MAMUD - Measuring and modeling urban dynamics: impact on quality of life and hydrology

You are here

Context and objectives

The goal of the MAMUD project was to investigate how earth observation can contribute to a better monitoring, modelling and understanding of urban dynamics and its impacts on the urban and suburban environment. Research in the project focused on two urban areas in Europe (Dublin, Istanbul). Both recent, high- resolution data, as well as medium- resolution time series was used to improve the monitoring and modelling of urban change processes and their environmental impact, based on innovative land-use/land-cover mapping approaches, spatial metrics and spatial dynamic modelling. The most important objectives of the project were:

1. To improve the extraction of urban land-use/land-cover information and elevation data from high- and medium- resolution imagery.

2. To investigate how spatial metrics calculated from remotely sensed data may contribute to a more detailed, more objective, and more generic representation of urban form and function.

3. To examine how remote sensing imagery may complement existing, detailed land-use maps in the calibration and validation of the MOLAND urban growth model.

4. To study the impact of urban dynamics on population distribution, environmental quality and hydrological run-off in the urban/rural interface using remote sensing based approaches.

Project outcome

1. A multi- resolution strategy to obtain detailed information on urban dynamics from time-series of medium- resolution satellite data. The method is based on sub- pixel estimation of impervious surface cover, which is one of the most direct indicators of urban development, and allows accurate monitoring of urban sprawl and intra-urban change processes.

2. Two alternative approaches (kernel-based, region-based) to transform time-series of impervious surface cover data into land-use maps. Both methods succeed well in distinguishing areas with predominantly residential activity from areas that are mainly characterised by non-residential (employment related) activities.

3. A remote sensing based calibration framework for land-use change modelling based on metrics describing changes in the spatial pattern of residential and employment related activities. In the approach proposed model parameters are tuned in such a way that simulated patterns of urban growth, as described by the metrics, match with the patterns observed in the remote sensing derived maps.

4. A set of strategies to generate digital surface models from 2D imagery, using stereoscopic imagery as well as image triplets. Particular attention was paid to the development of a workflow to improve 3D surface models, extracted from satellite imagery, based on 2D vector data of building footprints.

5. An adapted version of the WetSpa model, integrating data on sub- pixel imperviousness in the calculation of potential runoff. A rainfall-runoff model was built for the Tolka catchment using remote sensing data for deriving parameters related to the LAI, and for estimating energy balance terms required for model calibration (SEBAL). Specific attention was paid to thermal sharpening by adapting the DisTrad model for use in urban areas.

6. A set of spatio-temporal environmental indicators, quantifying the effect of urban growth on the environment. The indicators proposed demonstrate the importance of expressing the impact of complex processes like urban growth by means of simple measures that allow planners, managers and policy makers to better assess and anticipate the effect of (local) policies.

7. A model for estimating population distribution, based on expected land-use change and estimated increase in impervious surface cover. Developing such models is essential for estimating impacts of increased population pressure on the environment, impacts of urbanisation on population, as well as for estimating infrastructure related costs linked to urban population growth (roads, sewage systems,...).