Risk modelling for human displacement: what we’ve learnt and what’s next?

Weather-related events were responsible for nearly 95 per cent of all disaster displacement recorded over the last decade. An average of 21.5 million internal displacements were triggered per year by weather related hazards. Although displacement can be a “short-term” pre-emptive evacuation measure that effectively prevent injuries or loss of lives, people whose home or livelihoods are destroyed or threatened might be forced into medium to long-term displacement.

Risk assessment for displacement could help inform anticipatory action that protects people from the harmful impacts of being displaced. Past studies used CLIMADA (CLIMate ADAptation), an open-source probabilistic natural catastrophe risk assessment platform, to estimate the displacement risk in future for weather related hazards taking in account climate change variation in intensity and frequency and population growth. The platform also enables the implementation of impact forecasting for displacement for impending tropical cyclone events, with the possibility to transfer the implementation to other hazards and impact types.

However, displacement is much more complex and context-dependent. The modelling assumptions may not be able to represent all the drivers and complex processes of displacement. From our modelling experience we will shed some light on the potential of probabilistic risk assessment for displacement, and the importance of uncertainty and sensitivity analysis that quantify the confidence of model outputs. We will identify and discuss the current scientific gaps of displacement risk modelling, and the way forward to support decision making process in mitigating displacement risk.