Author:

Abstract:

Impervious surfaces – constructed surfaces such as buildings, roads, squares and parking lots, covered by, to some extent, impermeable materials - are inherently linked to urbanization. Although the environmental impacts of impervious cover, generally resulting from an increase in storm water runoff volumes and in higher peak discharges to the receiving water systems, are well described in the last decades, little is known about the extent of impervious cover in our cities. However, as much as impervious surfaces are linked to urbanization, residential gardens are too. Although decentralized and often overlooked by because of their private character and individually small sizes, the entire garden complex in a city generally occupies a large portion of the cities territory and its total green space area. The green structures in these private gardens have the potential to mitigate the impacts of the impervious surfaces adjacent to them. The overall objectives of this research were (1) to gain insight in the extent and growth of impervious surfaces in residential neighborhoods and (2) develop possible strategies to monitor and (3) evaluate impervious cover in urban, residential environments. For the monitoring aspects and data gathering, we mainly focused on a selection of social housing neighborhoods, as they are in fact a collection of individual structures, yet constructed and planned as a whole. This facilitates the collection and digitization of their building plans to reconstruct the situation in the year of the development. Furthermore, as their planning and construction were planned and constructed by (semi-)public agencies, they may serve as examples of the policy developed by public authorities. As urban and regional planners rely increasingly on data about the extent of imperviousness as an indicator to prepare, monitor and adjust policies and regulations, in Chapter 2 we aimed to provide some fundamental information regarding the extent and change over time of impervious elements within the urban parcel. Additionally, we tried to identify predictive variables of imperviousness both on a parcel scale and a neighborhood scale. Correlations between impervious coverage and parcel characteristics such as size, width and age can give good insights in land development regulations that can influence the area of impervious surfaces within residential parcels without the need to tightly monitor and prohibit imperviousness. In 2008, an averaged 56% of the studied social neighborhoods area was covered with impervious surfaces. The growth from an average 38% impervious cover on the building plans to this number indicates the potential increase of impervious cover after the first construction of a site and thus demonstrates the need to closely monitor such developments. New impervious surfaces were generally added close to the house and close to the street, resulting in a high connectivity of the impervious cover. As urban areas and especially residential parcels have complex configurations and diverse compositions, systematically obtaining data concerning impervious surfaces in these areas is data-intensive and time-consuming. In Chapter 3, we aimed to develop an easy to use method for the generation of high resolution maps of imperviousness in urban areas relying on a minimum of expert knowledge and input data. Several methods, and especially some object-based image analyses, were compared. The object-based approach of image classification entails a segmentation of the image into spatial objects that are internally relatively homogenous and are later classified using a nearest neighbor classification method. The objects created in the segmentation step of the object-based image analysis are richer in spectral (mean values, minimum and maximum values per band, etc.) and spatial (shape, distances, context, etc.) information compared to the single pixels they consist of. This approach resulted in higher classification accuracies than the more traditional pixel-by-pixel approach used on the same very high resolution satellite image. The incorporation of a high resolution vectorial dataset, containing street and building outlines and currently in development or yet available for every municipality in Flanders, further improved both segmentation and classification accuracy of impervious surfaces in the studied neighborhood. However, the classification accuracy was not high enough to really substitute for field survey data and therefore we encourage more research on the improvement of object-based classification and on opportunities for more data fusion.Lastly, to evaluate hydrological effects of impervious cover on a parcel, in Chapter 4 we developed a performance standard (REP – retention efficiency of a parcel) based on the runoff coefficients of possible parcel elements. We aimed the tool to have a scientific background, yet be accessible and straightforward for a layman too. The calculations of the REP tool were automated in a GIS environment to allow for easy processing and spatially explicit input in a modeled geographic database. The tool allowed evaluating the impact of different management and design opportunities as demonstrated with two scenarios: a greening scenario – adding green structure where possible and minimizing the impervious area without compromising the current use - and a business as usual scenario – extrapolating the observed growth trend of impervious cover.