Multicollinearity and correlation among local regression coefficients in geographically weighted regression

Present methodological research on geographically weighted regression (GWR) focuses primarily on extensions of the basic GWR model, while ignoring well-established diagnostics tests commonly used in standard global regression analysis. This paper investigates multicollinearity issues surrounding the local GWR coefficients at a single location and the overall correlation between GWR coefficients associated with two different exogenous variables. Results indicate that the local regression coefficients are potentially collinear even if the underlying exogenous variables in the data generating process are uncorrelated. Based on these findings, applied GWR research should practice caution in substantively interpreting the spatial patterns of local GWR coefficients. An empirical disease-mapping example is used to motivate the GWR multicollinearity problem. Controlled experiments are performed to systematically explore coefficient dependency issues in GWR. These experiments specify global models that use eigenvectors from a spatial link matrix as exogenous variables.This study was supported by grant number 1 R1 CA95982-01, Geographic-Based Research in Cancer Control and Epidermiology, from the National Cancer Institute. The author thank the anonymous reviewers and the editor for their helpful comments.     

The impact of binary-star yields on the spectra of galaxies

One of the complexities in modelling integrated spectra of stellar populations is the effect of interacting binary stars besides Type Ia supernovae (SNeIa). These include common envelope systems, cataclysmic variables, novae, and are usually ignored in models predicting the chemistry and spectral absorption line strengths in galaxies. In this paper, predictions of chemical yields from populations of single and binary stars are incorporated into a galactic chemical evolution model to explore the significance of the effects of these other binary yields. Effects on spectral line strengths from different progenitor channels of SNeIa are also explored. Small systematic effects are found when the yields from binaries, other than SNeIa, are included, for a given star formation history. These effects are, at present, within the observational uncertainties on the line strengths. More serious differences can arise in considering different types of SNIa models, their rates and contributions.     

The use of7LiOH-impregnated filters for the collection of acidic gases and analysis by instrumental neutron activation analysis

A simple system was designed using⁷LiOH-impregnated filters to collect acidic gases from ambient air and from highly concentrated volcanic plumes or gas streams. These filters were developed for analysis using instrumental neutron activation analysis, but other analytical techniques could be used as well. The sampling system was designed to use a series of 1–3 M⁷LiOH-impregnated filters to collect high concentrations of acidic gases found in gas plumes of active volcanoes. The filters are quantitative for SO₂ and the halide acids when sufficient base is present to neutralize the acidic species. Extremely high concentrations of SO₂ may not be collected quantitatively since SO₂ is a relatively weak Lewis acid compared to the halide acids. The acidic oxides of Sb, As, and Se were also collected quantitatively. A particle filter preceded the impregnated filters in order to remove particles from the fumes. This system has proven effective under difficult sampling conditions and, since it is portable and light weight, it could be used for many volcanological applications where high levels of acidic gaseous phase species need to be collected.     

Optimal instrumentation of structures on flexible base for system identification

A criterion previously developed by Heredia-Zavoni and Esteva for selecting optimal sensor locations is used to analyse the optimal instrumentation of structures on soft soils. The stochastic response of a linear structural system on a flexible base is formulated for use of the criterion. The case of MDOF shear systems on flexible base, with uncertain lateral stiffness and subjected to random earthquake ground motions, is studied. The optimal location of accelerometers, the reduction of prior uncertainty on the lateral stiffness, the effects of the base flexibility, the relative influence of translation and rocking of the base, and the influence of recording noise are assessed and discussed. Copyright © 1999 John Wiley & Sons, Ltd.