Inverse Modeling

Urban population is rapidly increasing. By com­parison to last decades, there is nowadays more pressure on planners to find and to develop solutions to problems associated with urbanization. Urban planning has mainly addressed urbanization problems by identifying and selecting a set of mea­sures to be taken in the present in order to achieve a desired future. Various mathematical models exist in the literature to support urban decision-making processes. Most of them use current data to forecast or simulate future scenarios. In this project, we pro­pose a new approach for spatial planning, where the point of departure is not current data, but a fu­ture desired by stakeholders. By doing so, the set of measures (defined by a set of key variables) to be taken in the present is derived from a future vision by an inverse model formulation.

We test the power of the proposed method using a hedonic house price model in a metropolitan area in Switzerland to investigate the negative effects of densification on house prices. In order to account for the spatial structure of key variables, we fit the model using geographically weighted regression (GWR), a popular local regression technique for spa­tial data modeling and spatial data exploration. Results show how devaluation of house prices cau­sed by densification can be economically compensated by different levels of socioeconomic, environmental or house-specific structural variables. In the context of urban planning, such new levels of key variables define the actions to be taken in the present in order to achieve a desired future in which economic losses caused by densification are compensated. We also discuss how trade-offs between key variables lead to more feasible or alternative compensation schemes from an urban planning perspective. This trade-off analysis provides planners with valuable input for decision-making processes as more solutions might be chosen to reach the same desired future scenario.

The inverse modeling approach can also be employ­ed to deal with other conflicts associated with ur­­­ba­­­nization, such as urban sprawls, undesired environmental or social changes, and land-use con­flicts. Similarly, mathematical formulations other than hedonic pricing methods can be used to relate a desired output to current data. In fact, with the use of a proper model, applications of the inverse model approach can be extended to other fields of spatial planning, such as regional, environmental, and transport planning. Inverse modeling can also be used in conjunction with Backcasting. In this way, while participative Backcasting can be employed in strategic planning, the use of quan­titative spatially explicit inverse modeling approaches might support the efforts of planners to achieve long-term goals of sustainability.