Kaupunkimaisuuden mittaaminen : Suomeen soveltuva alueluokittelumenetelmä

Cities are formed when a number of people, different societal functions and structures are densely concentrated in certain areas. Because of continuing urbanization and changes in society, cities will face a number of challenges in the future. Therefore it is necessary to define urban and rural areas even if dividing them strictly in two has attracted much debate over the years. To be able to evaluate for example urban development and to be able to recognize differentiation of areas, developing and testing a new classification approach was appropriate. Spatial and statistical data from various sources were used when developing the new classification method. The main source was Grid Database of Statistics Finland. Gross floor area values from Population Register Centre database were linked to the same 250 m x 250 m spatial grid. Distance calculations were performed by using Digiroad, a National Road and Street Database. YKR-villages and localities developed in Finnish Environment Institute were used in defining and categorizing chosen variables. The aim of this paper was to develop a method to be used in classification of areas according to how urban they are in nature i.e how high is the degree of urban features they contain. The motivation for the work was linked to the practical need to develop a classification which is independent of administrative boundaries but which is in further analysis applicable to municipalities as well. The method developed in this paper was based on the idea that the areas which are classified into the same urbanity class can contain different features if the intensity of the variables is high enough. The variables were chosen so that they represent different dimensions of urbanity but the emphasis was put on built environment. The key variables were: population, jobs, estimated daytime population, gross floor area and accessibility. At first all variables were reclassified into four categories according to quartile ranges calculated from YKR-locality values. An exception was the accessibility variable; the highest decile formed its highest class. Thus every square in the grid was given values 1–4 from every variable. The variables were then added together and every square was finally scored with urbanity values 1–20. Urban classification for municipalities was constructed by using these square values. By using the developed method, urbanity is described through concentrations of functions and structures. The grid data results show that the urban population share in Finland is 60–70 % depending on the interpretation. The urban population share is 60–80 % if the focus is on municipalities. The method was especially effective in making the degree of urbanity visible on a grid data, but a limitation lies in classifying municipalities: an area has still an impact on municipality ranking. Examining different variables, grid sizes and threshold values could be beneficial in the future. In addition it turned out that describing urbanity in just one scale is challenging. Despite these issues the method provides interesting insights into the urban phenomena in Finland. In further studies this kind of approach can be used in many ways because of its simplicity and because of the various modification possibilities it offers for different areal research purposes.