Derivation of robust predictor variables for modelling urban shrinkage and its effects at different scales

Currently, we observe diverging processes of growth and shrinkage in European Cities. Whereas in the 80ies and 90ies partially accelerated through the crash of the socialist system mostly urban growth and suburban development occurred in European Cities, today we find a general decline of population as well as an increase of aged people (as results of the demographic change in Europe and worldwide, Cloet 2003, Lutz 2001). These processes influence land use pattern (state of the environment) and land use changes in urban areas enormously. Land use pattern reflect the current socio-economic development of an urban area and give an idea of how the urban ecosystem is influenced by man. In doing so, for instance, surface sealing reduces the filtering and remediation capacity of soils and the water retention in general as well as minimises habitat quality for wetland species. At the same time, the ecosystem(s) provide so-called ecosystem services, benefits people obtain from ecosystems: water availability, drinking water, remediation and filtering of waste, places to settle, recreation facilities in nature and others. Their quantification enables to bring the change (availability/loss) of ecosystem services into relation with effective costs (economic sphere, Farber 2002, De Groot et al. 2002). The above mentioned population decline and related shrinkage processes will have enormous consequences on the demand and availability of ecosystem services needed to sustain a high and even increasing status of quality of life for European citizens in the next future. Therefore, the predictor variables describing on the one hand shrinkage-related land use changes and on the other its effects are most important but at the same time it is still a challenge; to extract such predictor variables from a huge catalogue of urban socio-economic and environmental indicators elaborated by many studies for different landscape types and scales; to derive relevant digital and spatially explicit data as model input to calculate the effects of land use (change) and; to validate the model results at the city and the quarter level (scale) as well as to prove the response of the (gained/released) ecosystem service (environmental quality) at the city and at quarter level (closing the circle). Here, the author will give some expressive examples showing the derivation of predictor variables for modelling peri-urban growth and inner city shrinkage as well as its effects on water balance, habitat quality (urban green network) and recreational space. Of major interest is the approach of how to tackle the problem of urban shrinkage in spatially explicit land use (change) modelling (Haase et al. 2004).