Tim Van de Voorde

Comparing Different Approaches for Mapping Urban Vegetation Cover from Landsat ETM+ Data: A Case Study on Brussels

Urban growth and its related environmental problems call for sustainable urban management policies to safeguard the quality of urban environments. Vegetation plays an important part in this as it provides ecological, social, health and economic benefits to a citys inhabitants. Remotely sensed data are of great value to monitor urban green and despite the clear advantages of contemporary high resolution images, the benefits of medium resolution data should not be discarded. The objective of this research was to estimate fractional vegetation cover from a Landsat ETM+ image with sub-pixel classification, and to compare accuracies obtained with multiple stepwise regression analysis, linear spectral unmixing and multi-layer perceptrons (MLP) at the level of meaningful urban spatial entities. Despite the small, but nevertheless statistically significant differences at pixel level between the alternative approaches, the spatial pattern of vegetation cover and estimation errors is clearly distinctive at neighbourhood level. At this spatially aggregated level, a simple regression model appears to attain sufficient accuracy. For mapping at a spatially more detailed level, the MLP seems to be the most appropriate choice. Brightness normalisation only appeared to affect the linear models, especially the linear spectral unmixing.

Quantifying uncertainty in remote sensing-based urban land-use mapping

Abstract Land-use/land-cover information constitutes an important component in the calibration of many urban growth models. Typically, the model building involves a process of historic calibration based on time series of land-use maps. Medium-resolution satellite imagery is an interesting source for obtaining data on land-use change, yet inferring information on the use of urbanised spaces from these images is a challenging task that is subject to different types of uncertainty. Quantifying and reducing the uncertainties in land-use mapping and land-use change model parameter assessment are therefore crucial to improve the reliability of urban growth models relying on these data. In this paper, a remote sensing-based land-use mapping approach is adopted, consisting of two stages: (i) estimating impervious surface cover at sub-pixel level through linear regression unmixing and (ii) inferring urban land use from urban form using metrics describing the spatial structure of the built-up area, together with address data. The focus lies on quantifying the uncertainty involved in this approach. Both stages of the land-use mapping process are subjected to Monte Carlo simulation to assess their relative contribution to and their combined impact on the uncertainty in the derived land-use maps. The robustness to uncertainty of the land-use mapping strategy is addressed by comparing the most likely land-use maps obtained from the simulation with the original land-use map, obtained without taking uncertainty into account. The approach was applied on the Brussels-Capital Region and the central part of the Flanders region (Belgium), covering the city of Antwerp, using a time series of SPOT data for 1996, 2005 and 2012. Although the most likely land-use map obtained from the simulation is very similar to the original land-use map – indicating absence of bias in the mapping process – it is shown that the errors related to the impervious surface sub-pixel fraction estimation have a strong impact on the land-use maps uncertainty. Hence, uncertainties observed in the derived land-use maps should be taken into account when using these maps as an input for modelling of urban growth.

Inferring urban land use using the optimised spatial reclassification kernel

In the 1990s, promising results in land-use classification were obtained by kernel-based contextual classification algorithms. Soon, however, it was recognised that kernel-based reclassifiers have important shortcomings and research instead focused on object-based image analysis. This study proposes a solution to two of the most important drawbacks of kernel-based reclassifiers: (1) the use of kernels tends to smooth boundaries between discrete land-use/land-cover parcels, and (2) it is difficult to determine a priori the optimum kernel size of the classifier. The Spatial Reclassification Kernel (SPARK) algorithm has been adapted in order to automatically optimise the kernel size depending on the spatial variation found in the neighbourhood of a pixel to be classified, resulting in the Optimised SPARK (OSPARK) algorithm. The performance of SPARK and OSPARK for land-use classification has been evaluated for the Dublin urban area (Ireland), using a Landsat TM image. The MOLAND land-use map of 1990 was used as a reference. Results show that the use of optimal kernel sizes instead of fixed kernel sizes reduces the omission and commission errors for most land-use classes.

Quantifying intra-urban morphology of the Greater Dublin area with spatial metrics derived from medium resolution remote sensing data

Spatial metrics derived from satellite imagery are useful measures to quantify structural characteristics of expanding cities, and can provide indications of functional land use types. Images of medium resolution are cheap, widely available and are often part of extensive historic archives. Their lower resolution, on the other hand, inhibits studying urban morphology and change processes at a more detailed, intra-urban level. In this study, we develop spatial metrics for use on continuous sealed surface data produced by a sub-pixel classification of Landsat ETM+ imagery. The metrics characterise the shape of the cumulative frequency distribution of the estimated sub-pixel fractions within a building block by fitting an exponential and a sigmoid function with a least-squares approach. A classification tree is then used to relate the metric variables to urban land-use classes selected from the European MOLAND topology. This approach shows promising results, but still needs improvement which may be achieved by including spatially explicit metrics in the analysis.

Using remote sensing derived spatial metrics for the calibration of land-use change models

More than ever before, planners and policy makers need tools to anticipate and assess the impact of their decisions on the spatial system that they are to manage. A growing number of high resolution models is currently being developed for this purpose. The calibration of these models remains a major challenge. Typically the required time series of land-use maps based on identical and consistent mapping methodologies, legends and scales are missing. The availability of images from earth observation satellites is much larger. However, conventional remote sensing based land-use classifications result in land cover maps, based on reflective properties of the surface, rather than land-use maps representing the functional classes needed for urban land-use change modeling. Recently, landscape metrics or spatial metrics have been introduced in the field of urban land-use mapping and modeling to characterize the spatial dynamics of such systems. The question raised in the study presented is whether spatial metrics directly applied to remote sensing images can be used to calibrate and validate land-use models of urban systems. The underlying hypothesis is that a methodology can be developed which enables to calculate metrics on both the remote sensing image and the predicted land-use map, which quantify the same distinguishing spatial structures at some level of abstraction. The study demonstrates the potential of spatial metrics to simplify and speed up the calibration procedures in so far that the development of land-use maps could be avoided.

Full Hierarchic Versus Non-Hierarchic Classification Approaches for Mapping Sealed Surfaces at the Rural-Urban Fringe Using High-Resolution Satellite Data

Since 2008 more than half of the world population is living in cities and urban sprawl is continuing. Because of these developments, the mapping and monitoring of urban environments and their surroundings is becoming increasingly important. In this study two object-oriented approaches for high-resolution mapping of sealed surfaces are compared: a standard non-hierarchic approach and a full hierarchic approach using both multi-layer perceptrons and decision trees as learning algorithms. Both methods outperform the standard nearest neighbour classifier, which is used as a benchmark scenario. For the multi-layer perceptron approach, applying a hierarchic classification strategy substantially increases the accuracy of the classification. For the decision tree approach a one-against-all hierarchic classification strategy does not lead to an improvement of classification accuracy compared to the standard all-against-all approach. Best results are obtained with the hierarchic multi-layer perceptron classification strategy, producing a kappa value of 0.77. A simple shadow reclassification procedure based on characteristics of neighbouring objects further increases the kappa value to 0.84.

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

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.

Improving the calibration of the MOLAND urban growth model with land-use information derived from a time-series of medium resolution remote sensing data

Calibrating land-use change models requires a time-series of reliable and consistent land-use maps, which are often not available. Medium resolution satellite images have a temporal and spatial resolution that is ideally suited for model calibration, and could therefore be an important information source to improve the performance of land-use change models. In this research, a calibration framework based on remote sensing data is proposed for the MOLAND model. Structural land-use information was first inferred from the available medium resolution satellite images by applying supervised classification at the level of predefined regions using metrics that describe the distribution of sub-pixel estimations of artificial sealed surfaces. The resulting maps were compared to the model output with a selected set of spatial metrics. Based on this comparison, the model was recalibrated according to five scenarios. While the selected metrics generally demonstrated a low sensitivity to changes in model parameters, some improvement was nevertheless noted for one particular scenario.

Mapping sealed surfaces from CHRIS/Proba data: A multiple endmember unmixing approach

Previous work on spectral unmixing of medium-resolution multispectral data for mapping of sealed surfaces has pointed out the limitations of the approach, which are mostly related to the confusion between sealed surface materials and spectrally similar non-artificial land-cover types. Use of hyperspectral data may improve the accuracy of sealed surface mapping in urbanized areas. In this paper the potential of multiple endmember unmixing for sealed surface mapping from hyperspectral CHRIS/Proba data is examined using a modeling scenario based on endmembers for four major classes: grey sealed surfaces, red sealed surfaces, bare soil and vegetation. A reference database was developed for validating the sub-pixel fractions using 25 cm resolution aerial photographs. The average proportional error for sealed surfaces, vegetation and bare soil is around 15%. Defining a model selection criterion that favors the use of models with few endmembers leads to a substantial improvement of the accuracy of the unmixing.

Spatially explicit urban green indicators for characterizing vegetation cover and public green space proximity: a case study on Brussels, Belgium

Cities often have a substantial green infrastructure, which provides local ecosystem services that improve the quality of life of urban residents. These services should be explicitly addressed in u...ABSTRACT Cities often have a substantial green infrastructure, which provides local ecosystem services that improve the quality of life of urban residents. These services should be explicitly addressed in urban development policies, and areas with insufficient vegetation and limited access to public green spaces should be identified. This paper presents two spatially explicit urban green indicators that are derived using remote sensing imagery, freely available map data and spatial analysis tools from open source geospatial libraries and commercial software. The first indicator represents proportional green cover (public as well as private) in the vicinity of each building within a city. The second indicator quantifies the proximity of public green spaces as walking distances from buildings to actual park entrances. A dasymetric mapping approach was used to take spatial variations in population density into account. This allows representing the indicators from the perspective of citizens instead of buildings, which may be more meaningful for deriving statistics at city level or at the level of neighbourhoods or administrative zones. The potential use of these indicators in a planning context is discussed on a case study carried out for the city of Brussels, Belgium.