Robert Haining

Integrating GIS and spatial data analysis: problems and possibilities

Abstract This article is an agreed summary of a workshop held in Sheffield between 18-20 March 1991. The focus here is on three of the themes of the workshop: the mutual benefits of closer links between geographical information systems (GIS) and the methods of spatial data analysis (SDA); the specific areas of SDA that should be linked with GIS; how the linkage should be made in practice. Directions for future research are also reviewed. The emphasis throughout is on statistical SDA and principally from the perspective of human rather than physical geography.

Outdoor Air Pollution and Stroke in Sheffield, United Kingdom: A Small-Area Level Geographical Study

Background and Purpose— Current evidence suggests that stroke mortality and hospital admissions should be higher in areas with elevated levels of outdoor air pollution because of the combined acute and chronic exposure effects of air pollution. We examined this hypothesis using a small-area level ecological correlation study. Methods— We used 1030 census enumeration districts as the unit of analysis and examined stroke deaths and hospital admissions from 1994 to 1998, with census denominator counts for people ≥45 years. Modeled air pollution data for particulate matter (PM10), nitrogen oxides (NOx), and carbon monoxide (CO) were interpolated to census enumeration districts. We adjusted for age, sex, socioeconomic deprivation, and smoking prevalence. Results— The analysis was based on 2979 deaths, 5122 admissions, and a population of 199 682. After adjustment for potential confounders, stroke mortality was 37% (95% CI, 19 to 57), 33% (95% CI, 14 to 56), and 26% (95% CI, 10 to 46) higher in the highest, relative to the lowest, NOx, PM10, and CO quintile categories, respectively. Corresponding increases in risk for admissions were 13% (95% CI, 1 to 27), 13% (95% CI, −1 to 29), and 11% (95% CI, −1 to 25). Conclusion— The results are consistent with an excess risk of stroke mortality and, to a lesser extent, hospital admissions in areas with high outdoor air pollution levels. If causality were assumed, 11% of stroke deaths would have been attributable to outdoor air pollution. Targeting policy interventions at high pollution areas may be a feasible option for stroke prevention.

Error propagation modelling in raster GIS: overlay operations

Performing data manipulations on maps that possess error as a result of the process of data collection leads to error propagation. The errors that are present in maps are modified by such operations in ways that may undermine the purposeofanalysisand lead to increased uncertainty in thevalidity ofthe conclusions that are drawn. This paper analyses how source map error propagates as a result of overlay operations. Geman and Gemans corruption model for individual source map error is used for the analysis which allows for attribute measurement error and location error that can then interact with the (true) source map geography. This paper reports theoretical results on the univariate overlay problem and then extends these results through simulation. Throughout a set of source maps and error processes are used with specified properties in order to examine in detail the interactions that can take place between the different elements of the source map structure and the error process. The paper uses ANOVA metho...

Sample surveying to estimate the mean of a heterogeneous surface: reducing the error variance through zoning

One of the major sources of uncertainty associated with geographical data in GIS arises when they are the outcome of a sampling process. It is well known that when sampling from a spatially autocorrelated homogeneous surface, stratification reduces the error variance of the estimator of the population mean. In this study, we evaluate the efficiency of different spatial sampling strategies when the surface is not homogeneous. When the surface is first-order heterogeneous (the mean of the surface varies across the map), we examine the effects of stratifying it into first-order homogeneous zones prior to the usual stratification for a systematic or stratified random sample. We investigate the effect of this form of spatial heterogeneity on the performance of different methods for estimating the population mean and its error variance. We do so by distinguishing between the real surface to be surveyed (ℜ), the sampling frame (ℑ) including the choice of zoning, and the statistical estimators (Ψ). The study shows that zoning improves estimator efficiency when sampling a heterogeneous surface. Systematic comparison provides rules of thumb for choice of sample design, sample statistics and uncertainty estimation, based on considering different spatial heterogeneities on real surfaces.

Exploring offence statistics in stockholm city using spatial analysis tools

The objective of this article is to investigate changes since the early 1980s in offence patterns for residential burglary, theft of and from cars, and vandalism in Stockholm City using methods from spatial statistics. The findings of previous Swedish studies on crime patterns and the insights provided by different theories, notably one propounded by Wikström (1991), provide a background for this study and are briefly reviewed. The analytical elements of the article are presented in two main parts. The first consists of a brief description of methodological procedures to obtain robust estimates of small-area standardized offence ratios. Attention is paid to both the spatial framework and the method of calculating rates. Standardized offence ratios (SORs) are calculated and mapped using GIS, and the Getis-Ord statistic is used to identify areas of raised incidence. The variation in a relative risk is modeled as a function of socioeconomic variables using the linear regression model, recognizing the complications raised by the spatial nature of the data. Results suggest that while there have been no dramatic changes in the geographies of these offences in Stockholm City during the last decade, there have been some shifts both in geographical patterns and in their association with underlying socioeconomic conditions.

Modelling small area counts in the presence of overdispersion and spatial autocorrelation

The problems arising when modelling counts of rare events observed in small geographical areas when overdispersion and residual spatial autocorrelation are present or anticipated are considered. Different models are presented for handling inference in this case. The different strategies are implemented using data on offender counts at the enumeration district scale for Sheffield, England and results compared. This example is chosen because previous research suggests that social processes and social composition variables are key to understanding geographical variation in offender counts which will, as a consequence, show evidence of clustering both at the scale of the enumeration district and at larger scales. This in turn leads the analyst to anticipate the presence of overdispersion and spatial autocorrelation. Diagnostic measures are described and different modelling strategies are implemented. The evidence suggests that modelling strategies based on the use of spatial random effects models or models that include spatial filters appear to work well and provide a robust basis for model inference but gaps remain in the methodology that call for further research.

Exploratory Spatial Data Analysis

The paper describes SAGE, a software system that can undertake exploratory spatial data analysis (ESDA) held in the ARC/INFO geographical information system. The aims of ESDA are described and a simple data model is defined associating the elements of ‘rough’ and ‘smooth’ with different attribute properties. The distinction is drawn between global and local statistics. SAGEs region building and adjacency matrix modules are described. These allow the user to evaluate the sensitivity of results to the choice of areal partition and measure of interarea adjacency. A range of ESDA techniques are described and examples given. The interaction between the table, map and graph drawing windows in SAGE is illustrated together with the range of data queries that can be implemented based on attribute values and locational criteria. The paper concludes with a brief assessment of the contribution of SAGE to the development of spatial data analysis.

Crime in Border Regions: The Scandinavian Case of Öresund, 1998–2001

Abstract This article compares offense patterns at two points in time in Öresund, a Scandinavian border region that spans Sweden and Denmark. The aim of the analysis is to contribute to a better understanding of the relationships between crime and demographic, socioeconomic, and land use covariates in a border area that has been targeted with long-term investments in transport. The changes effected by the construction of the Öresund bridge might be expected to have an impact on both the levels and the geographies of different offenses by creating new sites for offending and new, more vulnerable, transient groups. The article focuses on identifying and explaining changes in the geography of crime before and after the bridge was built. Spatial statistical techniques and GIS underpin the methodology employed. The article shows that there have been changes in the levels and the geography of some offenses. Crime in border regions is likely to be of growing interest in Europe as a result of European Union (EU) enlargement and increasing intra-European cross-border movement facilitated by improved communication systems.

Modelling High-intensity Crime Areas in English Cities

Police forces responsible for large metropolitan areas in England and Wales have claimed that within certain parts of their urban areas there exist high-intensity crime areas (HIAs). These are areas that raise special policing problems because of the particularly violent forms of crime sometimes found within them and because of the unwillingness or inability of the resident population to co-operate fully with the police in part because of fears for their own safety. A sample of metropolitan police forces was asked to identify the location of their HIAs and this paper reports the results of a GIS-based spatial analysis to try and model the location of these areas using census data. Three police force areas were used to develop the model. This was subsequently validated against a further set of HIA data from different police forces. The model suggests that HIAs are characterised by populations that are deprived and live at high density, and by higher levels of population turnover.

A Bayesian Approach to Modeling Binary Data: The Case of High-Intensity Crime Areas

This paper reports the fitting of a number of Bayesian logistic models with spatially structured or/and unstructured random effects to binary data with the purpose of explaining the distribution of high-intensity crime areas (HIAs) in the city of Sheffield, England. Bayesian approaches to spatial modeling are attracting considerable interest at the present time. This is because of the availability of rigorously tested software for fitting a certain class of spatial models. This paper considers issues associated with the specification, estimation, and validation, including sensitivity analysis, of spatial models using the WinBUGS software. It pays particular attention to the visualization of results. We discuss a map decomposition strategy and an approach that examines properties of the full posterior distribution. The Bayesian spatial model reported provides some interesting insights into the different factors underlying the existence of the three police-defined HIAs in Sheffield. High-intensity crime areas, or HIAs, are areas identified by urban police forces in England that experience high levels of violent, often drug-related, crime. Violence involves the use of knives and/or firearms. There may be further problems when bringing charges because of high levels of witness intimidation. The reason for this is that individuals or families resident in the neighborhood often perpetrate the crimes. HIAs therefore are more than simply areas with high levels of particular types of offenses (“hot spots”); they are areas with a particularly dangerous cocktail of violent crime perpetrated by offenders who are also resident in the area. They present particularly difficult policing problems. Craglia, Haining, and Signoretta (2001) reported the results of work into the spatial distribution of police-defined HIAs for a sample of English cities. The boundaries of HIAs were defined by senior police officers familiar with their cities. They first identified which of their basic command units (BCUs) had HIAs within them and