The drivers of human health and changing climate are important areas of environmental and atmospheric studies. Among many environmental factors present in our biosphere, small particles, also known as ultrafine particles or UFPs, have direct and indirect pathways to affect human health and climatic processes. The rapid change in their properties makes UFPs dynamic and often challenging to quantify their effect on health and radiative forcing. To reduce uncertainty in the climate effects of UFPs and to strengthen the evidence on health effects, accurate characterizations of physical and chemical properties of UFPs are needed. In this thesis, two broad aspects of UFPs were investigated: (1) the development of particle instrumentation to study particle properties; and (2) measurement of physical and chemical properties of UFPs relevant to human health and climate. These two broad aspects are divided into four specific aims in this thesis. The measurement of UFP concentration at different locations in an urban location, from roadside to various residential areas, can be improved by using a mobile particle counter. A TSI 3786 Condensation Particle Counter (CPC) was modified for mobile battery-power operation. This design provided high-frequency, one second time resolution measurements of particle number and carbon dioxide (CO2). An independent electric power system, a central controller and robust data acquisition system, and a GPS system are the major components of this mobile unit. These capabilities make the system remotely deployable, and also offer flexibility to integrate other analog and digital sensors. A Volatility Tandem Differential Mobility Analyzer (V-TDMA) system was designed and built to characterize the volatility behavior of UFPs. The physical and chemical properties of UFPs are often challenging to measure due to limited availability of instruments, detection limit in terms of particle size and concentration, and sampling frequency. Indirect methods such as V-TDMA are useful, for small mass (<1 µg/m3), and nuclei mode particles (<30nm). Another advantage of V-TDMA is its fast response in terms of sampling frequency. A secondary motivation for building a V-TDMA system was to improve instrumentation capability of our group, thus enabling study of kinetic and thermodynamic properties of novel aerosols. Chapter four describes the design detail of the built V-TDMA system, which measures the change in UFP size and concentration during heated and non-heated (or ambient) condition. The V-TDMA system has an acceptable penetration efficiency of 85% for 10 nm and maintains a uniform temperature profile in the heating system. Calibration of V-TDMA using ammonium sulfate particles indicated that the system produces comparable evaporation curves (in terms of volatilization temperature) or volatility profiles to other published V-TDMA designs. Additionally the system is fully programmable with respect to particle size, temperature and sampling frequency and can be run autonomously after initial set up. The thesis describes a part of yearlong study to provide a complete perspective on particle formation and growth in a rural and agricultural Midwestern site. Volatility characterizations of UFPs were conducted to enable inference about particle chemistry, and formation of low volatile core or evaporation resistant residue in the UFP in the Midwest. This study addresses identification of the volatility signature of particles in the UFP size range, quantification of physical differences of UFPs between NPF1 and non-NPF events and relation of evaporation resistant residue with particle size, seasonality and mixing state. K-means clustering was applied to determine three unique volatility clusters in 15, 30, 50 and 80 nm particle sizes. Based on the proposed average volatility, the identified volatility clusters were classified into high volatile, intermediate volatile and least volatile group. Although VFR alone is insufficient to establish chemical composition definitively, least volatile cluster based on average volatility may be characteristically similar to the pure ammonium sulfate. The amount of evaporation residue at 200 °C was positively correlated with particle size and showed significant correlation with ozone, sulfur dioxide and solar radiation. Residue also indicated the presence of external mixture, often during morning and night time. Air quality science and management of an accidental urban tire fire occurring in Iowa City in May and June of 2012 were investigated. Urban air quality emergencies near populated areas are difficult to evaluate without a proper air quality management and response system. To support the development of an appropriate air quality system, this thesis identified and created a rank for health-related acute and chronic compounds in the tire smoke. For health risk assessment, the study proposed an empirical equation for estimating multi-pollutant air quality index. Using mobile measurements and a dispersion model in conjunction with the proposed air quality index, smoke concentrations and likely health impact were evaluated for Iowa City and surrounding areas. It was concluded that the smoke levels reached unhealthy outdoor levels for sensitive groups out to distances of 3.1 km and 18 km at 24 h and 1h average times. Tire smoke characterization was another important aspect of this study which provided important and new information about tire smoke. Revised emission factors for coarse particle mass and aerosol-PAH and new emission factors and enhancement ratios values for a wide range of fine particulate mass, particle size (0.001-2.5 µm), and trace gas were estimated. Overall the thesis added new instrumentation in our research group to measure various physical properties such as size, concentration, and volatility UFP. The built instruments, data processing algorithm and visualization tools will be useful in estimation of accurate concentration and emission factors of UFP for health exposure studies, and generate a fast response measurement of kinetic and thermodynamics properties of ambient particles. This thesis also makes a strong case for the development of an air quality emergency system for accidental fires for urban location. It provides useful evaluation and estimation of many aspects of such system such as smoke characterization, method of air quality monitoring and impact assessment, and develops communicable method of exposure risk assessment.

Background: Air pollution has become a serious environmental issue. It is responsible for hundreds of fatalities each year and it poses a serious threat to human health and environment. It leads to global warming, greenhouse effect and it also causes respiratory problems like asthma, lung cancer etc. It is important to predict the quality of the air to regulate air pollution. Air quality index (AQI) is a measure of air quality which describes the level of air pollution. Machine learning algorithms can help in predicting the AQI. Linear regression, LASSO regression, ridge regression, and SVR algorithms were used to forecast the AQI.  Objectives: The main objective of the thesis is to build and train a models using machine learning algorithms and find out the most accurate model in predicting the AQI.  Methods: Literature Review and Experimentation were chosen as methods to answer the research questions. There are a number of research papers written on prediction of AQI and literature review helped us a lot in research and references. Experimentation is also used to find out the most accurate machine learning model in predicting the air quality. In the experimentation phase, four machine learning algorithms were trained with air quality data to create predictive models for fore- casting AQI.  Results: Algorithms like Logistic Regression, Ridge Regression, LASSO Regression, and SVR are selected through literature review. Upon experimentation and training the algorithm with "Air Quality Data in India (2015-2020)" data set has showed that Ridge regression has the least MAE and RMSE and the highest R- square, which shows that it has the highest performance in predicting the AQI.  Conclusions: Four models are built by training with machine learning algorithms like Logistic Regression, Ridge Regression, LASSO Regression, and SVR and "Air Quality Data in India (2015-2020)" data set. After experimentation, it was found that Ridge Regression and LASSO regression has the better performance in the prediction of AQI.

No description available.

No description available.

No description available.