No abstract

No abstract

No abstract

ABSTRACT:Building urban growth models typically involves a process of historic calibration based on historic time series of land-use maps, usually obtained from satellite imagery. Both the remote sensing data analysis to infer land use and the subsequent mode lling of land-use change are subject to uncertainties, which may have an impact on the accuracy of future land-use predictions. Our research aims to quantify and reduce these uncertainties by means of a particle filter data assimilation approach that incorporates uncertainty in land-use mapping and land-use model parameter assessment into the calibration process. This paper focuses on part of this work, more in particular the modelling of uncertainties associated with the impervious surface cover estimation and urban land-use classification adopted in the land-use mapping approach. Both stages are submitted to a Monte Carlo simulation to assess their relative contribution to and their combined impact on the uncertainty in the derived land-use maps. The approach was applied on the central part of the Flanders region (Belgium), using a time-series of Landsat/SPOT-HRV data covering the years 1987, 1996, 2005 and 2012. Although the most likely land-use map obtained from the simulation is very similar to the original classification, it is shown that the errors related to the impervious surface sub-pixel fraction estimation have a strong impact on the land-use map's uncertainty. Hence, incorporating uncertainty in the land-use change model calibration through particle filter data assimilation is proposed to address the uncertainty observed in the derived land-use maps and to reduce uncertainty in future land-use predictions.* Corresponding author. This is useful to know for communication with the appropriate person in cases with more than one author.

<p>Soil sealing results in an increased flood risk, reduced water infiltration, higher temperatures in urban areas, decreased carbon storage and biodiversity losses. As the Flanders region in Belgium has one of the highest soil sealing rates in Europe, it is especially prone to these effects. Hence, sustainable spatial management is crucial to ensure a healthy living environment and address climate change. The Flemish government therefore wants to promote de-sealing and limit additional sealing. An example of a concrete policy measure is that by 2050 sealed surfaces in open space zoning have to be reduced by 20% compared to 2015, which corresponds with a decrease of 9 000 ha.   <br />In order to monitor the evolution of sealed surfaces, and thus the effectiveness of sealing-related policy measures, a method was developed for automatically generating annual and spatially detailed soil sealing maps of Flanders. These maps combine “known” sealing from administrative databases (buildings and transport infrastructure) with modelled sealing based on artificial intelligence. Administrative databases do not (adequately) cover parking lots, private driveways and garden terraces, which are a substantial part of the sealed area in Flanders. Hence, a machine learning model was built for deriving this remaining sealing from 25 cm resolution aerial imagery. For this purpose, an assessor manually labeled the sealed parts on a subset of the images. Based on this training set, a convolutional neural network model was used to produce a sealing probability map, which was converted to a binary modelled sealing map. Finally, a continuity correction was applied to ensure a temporally consistent result across the yearly maps.             <br />According to this method, 1 m resolution soil sealing maps were obtained for 2013 until 2021. These maps show that an additional area of sealed surfaces of approximately 15 000 ha was constructed in Flanders between 2013 and 2020. They also reveal that de-sealing takes place in Flanders every year, but also is always over-compensated by new sealing. As a result, there is a net increase of soil sealing year after year. The annual soil sealing maps also show an upward trend in open space zoning that moves in the opposite direction to the target of decreasing by 9 000 ha.</p>