The Attribution of Temporally Compounding Events: A Study on North-East Kenya

In early 2016, Kenya experienced a whiplash between two opposing extreme events: extreme heat in March followed by heavy rainfall in April. In particular, the North-East region (Mandera, Wajir, Isiolo, Marsabit, and Samburu counties) endured an ‘ultra extreme’ 20-day heat event, defined using the Heat-Wave Magnitude Index daily (HWMId), followed closely by a 4-day heavy rainfall period. This type of compound event, which involves a succession of individual events, is termed a ‘temporally-compounding event’ and can be particularly devastating as the initial event ‘preconditions’ the human and physical environment, thereby exacerbating the impacts of the second event.

There is a dearth of literature on compound events in East Africa, despite their increasingly common nature. Here we present an attribution methodology to disentangle the mechanisms driving temporally-compounding events to fill this gap.  While attribution studies are still predominantly performed on individual extreme events, those which do consider compound events tend to focus on co-occurring multivariate events. The attribution of temporally-compounding events is, however, still in its infancy.

There are an additional range of factors to consider when attributing the drivers of a succession of hazards when compared to an individual extreme event. We build upon existing proposed methodologies to navigate these complicating factors, such as deciding between univariate or multivariate thresholds for event definitions, and deciding the ‘reasonable’ time interval between the cessation of the first event and the instigation of the second.

This research aims to contribute to the shared understanding of the interactions between the mechanisms driving compound events, specifically temporally-compounding events, within an East African context. This improved understanding can be used to inform locally-specific compound event definitions which can ultimately inform effective early-warning systems. By determining the relative contributions of anthropogenic climate change and natural variability on the 2016 Kenyan event, this research also hopes to lay the foundation for future attribution studies on compound events in the region.