Urban areas are usually warmer than their surrounding natural areas, an effect known as the urban heat island effect. As such, they are particularly vulnerable to global warming and associated increases in extreme temperatures. Yet ensemble climate‐model projections are generally performed on a scale that is too coarse to represent the evolution of temperatures in cities. Here, for the first time, we combine unprecedented long‐term (35 years) urban climate model integrations at the convection‐permitting scale (2.8 km resolution) with information from an ensemble of general circulation models to assess temperature‐based heat stress for Belgium, a densely populated midlatitude maritime region. We discover that the heat stress increase toward the mid‐21st century is twice as large in cities compared to their surrounding rural areas. The exacerbation is driven by the urban heat island itself, its concurrence with heat waves, and urban expansion. Cities experience a heat stress multiplication by a factor 1.4 and 15 depending on the scenario. Remarkably, the future heat stress surpasses everywhere the urban hot spots of today. Our results demonstrate the need to combine information from climate models, acting on different scales, for climate change risk assessment in heterogeneous regions. Moreover, these results highlight the necessity for adaptation to increasing heat stress, especially in urban areas.

Rapid expansion of settlements and related infrastructures is a global trend that comes with severe environmental, economic, and social costs. Steering urbanization toward well-balanced compactness is thus acknowledged as an important strategic orientation in UN Sustainable Development Goal 11 (SDG-11) via the SDG-indicator “Ratio of land consumption rate to population growth rate.” The EU’s simultaneous commitment to being “a frontrunner in implementing […] the SDGs” and to striving for “no net land take until 2050” calls for relating the concepts of land consumption and land take to each other. Drawing on an EU-centred questionnaire study, a focus group and a literature review, we scrutinize definitions of land consumption and land take, seeking to show how they are interrelated, and questioning the comparability of respective indicators. We argue that conceptual clarifications and a bridging of the two notions are much needed, and that the precision required for definitions and applications is context-dependent. While approximate understandings may suffice for general communication and dissemination objectives, accurate and consistent interpretations of the discussed concepts seem indispensable for monitoring and reporting purposes. We propose ways of addressing existing ambiguities and suggest prioritizing the term land take in the EU context. Thereby, we aim to enhance conceptual clarity around land consumption and land take—a precondition for solidly informing respective policies and decisions.

Keywords: Error, Land Use, Mapping, Urban, Classification, Impact Analysis, Simulation, Processing 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 modelling 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.