Blog: Wildfires and the rising risk of ‘cascading’ natural disasters

Cascading, compound and interconnected risks are all around us. Yet, they are still rarely captured explicitly by catastrophe models, which largely focus on a single peril. In some cases, combined impacts may be partially and implicitly represented in models, given that the claims used for calibration may cover multiple perils – for example, claims from inland flooding and storm surge caused by tropical storms – but explicit understanding and acknowledgement of these interactions would support greater resilience.

In a recent report, 222 scientists across 52 countries highlighted cascading crises as the biggest threat to planetary health and future generations. The Covid-19 pandemic has exposed many examples of these complexities, not least when it collides with natural disasters. A number of climate-related hazards – such as tropical storms and flooding – have already intersected with the pandemic, testing the ability of governments to cope with compound risks. Another consideration is demand surge: demand for reconstruction material and labour after a natural disaster can typically surge 15% to 30% and last for a year, and this could be further compounded by this pandemic.

Wildfires

Wildfires provide an interesting lens to understand the real-world interactions between risks, which could make things much worse than standalone models predict. Even before the US wildfire season began, the usual mitigation strategy had already been derailed. Controlled burns were suspended in many states. Projects intended to help homeowners reduce their vulnerability were paused, despite investment in resilience (for example retrofitting homes) being worthwhile, with estimates of $1 (£0.73) spent now saving $6 in future emergencies.

In emerging economies, wildfires can have a particularly devastating impact, often largely uninsured. In 2015, fire and haze cost Indonesia $16bn - twice the cost of reconstruction following the 2004 tsunami. 26% of the costs relate to natural capital, including losses to biodiversity, with losses mostly driven by lost capacity for carbon storage. In other areas, the combined impact of Covid-19 and wildfire on indigenous populations is worrying: although only 5% of the world population, they are estimated to protect 80% of the world’s biodiversity.

Expect more cascading risks

Modern societies are now highly dependent on networks, which means that failures can easily propagate through critical infrastructure and supply chains, affecting multiple sectors and geographies. Infrastructure assets are not only exposed to risks, they also create risks and increase exposure.

The World Bank has highlighted the contribution of poor quality infrastructure to wildfire risk: in particular the lack of operations and maintenance investment for electricity transmission and distributional lines, creating greater potential for wildfire ignition.

This has led researchers at the Cascading Disasters Research Group at University College London to argue that going forward “most disasters will be cascading events to a greater or lesser extent”. Multi-hazard analyses are becoming more common, but an all-encompassing model capturing all those complex relationships remains a utopia, especially given that in many crises, human agency is one of the critical factors.

Scenario and event trees, however, can be a powerful and engaging way to investigate the vulnerability paths by which cascading impacts are propagated. This can support policy and decision making, in particular, assessing the preventive or adaptation measures that can be taken to reduce risks in advance and increase resilience of critical assets. Taking extreme events and stress-testing them, whether through quantitative modelling or qualitative scenarios is one way to build resilience to global, complex risks before deciding what to do next.

Climate change as a threat multiplier makes action even more urgent and there is no doubt that we must act. Understanding immediate impacts and ramifications will help determine “how”.