Impact-based forecasting for human displacement by tropical cyclones to support anticipatory humanitarian action

Tropical cyclones (TCs) displace the second-largest number of people each year among all natural hazards, following floods.  While TCs impose hardships and threaten lives, the negative impacts can be mitigated through anticipatory action such as evacuation, emergency protection, and humanitarian aid coordination. An impact-based forecast can support anticipatory action planning by providing detailed information about the numbers and locations of people at risk of displacement.

Here we introduce the first implementation of a globally consistent and regionally calibrated TC-related displacement forecast that combines the (1) TC weather forecast with (2) the spatially explicit representation of population distribution and (3) their vulnerability. Furthermore, we emphasise the importance of incorporating uncertainties from all three components in a global uncertainty analysis to reveal the full range of possible outcomes. Additionally, sensitivity analysis can help us helps us understand how the forecast outcomes depend on uncertain inputs.

We demonstrate the TC displacement forecast through a case study of storm Yasa in the Fidji in 2020. Additionally, we conduct a global uncertainty and sensitivity analysis for all recorded TC displacement events from 2017 to 2020. Our findings suggest that for longer forecast lead times, decision-making should focus more on meteorological uncertainty, while greater emphasis should be placed on the vulnerability of the local community shortly before TC landfall. The open-source code and implementations are also readily transferable to other hazards and impact types.