Mapping, monitoring and modelling urban areas with medium-resolution satellite imagery

Onderzoeksoutput: BookResearch

Samenvatting

The general objective of the dissertation is to exploit the advantages of mid-scale earth observation data for mapping, monitoring and modelling urban areas.
In terms of mapping, the focus lies on two elementary land cover components of cities: built-up land (i.e. "impervious surfaces") and vegetation. The disadvantage of using imagery with a lower spatial resolution for this purpose is that a single sensor observation usually covers multiple land-cover types and therefore represents a composite spectral signal. Using conventional classifiers therefore inevitably leads to classification errors and wrong estimations of areal coverage. This problem is addressed with subpixel classification. Variants of three types of models for inferring land-cover proportions from a mixed spectral response were built and validated in a multi-resolution framework. This involved the use of land-cover maps derived from high-resolution imagery as reference data. In order to ensure the reliability of such maps, post-classification techniques were used for improving map accuracy and increasing the information content. The performance of the different sub-pixels models we have tested was similar enough to prefer the most straightforward and well-known approach: linear regression analysis.
The subpixel approach was applied in a case study on Brussels, for which changes in impervious surface and vegetation cover were examined for 4 dates between 1978 to 2008. While a conventional classifier showed that at least 4500 hectares of land were consumed by urban expansion in the observed period, not all of this newly urbanised land consists of impervious surfaces given that the subpixel classification indicated a more moderate decline in vegetation cover of around 1900 hectares. These differences are due to the morphology of the new developments, which is mainly residential land use with a low built-up density. Land use is linked to socio-economic activities and can therefore not be directly inferred from spectral image information. In this thesis, the potential for inferring broadly defined land-use categories from maps showing the impervious surface distribution at sub-pixel level is examined and tested on the Greater Dublin Area. A good overall accuracy was achieved when a distinction was made between two broadly defined classes representing employment-related and residential land-use. These classes represent important drivers of urban land-use change. Changes in their spatial pattern are therefore important in the context of calibrating urban growth models.
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. In a third part of this thesis, a framework is proposed for calibrating an urban growth model with information on urban form and function that is derived from medium-resolution remote sensing data. The remote sensing derived maps were compared to model output of the same date for two growth scenarios, using well-known spatial metrics as goodness-of-fit measures. The analysis indicated that many of the selected metrics were sensitive enough to detect the differences between a reference scenario from a previous calibration exercise and a deliberately unrealistic urban growth scenario. Most of these sensitive metrics also indicated a good fit between the reference scenario and the remote sensing derived maps, which allowed a tentative suggestion of a suitable set of spatial metrics for assessing the goodness-of-fit between model output and remote sensing derived information in the calibration of urban growth models.
Originele taal-2English
UitgeverijVUBPRESS
Aantal pagina's199
ISBN van geprinte versie978-90-5487-969-5
StatusPublished - 10 nov 2011

Vingerafdruk

Duik in de onderzoeksthema's van 'Mapping, monitoring and modelling urban areas with medium-resolution satellite imagery'. Samen vormen ze een unieke vingerafdruk.

Citeer dit