A 10-day winter sampling campaign was conducted in downtown Toronto for particulate matter (PM) air pollution in the fine (<2.5 microm) size range. An aerosol laser ablation mass spectrometer (LAMS), a tapered-element oscillating microbalance (TEOM), and an aerodynamic particle sizer (APS) were operated in parallel to characterize the PM on-line. In this study, the LAMS observed differences in the chemical composition between three separate episodes with higher PM2.5 mass and APS counts. LAMS results showed that in one instance of elevated PM, organic amines were present in the particulates. Temporal analyses of this episode revealed chemical transformations as the amines, characterized by m/z peaks 58(C3HeN)+, 86(C5H2N)+, and nitrates, increased in number concentration while Ca and hydrocarbon particle classes concurrently decreased. On another day, sulfates were found to have increased significantly. The third event was only 4 h in duration and exhibited an increase in the number of submicron-sized K/hydrocarbons and sulfate-containing particles. In this last event, the hydrocarbons and a K to Fe ratio enrichment indicated there was likely a contribution from a combustion source. This work offers some of the first insights into single particle size and chemistry in a cold winter climate.

No abstract

This paper presents an automatic and more robust implementation of multivariate adaptive regression splines (MARS) within the orthogonal array (OA)/MARS continuous-state stochastic dynamic programming (SDP) method. MARS is used to estimate the future value functions in each SDP level. The default stopping rule of MARS employs the maximum number of basis functions (M max ), specified by the user. To reduce the computational effort and improve the MARS fit for the wastewater treatment SDP model, two automatic stopping rules, which automatically determine an appropriate value for M max , and a robust version of MARS that prefers lower-order terms over higher-order terms are developed. Computational results demonstrate the success of these approaches.

There has been a blurring with respect to the retail formats because of competition and proliferation of different types of formats. In this research, we use a unique scanner panel dataset to investigate how brand choice behavior varies for the same consumer shopping for the same brand across different retail formats. We develop hypotheses pertaining to promotion sensitivity, price sensitivity, package size preference, and effects of demographic and shopping variables on consumer brand choice behavior and test them using a multi-format probit choice model that allows for the estimation of the cross-format differences with respect to the above. We find that consumers exhibit different promotion and price sensitivities in brand choice behavior between the mass merchandise format and supermarkets. Discussions and insights are provided.

In this paper, we present a Decision-Making Framework (DMF) for reducing ozone pollution in the metropolitan Atlanta region. High ground-level concentrations of ozone continue to be a serious problem in several US cities, and Atlanta is one of the most serious of these cases. In contrast to the "trial and error" approach utilized by state government decision-makers, our DMF searches for dynamic and focused control strategies that require a lower total reduction of emissions than current control strategies. Our DMF utilizes a rigorous stochastic dynamic programming formulation and includes an Atmospheric Chemistry Module to represent how ozone concentrations change over time. This paper focuses on the procedures within the Atmospheric Chemistry Module. Using the US EPA's Urban Airshed Model for Atlanta, we use mining and metamodeling tools to develop a computationally efficient representation of the relevant ozone air chemistry. The proposed approach is able to effectively model changes in ozone concentrations over a 24-hour period.