多伦多都市区的居住空间结构研究

本文采用因子生态分析方法剖析多伦多都市区的居住空间结构。主成分分析表明,家庭构成、年龄构成、职业构成是形成城市居住分异的主要因素。聚类分析表明,家庭构成和年龄构成的分异呈同心园格局,职业构成的分异呈扇形格局。形成这种居住分异格局的机制包括动力机制、管理机制和行为机制三个方面。     

基于GIS的西安市城市扩展与模拟研究

本文以西安市为研究区域,以2002年ETM影像和2007年TM影像遥感数据为城市扩展动态变化的主要信息来源,采用逻辑回归模型并结合GIS技术对西安市的城市空间扩展进行模拟。研究选取了道路、河流、地裂缝、地面沉降、人口和GDP等6个因子,建立了西安市城市扩展的逻辑回归模型,并根据所建立的模型对西安市的城市扩展进行了模拟,结果表明,所建立的逻辑回归模型能够较好地模拟西安市的城市扩展,对西安市城市扩展的驱动力具有较强的解释作用,其中对城市扩展影响最重要的因素为距道路距离。     

A missing link? Network analysis as an empirical approach for critical physical geography

Critical physical geography (CPG) calls for integrative research on material landscapes and the socio-political dynamics of scientific knowledge production. Network analysis, a rich tradition of tools and approaches for analyzing relational information, has seen little use in the CPG literature to date. This represents a fruitful opportunity, as many of CPG's core interests—knowledge politics, histories of scientific concepts, and ecosocial relations—can be effectively analyzed using network techniques. In this article, I argue for adapting network approaches to CPG. First, I provide an overview of various network concepts, approaches, and their origins. I then discuss bibliometric network techniques for “science mapping” including co-word, co-authorship, and citation analyses. Next, I describe discourse network analysis, a recent mixed-method approach from political science. Finally, I discuss overlaps with emerging approaches from qualitative and visual network analysis. In each section, I provide existing and hypothetical examples, as well as software and visualization techniques, that demonstrate how network approaches could add new insights to CPG and related scholarship. Linking CPG with the diverse traditions of network analysis has the potential to produce new empirical understandings and bring the field into conversation with a growing body of research that spans the social sciences, natural sciences, and humanities.     

L'ANALYSE DES CORRESPONDANCES: UNE AIDE À LA DESCRIPTION DE DONNÉES SPATIO-TEMPORELLES — APPLICATION À L'OCCUPATION DU SOL EN RÉGION WALLONNE, BELGIQUE

L'analyse des correspondances (AC) est une méthode exploratoire d'analyse multivariée destinée à représenter graphiquement et de façon synthétique les lignes et les colonnes d'un tableau de contingence. Dans la présente application, nous considerons un ensemble de tables annuelles qui ventilent la surface de la Région wallonne selon ses cinq provinces et ses classes d'occupation du sol. Ces tables annuelles sont juxtaposées et traitées simultanément par AC, de sorte qu'on peut suivre sur les graphiques l'évolution des provinces et celle des occupations. Grâce à un critère de maximisation de l'inertie temporelle, ces dynamiques sont paticulièrement bien mises en évidence en une représentation toutefois très concise.
Correspondence analysis (CA) is an exploratory multivariate technique used to represent graphically and in a synthetic way the rows and columns of a contingency table. In this application, we consider a set of annual tables that cross-classify the Walloon Region surface according to its five provinces and land-cover classes. These annual tables are juxtaposed and processed simultaneously by CA, so that one can observe on the graphical displays the temporal changes of the provinces and those of the land-cover classes. Thanks to a criterion of maximizing temporal inertia, these dynamics are particularly well emphasized in a very concise representation.