Search

This chapter uses the case of Hurricane Katrina to show how an advance in geospatial data collection initially used to improve damage assessment can also provide utility throughout other phases of the emergency management cycle. A spatial video acquisition system (SVAS), employed as a disaster assessment tool because of Katrina-related response deficiencies, is described in terms of its genesis, and its application in the ongoing recovery phase of the disaster. The SVAS is discussed conceptually as a new approach for capturing the spatially and temporally dynamic urban environments of a post-disaster neighborhood; for example, collecting data that can be used as a proxy for disaster-related health vulnerability, especially psychopathology. The ability to collect such a dynamic data set has previously been lacking in post-disaster health research.

A 2-D Euclidean shortest path with obstacles (ESPO) algorithm for pedestrian navigation is developed. ESPO is a classical algorithm in the field of computational geometry. We describe some common ESPO algorithms and discuss their application in pedestrian shortest path determination based on the generation of a network of paths within a polygon with interior obstacles. This algorithm can be applied to pedestrian navigation in open spaces, such as squares, parks and big halls. Path generation is based on the Dijkstra algorithm, which is extended to solve the path planning problem not only for path and road networks but also for open spaces. The algorithm takes human preferences based on walking conditions into account and can find different paths with minimum cost for different conditions. The results of this approach are illustrated through an experimental system. Further work to integrate the algorithm into a practical pedestrian navigation system is proposed.

Video imagery can be acquired from aerial, terrestrial and marine based platforms and has been exploited for a range of remote sensing applications over the past two decades. Examples include coastal surveys using aerial video, routecorridor infrastructures surveys using vehicle mounted video cameras, aerial surveys over forestry and agriculture, underwater habitat mapping and disaster management. Many of these video systems are based on interlaced, television standards such as North America's NTSC and European SECAM and PAL television systems that are then recorded using various video formats. This technology has recently being employed as a front-line, remote sensing technology for damage assessment post-disaster. This paper traces the development of spatial video as a remote sensing tool from the early 1980s to the present day. The background to a new spatial-video research initiative based at National University of Ireland, Maynooth, (NUIM) is described. New improvements are proposed and include; low-cost encoders, easy to use software decoders, timing issues and interoperability. These developments will enable specialists and non-specialists collect, process and integrate these datasets within minimal support. This integrated approach will enable decision makers to access relevant remotely sensed datasets quickly and so, carry out rapid damage assessment during and post-disaster.

Building and maintaining modern transportation infrastructure demands considerable expenditure for any nation. These terrestrial route corridor zones include road, rail and to a lesser extent waterways. Road networks range from the large highways and motorways covering hundreds of kilometres down to smaller street networks that may only be few hundred meters in length. These route networks attract their own unique set of spatial information requirements in terms of overall management. These include transportation planning, engineering and operation. High quality, timely spatial information is required of the entire route corridor which now extends past the narrow confines of the road surface and includes the area adjacent to the road edge as well as areas above and below the road surface. Comprehensive 3D spatial information is required, not only, of the network itself but also objects occurring along these route corridors. This information can be used to address the day to day engineering problems as well as more strategic issues such as road safety, congestions management and noise modelling. LiDAR systems are widely available and now used to record data from both aerial and terrestrial survey platforms. LiDAR outputs X,Y,Z points enabling reliable 3-D measurements as well as 2.5-D geometric surface to be produced. High quality imagery is also collected from similar airborne and terrestrial mobile mapping platforms. This paper examines the integration of road survey imagery and airborne LiDAR data-streams within a GIS in order to satisfy these spatial information requirements.

Hedonic price modelling attempts to uncover information on the determinants of prices - in this case the prices are those of houses in the Greater London area for the period between 1980 and 1998. The determinants of house prices can include house attributes (such as size, type of building, age, etc.), neighbourhood attributes (such as proportion of unemployed people in the neighbourhood or local tax rates) and geographic attributes (such as distance from the city centre or proximity to various amenities) (Orford 1999). Almost all applications of hedonic price models applied to housing are in the form of multiple linear regression models where price is regressed on various attributes. The parameter estimates from the calibration of this type of regression model are assumed to yield information on the relative importance of various attributes in influencing price. One major problem with this approach is that it assumes that the determinants of prices are the same in all parts of the study area. This seems particularly illogical in this type of application where there could easily be local variations in preferences and also in supply and demand relationships. Hence, it seems reasonable to calibrate local hedonic price models rather than global ones – that is, to calibrate a model form which is flexible enough to allow the determinants of house prices to vary spatially. Geographically Weighted Regression (GWR) (Fotheringham et al. 2002) is a statistical technique that allows local calibrations and which yields local estimates of the determinants of house prices. GWR was recently used to investigate spatial variations in house price determinants across London separately for each of the years between 1980 and 1998 (Crespo et al. 2007). The result of the GWR analysis is a set of continuous localised parameter estimate surfaces which describe the geography of the parameter space. These surfaces are typically visualised with a set of univariate choropleth maps for each surface which are used to examine the plausibility of the stationarity assumption of the traditional regression and different possible causes of non-stationarity for each separate parameter (Fotheringham and al. 2002). The downside of these separate univariate visualisations is that multivariate spatial and non-spatial relationships and patterns in the parameter space can not be seen. In an attempt to counter this inadequacy, in a previous study we suggested to treat the result space of one single GWR analysis as a multivariate dataset and visually explore it (Demšar et al. 2007). The goal was to identify spatial and multivariate patterns that the separate univariate mapping could not recognise. In this paper we extend this approach with the temporal dimension: we use Geovisual Analytical exploration to investigate the spatio-temporal dynamics in a time series of GWR hedonic price models. The idea is to merge the time series of GWR result spaces (one space per year) into one single highly-dimensional spatio-temporal dataset, which we then visually explore in an attempt to uncover information about the temporal and spatio-temporal behaviour of parameter estimates of GWR and consequently of underlying geographical processes.

This paper describes the creation of a set of small-areas for the reporting of census data in the Republic of Ireland. The current areas used for reporting the results of the quinquennial population censuses are known as Electoral Divisions; they are large compared with similar reporting areas in Northern Ireland, they have widely varying populations and considerable internal social heterogeneity which makes them unsuitable for a wide variety of planning tasks. We describe an automated method of creating a suitable census geography which uses existing digital map and gazetteer data. We describe its structure and operation, validation and its application to nationwide. The areas have a prescribed minimum size, are designed to be consistently small, nest into the existing ED geography, cover the whole country, are constrained by natural boundaries, use streets as their unifying feature, and are reasonably homogenous.

Although the disaggregation of consumers is crucial in understanding the fragmented markets that are dominant in many developed countries, it is not always straightforward to carry out such disaggregation within conventional retail modelling frameworks due to the limitations of data. In particular, consumer grouping based on sampled data is not assured to link with the other statistics that are vital in estimating sampling biases and missing variables in the sampling survey. To overcome this difficulty, we propose a useful combination of spatial interaction modelling and microsimulation approaches for the reliable estimation of retail interactions based on a sample survey of consumer behaviour being linked with other areal statistics. We demonstrate this approach by building an operational retail interaction model to estimate expenditure flows from households to retail stores in a local city in Japan, Kusatsu City.

n this article, we propose a novel analysis technique for geographical data, Geo- graphically Weighted Discriminant Analysis. This approach adapts the method of Geographically Weighted Regression (GWR), allowing the modeling and prediction of categorical response variables. As with GWR, the relationship between predictor and response variables may alter over space, and calibration is achieved using a moving kernel window approach. The methodology is outlined and is illustrated with an ex- ample analysis of voting patterns in the 2005 UK general election. The example shows that similar social conditions can lead to different voting outcomes in different parts of England and Wales. Also discussed are techniques for visualizing the results of the analysis and methods for choosing the extent of the moving kernel window.

Traditional regression analysis describes a modelled relationship between a dependent variable and a set of independent variables. When applied to spatial data, the regression analysis often assumes that the modelled relationship is stationary over space and produces a global model that is supposed to describe the relationship at every location in the study area. This can be misleading, as the relationships in spatial data are often intrinsically different across space. One of the spatial statistical methods that attempts to solve this problem and explain local variation in complex relationships is Geographically Weighted Regression – GWR (Fotheringham et al. 2000).

This paper reports on the smoothing/filtering analysis of a digital surface model (DSM) derived from LiDAR altimetry for part of the River Coquet, Northumberland, UK using loess regression and the 2D discrete wavelet transform (DWT) implemented in the S-PLUS and R statistical packages. The chosen method of analysis employs a simple method to generate ‘noise’ which is then added to a smooth sample of LiDAR data; loess regression and wavelet methods are then used to smooth/filter this data and compare with the original ‘smooth’ sample in terms of RMSE. Various combinations of functions and parameters were chosen for both methods. Although wavelet analysis was effective in filtering the noise from the data, loess regression employing a quadratic parametric function produced the lowest RMSE and was the most effective.

The planning support system described in this chapter is a prototype internal migration modelling system developed for the UK’s Department of Environment, Transport and the Regions during 2000. It forms part of the Government’s drive towards evidence-based policy-making. It is designed to be run by civil servants on stand-alone desk-top computers primarily in order to gauge the likely impact on between-area migration flows of alternative economic and policy scenarios. The system is based on a two-stage representation of the migration process, the first stage predicting out-migration from each of 100 areas of the UK, and the second predicting the distribution of these migrants between destinations. A user-friendly front-end allows the alteration of the levels of the determinant variables for running scenarios and the easy visualisation of the migration impacts through tables and maps. The system is now being refined in a second phase of work for the client, which includes the preparation of a scoping report on future enhancements. The chapter outlines the policy context of this work and the client’s requirements, describes the modelling approach and the basic structure of the initial model, and presents an example of running a trial scenario.

In Britain, the performance of all state primary schools is assessed by students' attainment levels in a set of standardized tests administered to pupils at the ages of 7 and 11 (the so-called Key Stages 1 and 2, respectively). These data are analysed for 3687 schools in northern England. In particular, school performance is linked to the number of students taking the test at each school and to various socioeconomic indicators of the estimated school catchment areas. The latter are based on a geographical weighting function that links census data, an areal coverage, to school locations, a point coverage. Following a traditional global regression analysis, spatial variations in the relationships are examined with geographically weighted regression (GWR) to reveal some interesting geographical variations in the results.

The location of a facility can be examined in the context of several classical frameworks in economic geography and operation research where the facility is located such that some objective function is minimised (ie. Love et al., 1988; Ghosh and McLafferty, 1987; Wrigley, 1988; Fotheringham and O’Kelly, 1989). A frequently encountered objective, for example, is to find the optimal location of a new facility in terms of minimising the average time it takes individuals to travel to the nearest facility. That is, a new facility is added to an existing spatial distribution of facilities in order to achieve the maximum reduction in average travel times (the p-median problem). Other objective functions can of course be used to locate the new facility: it could for example, be located so that the maximum distance any individual has to travel is minimised (the minimax problem). Another slant on the problem is to simultaneously locate a set of facilities and to determine the allocation of demand to each of these facilities (the location-allocation problem). Still another is to model the choice of a facility by an individual as a probabilistic function of the attributes of each facility rather than as a deterministic one (a spatial interaction problem).

Introduction Je suis heureux de presenter cette selection d'articles prepares pour le 11e Colloque Europeen de Geographie Theorique et Quantitative qui s'est tenu au château de Durham, Royaume-Uni, du 3 au 7 septembre 1999. La serie des colloques europeens remonte a 1976. Le colloque se tient tous les deux ans et attire la creme des geographes "quantitativistes" europeens, ainsi que regulierement des voyageurs d'origines plus lointaines telles l'Australie, la Nouvelle-Zelande et l'Amerique d...

A frequent aim of data analysis is to identify relationships between pairs of variables, often after negating the effects of other variables. By far the most common type of analysis used to achieve this aim is that of regression in which relationships between one or more independent variables and a single dependent variable are estimated. In spatial analysis, the data are drawn from geographical units and a single regression equation is estimated. This has the effect of producing ‘average’ or ‘global’ parameter estimates which are assumed to apply equally over the whole region. That is, the relationships being measured are assumed to be stationary over space. Relationships which are not stationary, and which are said to exhibit spatial nonstationarity, create problems for the interpretation of parameter estimates from a regression model. It is the intention of this paper to describe a statistical technique, which we refer to as Geographically Weighted Regression (GWR), which can be used both to account for and to examine the presence of spatial non-stationarity in relationships.

This paper describes how individuals process spatial information, how geographical location can affect the method of processing, how an understanding of spatial information processing is necessary to model spatial choices accurately, and how spatial variations in the way information is processed can be measured. Both nonhierarchical and hierarchical methods of processing information are discussed with the latter being the more likely means of processing spatial information. Results from an empirical example using migration flows within the U.S. suggest that individuals use a hierarchical method of spatial ordering and that there is a spatial pattern to the degree to which spatial information is processed hierarchically. This pattern is shown to be related to the amount of spatial information available to an individual.

In spatial analysis, the 1980’s could be characterized as the decade of discrete choice modeling and, more generally, categorized data analysis. Subsequent to the pioneering work undertaken in economics, marketing and transportation, it was quickly recognized that most spatial decisions are discrete (for example, the selection of a city in which to live or a store at which to shop) and the discrete choice framework was adopted enthusiastically. One of the most popular statistical models for the analysis of discrete choices has been the multinomial logit model (MNL). In this paper we describe the added complexities space introduces into Discrete Choice Modeling and then discuss how they can be incorporated into the framework to produce more realistic spatial choice models. Our discussion is centered around the MNL model, because of its popularity, and on the role of the Independence from Irrelevant Alternatives (IIA) assumption which is shown to be a key factor in highlighting the differences between aspatial and spatial choice. We also demonstrate that in certain models where IIA is relaxed the subtleties of space are not captured sufficiently and that these subtleties can only be captured by developing models from spatial theory. Some promising beginnings in this direction are reported.

As computers become faster and more powerful, it is increasingly possible to provide better solutions to many geographical problems by developing new, computationally dependent tools for analysis and modelling. The term ‘geocomputation’ has been coined to describe the development and application of these tools. There are already indications of a significant increase in the number of research activities and funding opportunities that involve leading-edge computer-based tools and techniques, and geocomputational techniques are poised to provide novel solutions to a mix of old and new problems in spatial analysis. An increasingly important component of much geocomputational analysis is the use of GIs. These provide the platform for data storage, abstraction, analysis, and visualization. However, although GIS have successfully created an infrastructure for gathering, storing, and manipulating spatial information, these technologies are often deficient when it come to analysis and modelling. Geography needs to use, process, and add value to the massive amount of spatial information now available and researchers in GIS are well-poised to exploit the impact of new technologies for performing these tasks. The leading parallel supercomputers are today running at speeds a few thousand times faster than those provided by even the fastest personal computers and offer memory spaces soon expected to reach a Terabyte. The papers in this special issue of Tranructions in GZS demonstrate the link between computational power and GIs. They were presented at the First International Conference on Geocomputation held at the University of Leeds in 1996. The principal objective of the conference was to bring together various emergent, computer-oriented activities for the analysis and modelling of spatial data. The paper by Brookes, for example, examines the problem of identifying optimal locations of facilities on raster maps when the facilities are larger than the cell size used in the raster. A genetic algorithm is described which searches for optimal or near-optimal clusters of cells and then employs a raster GIS to evaluate the utility scores for these clusters. The results suggest that the genetic algorithm can find good solutions to locational problems when exhaustive search methods are impractical. Park and Wagner investigate a coupling of cellular automata and GIS in which the cellular automata act as an analytical engine for the GIs. Examples given on spatial diffusion and spatiotemporal processes show how dynamic spatial modelling can be handled within a GIs. Lilburne et al address the interesting issue of interoperability and GIs. As we demand more and more of GIs, particularly in the arena of spatial analysis, the question of where to draw the balance between one all-embracing GIS and the establishment of an increasing number of flexible interfaces to other systems arises. Interoperability is an attempt to recognize that some middle ground might provide the optimal future for GIs. In two very different papers, Takeyama describes his vision of Geo-Algebra to facilitate geocomputational modelling within a GIS while

This paper introduces the mechanism of diffusion-limited aggregation (DLA) as a new basis for understanding urban growth. Through DLA, urban form is related to the processes of rural-to-urban migration and contiguous growth. However, despite being based on very simple principles, DLA simulations are shown to have properties found in most urban areas such as negative density gradients and ordered chaotic structures. The paper examines variations in the simulated urban structures produced by different assumptions regarding the rural-to-urban migration mechanism. An important finding is that urban density gradients can occur independently of the generally accepted reasons for their presence. We also comment on boundary effects in the measurement of urban density gradients.

In this paper, we propose a model of growth and form in which the processes of growth are intimately linked to the resulting geometry of the system. The model, first developed by Witten and Sander and referred to as the diffusion-limited aggregation or DLA model, generates highly ramified tree-like clusters of particles, or populations, with evident self-similarity about a fixed point. The extent to which such clusters fill space is measured by their fractal dimension which is estimated from scaling relationships linking population and density to distances within the cluster. We suggest that this model provides a suitable baseline for the development of models of urban structure and density which manifest similar scaling properties. A typical DLA simulation is presented and a variety of measures of its structure and dynamics are developed. These same measures are then applied to the urban growth and form of Taunton, a small market town in South West England, and important similarities and differences with the DLA simulation are discussed. We suggest there is much potential in extending analogies between DLA and urban form, and we also suggest future research directions involving variants of DLA and better measures of urban density.

Consumer store choice results from a process whereby information on various alternatives is evaluated by the consumer prior to the selection of one of these alternatives. In the application of store choice models it is often assumed that the information-processing strategy underlying store choice is a simultaneous one in which all possible alternatives are evaluated by an individual. A competing assumption, increasingly recognized in aspatial choice, is that individuals initially evaluate clusters of alternatives and then only evaluate alternatives within a chosen cluster. The nested logit model is a well-known formulation for capturing this type of choice process. This paper adds to the above understanding of consumer spatial choice in the following ways: (i) It describes an alternative to the nested logit model, known as a competing destinations model, which can be used to model hierarchical spatial choice processes. While this model has been described previously in the geography and planning literature, this paper contains a novel derivation of it from within the random utility maximization framework. (ii) A general choice formulation is derived from which the logit, nested logit, and competing destinations models can be obtained. This allows the behavioral assumptions embedded in each model to be seen clearly and it allows a comparison of the nested logit and competing destinations formulations. The latter is shown to be preferable in most spatial choice situations. In particular, the latter can be used when restricted choice sets are fuzzy. (iii) The errors that arise when spatial choice results from a hierarchical process and when a simultaneous choice structure is assumed are demonstrated and are related to the presence of consumer competition/agglomeration effects between stores which are thought to be important in store choice.