados experimentais de determinação de fatores de emissão. Para avaliar a performance das simulações foram utilizados dados observados de temperatura, umidade relativa, vento, concentração de ozônio, óxidos de nitrogênio e monóxido de carbono das estações de qualidade do ar da rede de monitoramento da Companhia de Tecnologia de Saneamento Ambiental (CETESB). Os resultados mostraram que as configurações físicas e químicas estabelecidas no modelo conseguiram representar adequadamente os ciclos de formação do ozônio, no entanto, com defasagens e intensidades máximas menores, características relacionadas com o cálculo das distribuições espaço/temporal das emissões como também pelo transporte não só do ozônio, mas também de seus precursores desde regiões externas à RMSP. Por outro lado, dadas as características de emissão em 2004 e 2011, encontrou-se que a atmosfera da RMSP estava caracterizada por um regime COV-limitante naqueles anos, principalmente no ano 2004. Finalmente, o cenário de emissão para 2011 considerando os mesmos fatores de emissão de 2004 mostrou um grande impacto na formação do ozônio troposférico, considerando a grade de 1 km de espaçamento. Esse resultado foi 8 produto do estabelecimento de uma razão média COV/NOx mais eficiente em termos de formação de ozônio, apesar das emissões dos seus precursores ter diminuído em relação a 2004. / The objective of this work was to evaluate the impact of vehicular emission factors change on tropospheric ozone formation in the Metropolitan Area of São Paulo (MASP), Brazil for 2004 and 2011. Weather Research and Forecasting with Chemistry (WRF-Chem) model was used as the photochemical modeling tool. WRF-Chem was set up to run with four nested grid cells: 27, 9, 3 and 1 km. The Regional Acid Deposition Model (RADM) version 2 was the chemical mechanism considered, turned it on barely in 3 and 1 km grid cells. The emission inventories were based on the estimated information of number of vehicles, emission factors and average driving kilometers for vehicle per day. The study periods September 06-09, 2004 and November 12-15, 2011 were chosen due to the high ozone concentrations observed, and the experimental data available for these periods. Observation of temperature, relative humidity, winds, and O3, CO and NOx concentrations from the air quality monitoring network of Companhia de Tecnologia de Saneamento Ambiental do Estado de São Paulo (CETESB) were used to evaluate WRF-Chem simulations performance. The results showed that the chemical and physics configurations in WRF-Chem represented adequately the ozone formation cycles; however, with mismatches and maximum concentrations lower than observations, characteristics related with the estimative of the spatial and temporal distributions of the emissions as well as by the transportation of ozone and its precursors from external regions to MASP. On the other hand, given the characteristics of emission in 2004 and 2011, we found that MASP was under a strong COV-limited regime, principally in 2004. Finally, the scenario considering the emission factor of 2004 valid for 2011 in MASP showed a great impact in tropospheric ozone formation with high concentrations calculated for 2011 considering 1 km grid cell (with the most realistic vehicular activity), as result of a more efficient COV/NOx ratio calculation in terms of ozone formation.

As emissões globais de gases do efeito estufa (GEE) devido a ações do ser humano tem levado a um aumento na temperatura média da superfície terrestre de 0,55oC, e mudanças climáticas como aumento de eventos climáticos extremos, elevação dos níveis dos oceanos, e mudanças nos regimes pluviométricos são alguns exemplos das possíveis implicações deste aquecimento. O carbono orgânico do solo (COS) é o principal reservatório terrestre de C, contendo mais que o dobro do C da atmosfera, portanto, dependendo do manejo os solos podem se transformar em importantes fontes ou drenos de C atmosférico, influenciando significativamente os efeitos do aquecimento global. O objetivo desta pesquisa foi estimar as mudanças nos estoques do COS devido às mudanças no uso da terra e sistemas de manejo nos estados de Rondônia e Mato Grosso entre 1970 e 1985 e 1985 a 2002 utilizando dados específicos da região; e realizar a análise de incerteza destas estimativas mediante o método de Monte Carlo. Para alcançar o objetivo principal, a presente pesquisa foi composta das seguintes etapas: i) cálculo dos estoques do COS sob vegetação nativa (carbono de referência); ii) obtenção dos dados (áreas) das principais categorias de uso da terra nos estados de Rondônia e Mato Grosso para os anos de 1970, 1985 e 2002, a partir da combinação de técnicas de sensoriamento remoto, dados dos censos agropecuários, e informações de especialistas do setor agropecuário; iii) desenvolvimento dos fatores de emissão específicos para os principais sistemas de manejo da região de estudo utilizando um modelo linear misto; iv) e a etapa final que consistiu em combinar as etapas anteriores para se estimar as mudanças nos estoques de COS, e realizar a análise das incertezas associadas. Sucintamente, foram derivados fatores de emissão para as pastagens degradadas (0,91 ± 0,14), típicas em Latossolos (0,99 ± 0,08), típicas nos demais tipos de solos (1,24 ± 0,07), e pastagens melhoradas em Latossolos (1,19 ± 0,07), todos os fatores representam a comparação entre as pastagens manejadas e a vegetação nativa. Nos sistemas agrícolas foi possível derivar fatores de emissão para sistemas de plantio direto (PD) em áreas de Cerrado (1,08 ± 0,06), PD em áreas de floresta Amazônica e Cerradão (1,01 ± 0,17), cultivo convencional (CC) (0,94 ± 0,04) e culturas perenes (0,98 ± 0,14), sendo que o fator para o CC foi comparado aos dados de PD, enquanto que os demais fatores foram obtidos a partir da comparação com os estoques sob vegetações nativas. Quanto às emissões de COS, foi encontrado que usando o método de Monte Carlo com 20000 simulações no período de 1970 a 1985, os solos minerais apresentaram uma perda de C com fluxos anuais de 4,28 e 1,14 Tg C ano-1, para Mato Grosso e Rondônia, respectivamente, e com 95% de intervalo de confiança as incertezas foram de ± 41,5 e 21,9%, respectivamente. No segundo período, as emissões foram de 2,86 e 0,91 Tg C ano-1, com incertezas de ± 40,1 e 33,8%, respectivamente, para Mato Grosso e Rondônia. Quanto às fontes de incerteza, o carbono de referência, a opinião dos especialistas sobre as condições das pastagens e os fatores de emissão para pastagens típicas e degradadas foram às variáveis responsáveis por mais de 90% das incertezas das estimativas das emissões de C do solo. / Global emissions of greenhouse gases (GHG) due to human beings actions have led to an increase in average temperature of the earth of 0,55oC, and climate changes such as increases of the extreme weather events, sea level rise and precipitation changes are some examples of the possible implications of this global warming. Soil organic carbon (SOC) is the largest terrestrial organic carbon pool, containing more than the double of the atmospheric C; therefore, depending on the management the soils can became a source or a sink for the atmospheric C, influencing the effects of global warming. The objective of this research was to estimate the changes in SOC stocks due to the land-use and management systems in the states of Rondônia and Mato Grosso from 1970 to 1985 and from 1985 to 2002 using regional specific data, and perform the uncertainty analysis of these estimates through the Monte Carlo method. To achieve the main objective, this research was composed by the following steps: i) the estimate of the SOC stocks under native vegetation (reference carbon), ii) obtain data (areas) of the main land-use categories in the states of Rondônia and Mato Grosso for the years 1970, 1985 and 2002, from a combination of remote sensing, agricultural census data, and information from experts of the agricultural sector, iii) to derive the emission factors specific to the major management systems in the region of study using a linear-mixed model; iv) the final step was to combine the above steps to estimate the changes in SOC stocks, and carry out the uncertainty analysis associated with them. Briefly, emission factors were derived to the degraded grasslands (0.91 ± 0.14), typical in Oxisols (0.99 ± 0.08), typical in other soil types (1.24 ± 0.07), and improved grasslands in Oxisols (1.19 ± 0.07), all factors represent the comparison between managed pastures and native vegetation. In agricultural systems could be derived emission factors for no tillage (NT) systems in the Cerrado areas (1.08 ± 0.06), NT in Amazon Forest and Cerradão areas (1.01 ± 0.17), conventional tillage (CT) (0.94 ± 0.04), and perennial crops (0.98 ± 0.14). However, the CT factor was obtained from the comparison with NT data, while the other factors were compared to SOC stocks under the native vegetation. Using the Monte Carlo approach with 20000 simulations it was estimated the changes in the SOC stocks and the uncertainties associated to them. In the period from 1970 to 1985, mineral soils had a loss of C with annual fluxes of 4.28 and 1.14 Tg C yr-1, in Mato Grosso and Rondônia, respectively, and with a 95% confidence interval the uncertainties were of ± 41.5 and 21.9% respectively. In the second period, emissions were of 2.86 and 0.91 Tg C yr-1, with uncertainty of ± 40.1 and 33.8%, respectively, in Mato Grosso and Rondônia. In terms of the sources of uncertainty, the reference carbon, the experts opinions about the grasslands, and the emission factors for typical and degraded grassland were the variables responsible for more than 90% of the uncertainties in the SOC emissions estimates.

Welding is a common industrial practice that has the potential to emit air pollutants. Emission factors are useful indicators to help in the understanding of the extent of pollution from a process and managing them to reduce or minimize health impacts. The objective of this thesis is to determine emission factors applicable to the gas metal arc welding (GMAW), under varying current and voltage conditions. The most used base metals and an electrode for the shipbuilding industry were considered. A weld fume chamber was used to achieve the project goals along with standard sampling and analytical procedures. Three test runs were performed for each sampling scenario to ensure repeatability. The EPA EF average for MS experiments with the ER70S-6 electrode is 5.2 g/kg, and for SS experiments with the ER316L-Si electrode is 3.2 g/kg, while the average results for this study were 6.81 g/kg and 3.28 g/kg respectively.

The rapid industrialization, population growth, urbanization and economic and social development cause many environmental problems, such as climate change which is the result of the increase in the emission of greenhouse gases (GHGs) especially CO2. Combustion of fossil fuels, particularly from electricity generation, has the major responsibility for CO2 emissions. Decreasing the amount of CO2 emission requires a significant shift from our present energy use pattern toward one of lesser reliance on fossil fuels. Using renewable energy sources is one of the ways to supply some of the electricity demand reducing the associated GHG emissions and thus decreasing the adverse effects of climate change. In this study, generated electricity associated CO2 emissions and the specific CO2 emission factors are calculated based on IPCC methodology for each fuel type and each thermal power plant for Turkey between 2001 and 2008. The electricity demand of Turkey is estimated to increase about 7% annually till to 2019. Based on the planned power plant data obtained from EMRA, it was found that the total electricity supply (planned + existing) will not be sufficient to provide the estimated demand between 2011 and 2019. To overcome supply deficiency problem, four different scenarios are developed and the mitigation potential of CO2 emission from electricity generation based on these scenarios are examined. The results from these scenarios show that there is a significant decrease in the amount of CO2 emission from electricity generation. Renewable Energy Scenario which is the best scenario in terms of mitigation of CO2 emissions, provides to mitigation of 192 millions of CO2 emissions between 2009 and 2019. with respect to BAU scenario

Particulate matter (PM) is a widespread air pollutant of microscopic particles that are suspended in the atmosphere. There is high concentration of PM in underground train stations and tunnels. The concentration of particulate matter (PM10) measured in Stockholm’s metro varies significantly between day and nighttime traffic of trains. Emission factors are a representative value of a pollutant released into the atmosphere. These factors can be expressed as the weight of pollutant divided by a unit distance, volume, or weight. In this document it is expressed as the amount of energy used to produce a unit weight. Such factor expresses estimation of emissions from train wheel-rail contact and brake. Simulation of particulate matter using IDA tunnel was conducted to evaluate and predict particulate matter (PM10) concentration levels. Previous measured data of PM10 from Mariatorget’s platform (performed by SLB analysis) was used for the validation of the simulation model. These data were collected during winter and summer periods. It was then used as fundamental reference to simulate and calibrate emission factors. Important parameters of the tunnel, platform and trains that were used in traffic were imported. There are ventilation shafts on each end of the platform. These ventilation shafts are open fully during summer and closed during winter periods. Thus, two case scenarios, winter and summer were investigated. The results obtained were later evaluated and analysed. Sensitivity analysis was made to test the effect of ventilation shaft’s openings on emission factors. The results from winter case study showed that emission factors are 0.57 g/kWh from wheel-rail contact and 0.031 g/kWh from brakes. Emission factors from wheel-rail contact give 70% of the measured PM10 concentration whereas brakes give 30% of PM10 concentrations. Results obtained from summer case study showed that emission factors are 0.61 g/kWh and 0.05 g/kWh from wheel-rail contact as well as from brakes, respectively. / Partiklar är en utbredd luftföroreningar av mikroskopiska partiklar som finns i luften. Det finns höga halter av PM (particulate matter) i underjordiska tågstationer och tunnlar. Partikelhalten (PM10) som är uppmätt i Stockholms tunnelbana varierar betydligt mellan dag- och natttågtrafik. Emissionsfaktorn är ett representativt värde för mängden partiklar som släpps ut i atmosfären. Dessa faktorer kan uttryckas som massan av partikel per enhetsavstånd, volym eller vikt. I detta dokument uttrycks detta som den mängd energi som går åt för att producera enhetsviktspartiklar. Faktorn uttrycker en uppskattning av partikelemissioner från hjul-rälskontakt och broms. Simulering har genomfördes i IDA tunnel för att utvärdera och förutsäga partikelhalten på tågplattformar. Tidigare uppmätta data under vinter och sommar från Mariatorgets plattform (utförd av SLB-analys) användes för validering av simuleringsmodellen. Detta användes sedan som grundläggande referens för att simulera och kalibrera emissionsfaktorer. Viktiga parametrar för tunneln, plattformen och tågen som användes i trafiken är inhämtade och evaluerade. Ventilationsschakt finns i varje ände av plattformen. Dessa ventilationsschakt är öppna under sommaren och stängda under vinterperioden. Således undersöktes två scenarier, vinter- och sommarfall. De erhållna resultaten utvärderades och analyserades senare. Känslighetsanalys gjordes för att testa effekten på emissionsfaktorerna av ventilationsschaktens öppningsgrad. Resultaten från vinterfallstudien visade att emissionsfaktorerna är 0,57 g/kWh från hjul-rälkontaktoch 0,03 g/kWh från bromsarna. Emissionsfaktorn från hjul-rälkontakten ger 70 % av den uppmätta PM10 koncentrationen, medan bromsarna ger 30 %. Resultaten från sommarfallstudien visade att emissionsfaktorerna är 0,61 g/kWh och 0,05 g/kWh från hjul-rälkontakten och bromsarna, respektive.