Impact of using observed land-use properties on the representation of heatwaves in Alaro-Surfex in Belgium

Abstract

Accurate simulation of the thermal characteristics of heatwaves forms the basis of climate adaptation and risk management, particularly because heat extremes are increasing under global warming. This thesis analyzes the impact of refreshing land-use property values—more specifically Leaf Area Index (LAI), albedo, and surface roughness—on the performance of the ALARO-Surfex regional climate model in simulating heatwaves over Belgium for the summers of 2018 and 2019. Two versions of the model were compared: a reference setup using default Ecoclimap-I inputs and an updated setup using observationally based, spatially distributed parameters obtained from MODIS remote sensing data products. The analysis used both model output and gridded observational climate data regridded onto a 5 km common resolution. Surface energy fluxes and temperature variables (Average temperature, Maximum temperature, Minimum temperature) of main interest were evaluated using Mean Absolute Error (MAE) and Mean Absolute Bias (MAB) over the entire domain and within dominant cropland classes (Ecoclimap-I land covers 166 and 182). Results showed that the updated model improved the simulation of mean and minimum temperature, particularly for the 2018 heatwave, by reducing MAB by up to 1.44 °C for Minimum temperature over cropland surfaces. These improvements are crucial for accurately representing nighttime cooling processes, which play a vital role in mitigating heat stress during extreme heat events. However, the updated model exhibited challenges in simulating maximum temperatures, with MAB increasing by more than 1 °C in both years. Surface energy balance analysis revealed that reductions in albedo and LAI led to increased net radiation and sensible heat fluxes, enhancing daytime heating. At the same time, reduced latent heat fluxes limited evaporative cooling, contributing to the overestimation of maximum temperatures. Evaluation against surface energy balance measurements from eddy covariance data within the domain further supported these findings: the novel model improved Minimum temperature simulation but degraded Maximum temperature performance significantly. These results highlight the complex trade-offs introduced by newer land surface characteristics and the importance of land-atmosphere interactions for regional heatwave simulations. The study concludes that while the incorporation of observational land-use updates enhances model realism and improves the simulation of certain temperature components, further improvements, perhaps through land-cover-specific parameterizations, are required to introduce a more balanced energy flux representation and avoid unintentional biases in extreme temperature simulations. This research contributes to the ongoing refinement of regional climate modeling approaches and identifies the need for improved model accuracy to enhance confidence in climate projections. By improving the accuracy of heatwave simulations, this study supports disaster risk reduction efforts and informs climate adaptation strategies, such as urban planning, agricultural optimization, and early warning systems. Ultimately, the findings emphasize the need for continued advancements in regional climate modeling to better anticipate and mitigate the impacts of future heatwave events. The findings of this study have significant implications for climate adaptation policy and disaster risk management, particularly in nations like Belgium, which are increasingly exposed to heatwaves due to climate change. Heatwaves are a multi-dimensional threat that impacts human health, agricultural production, energy systems, and urban infrastructure. Realistic modeling of heatwave behavior, as demonstrated in the current study, is thus essential to develop early warning systems and targeted interventions to mitigate related risks. For instance, the advanced simulation of nighttime thermal cooling processes (minimums) can be applied to inform public health policy, e.g., to ascertain locales where vulnerable populations may be placed at higher risk under prolonged heat episodes. Similarly, the traded-off simulations of daytime highs suggest the need for further parameterized improvements in an effort to better simulate daytime heating trends, which are critical to urban planning strategy aimed at reducing exposure to extreme heat. Urban heat island effect, wherein cities, particularly, are subject to the exaggeration of temperature extremes in urban cities, is minimized by incorporating observationally derived land-use data into regional climate models. Urban planners are then able to design heat-resilient cities through augmenting vegetation cover, applying reflective surfaces, and optimizing water consumption. These measures not only enhance urban heatwave resilience but also contribute to wider sustainability goals such as reducing carbon emissions and conserving water resources. Furthermore, the study emphasizes the importance of integrating climate science with disaster risk reduction programs like the Sendai Framework for Disaster Risk Reduction (2015–2030), which emphasizes upgraded risk analysis, early warning, and interventions in community resilience development. By bridging the gap between technological innovation and real-world applications, this research opens doors to stronger societies and ecosystems when it comes to increasing climate challenges. Although this study demonstrates potential from employing satellite-derived land-use properties, further research needs to be done to enhance the accuracy of the model. Significant areas consist of parameterizing land cover to adjust biases in the simulation of maximum temperature and timescales and climatic zones. Incorporating regional climate models with hydrological and socioeconomic models would also be a more complete method of appreciating heatwave impacts. Besides, mitigating uncertainties with sensitivity analyses and ensemble modeling is still crucial. These tasks can capitalize on this thesis for enhancing the representation of heatwaves in climate models and facilitating evidence-based decision-making for a sustainable future. Beyond the Belgian case study, the methodological insights from this thesis hold broader relevance for regions facing recurrent or intensifying heatwaves. By demonstrating how satellite-derived land-use properties can improve regional climate model performance, this research provides a transferable framework that can be applied in other European contexts and globally. Integrating high-resolution land surface data into climate modeling not only enhances hazard assessment but also strengthens the foundation for transboundary adaptation initiatives, such as those coordinated under the European Green Deal and international disaster risk frameworks. Thus, the outcomes of this study extend beyond academic advancement to inform practical strategies for climate resilience, aligning scientific innovation with urgent policy needs in an era of accelerating climate extremes.