Although environmental justice research has traditionally focused on environmental disamenities and health hazards, recent studies have begun to examine social inequities in the distribution of urban amenities such as street trees and parks that provide several direct and indirect health benefits to local residents. This thesis adds to this knowledge by evaluating distributional inequities in both distribution and access to parks in Pinellas County, the most densely populated and one of the most racially segregated counties in Florida. An important objective was to determine if neighborhoods with lower levels of park access are more likely to contain a significantly higher proportion of racial/ethnic minorities and low-income residents. The analysis uses precise locations of parks, street network data, and block group level census socio-demographic information. Parks are classified into three categories based on park size (acres). For the first research question, park service areas are constructed to determine the socio-demographic composition of residents closest to each park based on a 400-meter walking distance along the road network. Park service areas allow the calculation of potential park congestion, in acres per person, and the analysis of statistical associations between socio-demographic characteristics and park acreage. The results indicate less congested parks and higher acreage for racial/ethnic minority residents and those below poverty level, with respect to White residents and those above the poverty level. The second research question examines inequities in the geography of park access as measured through the creation of network-based buffer zones based on walking distances from each park. Statistical analysis, including basic comparisons and a multivariate least squares regression, indicate significantly lower accessibility to parks for residents who are Hispanic and 65 or more years in age. Parks are significantly more accessible to neighborhoods containing a higher proportion of individuals in poverty, vacant houses, and those within the cities of Clearwater or St. Petersburg. This research contributes to a growing body of literature on park inequity by using walking distances on local streets to define park service areas and focusing on an urban area (Pinellas County, Florida) that has not been examined in past studies of environmental justice.

The quantification and prediction of soil properties is fundamental to further understanding the Critical Zone (CZ). In this study we aim to quantify and predict soil properties within a forested catchment, Marshall Gulch, AZ. Input layers of soil depth (modeled), slope, Saga wetness index, remotely sensed normalized difference vegetation index (NDVI) and national agriculture imagery program (NAIP) bands 3/2 were determined to account for 95% of landscape variance and used as model predictors. Target variables including soil depth (cm), carbon (kg/m²), clay (%), Na flux (kg/m²), pH, and strain are predicted using multivariate linear step-wise regression models. Our results show strong correlations of soil properties with the drainage systems in the MG catchment. We observe deeper soils, higher clay content, higher carbon content, and more Na loss within the drainages of the catchment in contrast to the adjacent slopes and ridgelines.

No description available.

In this thesis I examined the impacts of variable retention harvesting on residual tree mortality and natural regeneration of white spruce [Picea glauca (Moench (Voss)] in northern Alberta. The VR was done in four overstory canopy compositions (ranging from deciduous dominated to conifer dominated) and at six rates of canopy retention (2%, 10%, 20%, 50%, 75% and 100%). After 10 years there was 32.9 % mortality of aspen (Populus tremuloides Michx.) and 16.9 % mortality of spruce in the VR cuts. Mortality of individual trees was greater with low density of trees, in the conifer stands and for trees with short live crowns, which are large and trees near machine corridors. Natural regeneration of spruce was greatest with higher availability of seed trees (>30 ha-1) and on machine corridors, where stocking reached 74%. By contrast, stocking was 14% on retention strips, when seed tree density was 11 seed trees ha-1. / Forest Biology and Management

国立情報学研究所で電子化したコンテンツを使用している。

Doctor of Philosophy / Department of Statistics / Weixing Song / In this dissertation, we proposed a new robust estimation procedure for two multivariate mixture regression models and applied this novel method to functional mapping of dynamic traits. In the first part, a robust estimation procedure for the mixture of classical multivariate linear regression models is discussed by assuming that the error terms follow a multivariate Laplace distribution. An EM algorithm is developed based on the fact that the multivariate Laplace distribution is a scale mixture of the multivariate standard normal distribution. The performance of the proposed algorithm is thoroughly evaluated by some simulation and comparison studies. In the second part, the similar idea is extended to the mixture of linear mixed regression models by assuming that the random effect and the regression error jointly follow a multivariate Laplace distribution. Compared with the existing robust t procedure in the literature, simulation studies indicate that the finite sample performance of the proposed estimation procedure outperforms or is at least comparable to the robust t procedure. Comparing to t procedure, there is no need to determine the degrees of freedom, so the new robust estimation procedure is computationally more efficient than the robust t procedure. The ascent property for both EM algorithms are also proved. In the third part, the proposed robust method is applied to identify quantitative trait loci (QTL) underlying a functional mapping framework with dynamic traits of agricultural or biomedical interest. A robust multivariate Laplace mapping framework was proposed to replace the normality assumption. Simulation studies show the proposed method is comparable to the robust multivariate t-distribution developed in literature and outperforms the normal procedure. As an illustration, the proposed method is also applied to a real data set.

In many high dimensional problems, the dependence structure among the variables can be quite complex. An appropriate use of the regularization techniques coupled with other classical statistical methods can often improve estimation and prediction accuracy and facilitate model interpretation, by seeking a parsimonious model representation that involves only the subset of revelent variables. We propose two regularized stochastic regression approaches, for efficiently estimating certain sparse dependence structure in the data. We first consider a multivariate regression setting, in which the large number of responses and predictors may be associated through only a few channels/pathways and each of these associations may only involve a few responses and predictors. We propose a regularized reduced-rank regression approach, in which the model estimation and rank determination are conducted simultaneously and the resulting regularized estimator of the coefficient matrix admits a sparse singular value decomposition (SVD). Secondly, we consider model selection of subset autoregressive moving-average (ARMA) modelling, for which automatic selection methods do not directly apply because the innovation process is latent. We propose to identify the optimal subset ARMA model by fitting a penalized regression, e.g. adaptive Lasso, of the time series on its lags and the lags of the residuals from a long autoregression fitted to the time-series data, where the residuals serve as proxies for the innovations. Computation algorithms and regularization parameter selection methods for both proposed approaches are developed, and their properties are explored both theoretically and by simulation. Under mild regularity conditions, the proposed methods are shown to be selection consistent, asymptotically normal and enjoy the oracle properties. We apply the proposed approaches to several applications across disciplines including cancer genetics, ecology and macroeconomics.

This thesis aimed to investigate and describe the influence that certain socioeconomical, demographical, and geographical variables had on the election results for the three parliamentary party groups in the 2018 Swedish parliamentary election on the municipal level. The study also aimed to compare the difference in effect of the variables between two different geographical study areas: Norrland and the Greater Stockholm area. The study has been conducted via a regression analysis.  The results indicated that income, education, population density and average age all have a noticeable influence on the election results for the different party blocks. Income was the factor with the overall largest influence on the election result. There was a difference in influence from different variables between the three different party blocks. The study also found that there was a difference in effect between Norrland and Greater Stockholm. Certain variables had more of an effect in Norrland, and vice-versa. Most notably, income and average age had the opposite effect in Norrland compared to Greater Stockholm. The reason for this is not clear, but differences in culture between the study areas may provide an explanation.

No description available.

No description available.