Air quality and climate impacts of alternative bus technologies in Greater London.

The environmental impact of diesel-fueled buses can potentially be reduced by the adoption of alternative propulsion technologies such as lean-burn compressed natural gas (LB-CNG) or hybrid electric buses (HEB), and emissions control strategies such as a continuously regenerating trap (CRT), exhaust gas recirculation (EGR), or selective catalytic reduction with trap (SCRT). This study assessed the environmental costs and benefits of these bus technologies in Greater London relative to the existing fleet and characterized emissions changes due to alternative technologies. We found a >30% increase in CO2 equivalent (CO2e) emissions for CNG buses, a <5% change for exhaust treatment scenarios, and a 13% (90% confidence interval 3.8-20.9%) reduction for HEB relative to baseline CO2e emissions. A multiscale regional chemistry-transport model quantified the impact of alternative bus technologies on air quality, which was then related to premature mortality risk. We found the largest decrease in population exposure (about 83%) to particulate matter (PM2.5) occurred with LB-CNG buses. Monetized environmental and investment costs relative to the baseline gave estimated net present cost of LB-CNG or HEB conversion to be $187 million ($73 million to $301 million) or $36 million ($-25 million to $102 million), respectively, while EGR or SCRT estimated net present costs were $19 million ($7 million to $32 million) or $15 million ($8 million to $23 million), respectively.

Effect of catalyst pretreatment on chirality-selective growth of single-walled carbon nanotubes

We show that catalyst pretreatment conditions can have a profound effect on the chiral distribution in single-walled carbon nanotube chemical vapor deposition. Using a SiO2-supported cobalt model catalyst and pretreatment in NH3, we obtain a comparably narrowed chiral distribution with a downshifted tube diameter range, independent of the hydrocarbon source. Our findings demonstrate that the state of the catalyst at the point of carbon nanotube nucleation is of fundamental importance for chiral control, thus identifying the pretreatment atmosphere as a key parameter for control of diameter and chirality distributions. © 2014 American Chemical Society.

Linear dimensionality reduction: Survey, insights, and generalizations

© 2015 John P. Cunningham and Zoubin Ghahramani. Linear dimensionality reduction methods are a cornerstone of analyzing high dimensional data, due to their simple geometric interpretations and typically attractive computational properties. These methods capture many data features of interest, such as covariance, dynamical structure, correlation between data sets, input-output relationships, and margin between data classes. Methods have been developed with a variety of names and motivations in many fields, and perhaps as a result the connections between all these methods have not been highlighted. Here we survey methods from this disparate literature as optimization programs over matrix manifolds. We discuss principal component analysis, factor analysis, linear multidimensional scaling, Fisher's linear discriminant analysis, canonical correlations analysis, maximum autocorrelation factors, slow feature analysis, sufficient dimensionality reduction, undercomplete independent component analysis, linear regression, distance metric learning, and more. This optimization framework gives insight to some rarely discussed shortcomings of well-known methods, such as the suboptimality of certain eigenvector solutions. Modern techniques for optimization over matrix manifolds enable a generic linear dimensionality reduction solver, which accepts as input data and an objective to be optimized, and returns, as output, an optimal low-dimensional projection of the data. This simple optimization framework further allows straightforward generalizations and novel variants of classical methods, which we demonstrate here by creating an orthogonal-projection canonical correlations analysis. More broadly, this survey and generic solver suggest that linear dimensionality reduction can move toward becoming a blackbox, objective-agnostic numerical technology.