A Remote Sensing Based Calibration Framework for the MOLAND Urban Growth Model of Dublin

A Remote Sensing Based Calibration Framework for the MOLAND Urban Growth Model of Dublin

Tim Van de Voorde (Vrije Universiteit Brussel, Belgium), Johannes van der Kwast (Flemish Institute for Technological Research (VITO), Belgium), Frank Canters (Vrije Universiteit Brussel, Belgium), Guy Engelen (Flemish Institute for Technological Research (VITO), Belgium), Marc Binard (Université de Liège, Belgium), Yves Cornet (Université de Liège, Belgium) and Inge Uljee (Flemish Institute for Technological Research (VITO), Belgium)
DOI: 10.4018/jaeis.2012070101
OnDemand PDF Download:
No Current Special Offers


Land-use change models are useful tools for assessing and comparing the environmental impact of alternative policy scenarios. Their increasing popularity as spatial planning instruments also poses new scientific challenges, such as correctly calibrating the model. The challenge in model calibration is twofold: obtaining a reliable and consistent time series of land-use information and finding suitable measures to compare model output to reality. Both of these issues are addressed in this paper. The authors propose a model calibration framework that is supported by information on urban form and function derived from medium-resolution remote sensing data through newly developed spatial metrics. The remote sensing derived maps are compared to model output of the same date for two model scenarios using well-known spatial metrics. Results demonstrate a good resemblance between the simulation output and the remote sensing derived maps.
Article Preview


The use of land to yield goods and services is seen as one of the most substantial human alterations of the earth system (Vitousek et al., 1997). Cities therefore play an important part in the global environmental change debate as they house the majority of the earth’s population and consequently form the main stage for interactions between man and environment. This calls for forceful and effective urban management policies that are based on the notion of sustainable development. Developing and assessing such policies requires reliable and sufficiently detailed information on the urban environment and its dynamics, as well as knowledge about the causes, chronology and effects of urban change processes (Herold et al., 2002). Land-use change models are useful tools in this respect as they offer planners and policy makers a spatial laboratory for assessing and comparing the environmental impact of alternative policy scenarios. As land-use change models are more and more leaving the purely academic sphere for practical applications set in a planning context, new scientific challenges also emerge as planners require robust and reliable tools rather than models that produce questionable output that can be legally challenged (Engelen & White, 2008). Prominent among these new challenges is adequate model calibration. Calibration involves adjusting the model’s parameters to ensure its behaviour matches the observed reality, and it therefore requires a time series of land-use maps. Land-use information is, however, often lacking or unavailable for short time intervals and if available, inconsistencies in mapping methodologies and mismatches in the definition of classes and spatial resolution of maps often lead to the detection of false changes (van der Kwast et al., 2009). For a successful calibration, however, the calibration period should be sufficiently long in order to give the underlying processes in the system enough time to manifest themselves representatively. The poor availability of high quality and temporally consistent land-use maps therefore often constrains the choice of the calibration period. Because manual mapping is time-consuming and expensive, alternative means for obtaining land-use information are thus worth investigating. In this paper, we propose a remote sensing based calibration framework that potentially increases the amount of data available for historic calibration and allows improving the consistency of available land-use maps. This in turn will result in better predictions of land-use change models.

The extensive archive of medium-resolution (MR) remote sensing data (e.g., Landsat, SPOT-HRV), with satellite images dating back to the early 1970s, may provide a solution for the lack of reliable land-use information. The likelihood that the required time series of MR remote sensing images can be successfully obtained is relatively high as the ground track repeat cycle is typically less than one month. The spatial resolution of the images also seems ideally suited for model calibration as it is comparable to the cell sizes of 50 to 500 m that are typically used for land-use change modelling. However, although commonly applied image classification algorithms can readily map land cover from the reflective properties of the earth’s surface, land use is linked to socio-economic activities and can therefore not be directly inferred from spectral information (Gong et al., 1992). The spatial structure of the built-up environment is nevertheless strongly related to its functional characteristics (Barnsley & Barr, 1997). Visual interpretation of remotely sensed imagery of urban areas, for example, is based on the close link between land-use and urban morphology. This link is the basic concept behind different (semi-)automated approaches for urban mapping that either rely on structural and contextual information present in the image or on information derived from ancillary data sources (de Jong & van der Meer, 2004).

Complete Article List

Search this Journal:
Volume 13: 2 Issues (2022): Forthcoming, Available for Pre-Order
Volume 12: 4 Issues (2021)
Volume 11: 4 Issues (2020)
Volume 10: 4 Issues (2019)
Volume 9: 4 Issues (2018)
Volume 8: 4 Issues (2017)
Volume 7: 4 Issues (2016)
Volume 6: 4 Issues (2015)
Volume 5: 4 Issues (2014)
Volume 4: 4 Issues (2013)
Volume 3: 2 Issues (2012)
Volume 2: 2 Issues (2011)
Volume 1: 2 Issues (2010)
View Complete Journal Contents Listing