Theme I: Climate science

Timely emission reductions remain the cornerstone of climate policy but pose an increasingly difficult challenge. Against this background, strategies of Climate Engineering by means of radiation management (CERM) are increasingly discussed as a potential addition to our portfolio of responses to climate change. Scientifically assessing the (in)feasibility and (un)desirability of such CERM suggestions is a highly interdisciplinary and multi-faceted endeavour. Current research at TU Delft ranges from the technical via policy to ethical aspects of CERM.

Urban areas are particularly vulnerable to extreme weather events, such as droughts, heat waves, unhealthy pollution levels, extreme precipitation and floodings. As these extremes are becoming more severe, both in intensity and in impact, as a result of global warming, there is in increasing need to better understand and predict how the built environment influences the local climate, both under present as well as for future climate conditions (i.e., urban heat island effect, dispersion of pollution, wind, and humidity).  Equally important is the need to predict and quantify the impact of urban mitigation measures (e.g. urban greening, water in the city, emission reductions) on the local urban climate.

Sea level rise is one of the main effects of climate change that the Netherlands faces. While coastal engineers and policymakers need accurate regional sea level projections, our physical understanding of how the circulation in deep oceans impacts sea level in shallow seas like the North Sea, and hence our ability to model this, is still limited. We will address this issue by studying the connections between sea level change on ocean basin scales and coastal scales, as well as the underlying dynamical processes in the ocean driving them, for present-day and for future climates. Possible approaches include the development and application of high-resolution numerical models and sophisticated analyses of observations.

Climate models are the primary tools used for generating projections of climate change under different future socio- economic scenarios and provide key input for regional decision making for a future climate resilient society. However, due to the large range of spatial and temporal scales and huge number of processes being modelled, these climate models are extremely computationally expensive to run, analyze and interpret using traditional tools and methods. Therefore, there is great interest in how machine learning (ML) might help to improve regional climate projections, especially with novel ML methodologies that are interpretable, show physical consistency, allow assimilation of observations and models across different scales, and that can handle complex and uncertain data. On the application side, these ML techniques should contribute potentially to the improvement of regional projections for the Dutch delta, where downscaling of global circulation models and uncertainties in circulation patterns are some of the main challenges.