Modelling the Drivers of Urban Densification to Evaluate Built-up Areas Extension: A Data-Modelling Solution Towards Zero Net Land Take

Abstract

The impact of urbanization is determined by the amount of land taken and the intensity with which it is used, such as soil sealing and population density. Land take can be referred to the loss of agricultural, forest, and other semi-natural and natural land to urban and other artificial land development. It is closely linked to urban expansion. City centers play an important role to curb such land take issues in allocating the growing population through urban densification. In order to assess how built-up, environmental, and socio-economic factors impacts zero net land take, this paper aims at using Multinomial regression model (MLR) to evaluate the built-up densification. This model is built, calibrated, and validated for the area of Brussels Capital region and its peripheral Brabant’s using cadastral data. Three 100 × 100 m built-up maps are created for 2000, 2010 and 2020 where the map for 1990–2000 were used for calibration and was further validated using 2000–2010 maps. The causative factors are calibrated using MLR and validated using ROC curve and goodness of fit. The results show that areas at closer periphery of the city center with high densities have high probability for allocating further growth as they provide a broad range of facilities and local services along with an established connectivity infrastructure. This can be observed as a pragmatic solution for the policy makers and urban planners to achieve the intended policy of “zero net land take”.