Towards a storyline approach for examining future flood risk: A Central American case study

Climate change will impact the probabilities of different weather conditions and make new weather conditions possible, with implications for societal exposure to extreme weather hazards. While there is agreement that the frequency and intensity of many hazards will increase at the global scale, there is uncertainty in the spatial distribution of the changes, which needs to be considered in assessments of future extreme weather risk. Typically, this uncertainty is quantified by exploring the range of hazard intensities across a climate model ensemble for a given climate forcing. An alternative approach is to consider the range of atmospheric circulation changes across an ensemble – the driver of much of the relevant uncertainty – and extract a limited set of “physical storylines”. Rather than viewing an ensemble as a continuum of possibilities from which percentiles can be drawn, this physical storyline approach identifies “scenarios within scenarios”, thereby enabling risk modelers to work with more tractable amounts of climate data, end users to explore a “plausible worst case”, and scientists to focus their efforts on which circulation changes might be most likely.

Here, we demonstrate a prototype storyline approach for future flood risk in Central America. We consider how the frequency and intensity of flooding might change by using a pattern-scaling approach to extract the climate signal from climate model output. As a first step towards quantifying the uncertainty in our future flood risk data, we use output from three CMIP6 models that span the range of climate sensitivities and provide different flood storylines for Central America because of their distinct precipitation and temperature trends. We show how return periods for precipitation and streamflow may change under a range of policy-relevant global warming levels, providing useful insights about future surface water and river flooding for the financial, insurance, and development sectors.