Classification of the Urban Morphology for Use in Residential Location Choice Models
- Doctoral Thesis
Rights / licenseIn Copyright - Non-Commercial Use Permitted
Geo-spatial data available to researchers and practitioners has increased substantially over the last decades, forming new opportunities of analyzing and characterizing locations and their spatial environment in an objective way. Such descriptives also allow for use in discrete locations choice models which model relocation behavior and generally implement three groups of variables, representing attributes of the alternative, the decision-taker and the location. With the growing availability of spatial data on a disaggregated level, a large number of location variables have been reported in these models, which reduces their comparability and their transferability to other study areas. <br/> Residential location choice models are a well-known example of such location choice models often reported by researchers. In this thesis a wide range of these studies is reviewed in order to propose a common classification for location variables and categorizes their findings. Similar preferences are found, but different measurement methods and market segments across different study regions. It becomes obvious that the inclusion of socio-economic environment is more common than the representation of the built environment due to the availability in census data. <br/> Various studies in the field of Urban Planning and Design have given recommendations on "good urban form", suggesting that specific spatial characteristics inform the quality of an urban landscape and the way persons perceive it, respectively behave in it. This demonstrates the need to reflect such spatial characteristics through quantitative attributes when modeling spatial behavior in form of location choice. <br/> The options to characterize the urban morphology with such attributes are numerious and leave the researcher with the question which attributes are necessary to reflect characteristics of the urban morphology and how these can be calculated from the given data. In this thesis this question is approached by giving an overview of quantitative descriptions of the urban morphology. The basis is a data model that is simple enough to allow the reproducibility in any study area. These attributes are classified in multiple scales to reflect different perceptions of the urban morphology, i.e. the object, the composition, the neighborhood and the municipality. A case study of the canton of Zurich furthermore shows how these characteristics allow for definition of urban typologies for the proposed scales. Three cluster algorithms are applied to define such typologies: kmeans, kmedoid and latent class clustering. The cluster results are compared on their descriptive statistics, their consistency for different number of clusters, their spatial distribution and their interpretation in maps. The final part of the thesis evaluates the impact of the proposed variables on residential location choice for the Canton of Zurich: first the standard set of variables of the review,then the newly proposed morphological attributes and cluster outcomes. The standard set of variables proves to be of high value as initial setup for residential location choice models, but a comparable model quality is found for the alternative model definitions with morphological attributes, which furthermore give the a better feedback on planning policies. The low number of observations in the survey does not allow dedicated models for different household types, but the descriptive analysis given at the end of the thesis show a variation of preferences which recommends such modeling in further studies. Show more
External linksSearch via SFX
SubjectWAHRNEHMUNGSGEOGRAFIE; STADTPHYSIOGNOMIE + STADTBILD + STADTFORMEN (STÄDTEWESEN); MATHEMATICAL MODELING AND SIMULATION IN GEOGRAPHY; CANTON OF ZURICH (SWITZERLAND); MODELLRECHNUNG UND SIMULATION IN DER GEOGRAFIE; KANTON ZÜRICH (SCHWEIZ); URBAN PHYSIOGNOMY + URBAN FORM + URBAN SETTLEMENT FORMS (URBAN STUDIES); PERCEPTIONAL GEOGRAPHY; BEHAVIOURAL GEOGRAPHY; VERHALTENSGEOGRAFIE; WOHNLAGE + WOHNQUALITÄT (HAUSWIRTSCHAFT); CLUSTER ANALYSIS + CLASSIFICATION (MATHEMATICAL STATISTICS); CLUSTERANALYSE + KLASSIFIKATION (MATHEMATISCHE STATISTIK); HOUSING SITE + HOUSING QUALITY (HOME ECONOMICS)
Organisational unit02120 - Dep. Management, Technologie und Ökon. / Dep. of Management, Technology, and Ec.
03521 - Axhausen, Kay W. / Axhausen, Kay W.
02226 - NSL - Netzwerk Stadt und Landschaft / NSL - Network City and Landscape
02655 - Netzwerk Stadt und Landschaft D-ARCH
Related publications and datasets
Is source of: http://hdl.handle.net/20.500.11850/334861
MoreShow all metadata