On the need and applications of intermediate-complexity climate risk models

Adaptation to climate change requires a timely implementation of effective adaptation actions. A major effort in this context is devoted to extreme weather and climate events that are a major driver of climate change impacts. Planning and designing effective and efficient adaptation actions ideally requires the assessment of risk posed by impactful events. Complex and conceptual risk models already cohabit in the scientific literature in the form of a disorganised ecosystem of non-homogeneous models. This study explores the concept and mathematical formalism of intermediate complexity (IC) models for assessing climate risk and their significance in supporting climate change adaptation efforts. Existing climate risk models are classified into categories of different complexity and intermediate-complexity climate risk models identified and conceptualised to fill gaps in the hierarchy. The formulation of an intermediate-complexity model for flood risk under future climate conditions is introduced, drawing inspiration from established large-scale methodologies and utilizing freely available data. The simulations combine data of hazard, exposure and vulnerability provided by the European Copernicus Services and local datasets to perform high resolution (order 100 m) risk assessments. We discuss the application of IC climate risk models as a conceptual laboratory to develop the concept of parametrizations to take into account the effect of small scale (order tents of meters) engineering actions in a risk model with a coarser resolution (typically hundreds meters or kilometers). In this sense, they will mimic traditional parametrizations used to model subgrid process for geophysical flows in numerical climate models. The use of IC models to rank the impact of extreme events simulated with climate models with the aim to replace hazard-based metrics with impact-based metrics for the definition of an extreme events is discussed. While acknowledging that these models may not encompass all factors influencing risk, the authors argue that they are a crucial component of model hierarchies, capable of promptly addressing critical knowledge gaps and providing valuable insights to decision-makers.