If it’s not one thing, it’s another: the challenge of compound weather events


Researchers are trying to improve ways of predicting the devastating effects of complex natural disasters. Michael Lucy reports.


Storms sometimes don't happen in isolation, and complex weather systems are a challenge for forecasters.
Storms sometimes don't happen in isolation, and complex weather systems are a challenge for forecasters.
Marcus Yam/Los Angeles Times via Getty Images

The deadly Russian wildfires of 2010, the Brisbane floods of 2011 and the inundation of New York by Hurricane Sandy in 2012 were all “perfect storms”, devastating combinations of weather and climate conditions that no-one saw coming.

“Compound events” such as these – where drought amplifies a heatwave, or one storm comes hard on the heels of another – could have been foreseen, according to one group of climate scientists, who argue that most analysis of the hazards of climate change underestimates the real risks.

In a paper published in the journal Nature Climate Change, the group – made up of Swiss, Dutch, American and Australian researchers – make the case that “most major weather and climate-related catastrophes are caused by compound effects”, and call for turning climate risk projections upside down.

After disastrous events such as fires and floods, says Michael Leonard of the University of Adelaide in South Australia, one of the authors of the paper, people sometimes want to think they were unavoidable:

“You’ll often hear phrases like ‘We didn’t see it coming’,” he says, “when, actually, maybe we could have seen it coming. I guess there’s a public need for people to excuse themselves a bit, but from a professional point of view planners and modellers can do better.”

Climate scientists have only begun coming to grips with the problem of compound events over the last half-dozen years. As recently as 2012, a 500-page report by the Intergovernmental Panel on Climate Change (IPCC) on the risks of extreme events spent barely half a page on the topic.

No-one really knew how to approach the problem, Leonard says. There were too many variables, and cross-referencing every possible set of conditions seemed impossible: “It was in the too-hard basket.”

By 2014, however, he and some colleagues had put together a framework for thinking about compound events and the risks they pose.

“What’s happened traditionally is a top-down approach,” he explains.

“You do some climate modelling and see that temperature or rainfall is going to change, and then you spread that out over some region and see what’s going to happen.”

The top-down approach can miss connections, however. A single day of heavy rainfall might not cause a flood, but two days in the same week might, or simultaneous heavy rainfalls in different parts of the same water catchment, especially if the soil is already wet.

Leonard proposed a bottom-up approach. “You have to start with the end users,” he says, such as emergency services. “You ask what the impact is [of fires or floods], and what are the circumstances that can lead to it. And you work back from there.”

Often, he says, it’s a question of the scale on which the problem is viewed.

“Say you have a single high-pressure [weather] system,” he explains.

“You might have drought conditions, where there is a lot of heat stress, and a heat wave, and then on top of that some bushfires in the same region.

“You’ve got landowners who are stretched by the burdens of maintaining farm under drought conditions, then simultaneously there’s a nursing home stressed by the heatwaves, and then bushfire conditions that require evacuation. Those are three very different types of impact, but they all come from the same underlying drivers.”

In this case, he argues, a government agency or emergency service should take a large enough view to be aware that these three things are likely to happen together, and plan accordingly.

Zooming out even further, even an enormous event like the hot Russian summer of 2010 – which sparked vast wildfires and led to more than 55,000 deaths – was only part of a large compound weather event that also caused enormous flooding in the Indus valley, affecting tens of millions of people in Pakistan.

In principle, an international agency might one day be able to plan for the risk of such events, though that is a long way off.

“If you’re just doing Pakistani regional response, you wouldn’t care about heatwave in Russia, and vice versa,” he says.

Awareness of the significance of compound events is percolating through the scientific community. While the IPCC’s fifth Assessment Report, the major document summarising the state of the world’s climate-change knowledge, published in 2014, made no mention of compound events, things are already changing.

Kathleen McInnes is a climate scientist at Australian research body CSIRO who is working on the IPCC Special Report on the Oceans and Cryosphere in a Changing Climate, one of three special reports to be published before the next full Assessment Report.

“These special reports are initiated to address gaps in what’s been assessed in IPCC reports before,” she says.

The report she is working on will not only take compound events into account, but move a step further to look at how one extreme event can reduce people’s protection against others.

“It’s not just that the hazards can be compound – like storm surge combined with rain – the vulnerability and exposure can also be compound,” says McInnes, giving the example of a storm that washes sand away from a beach.

“If you get a second storm a week later, the beach hasn’t recovered so the waves come right in and start attacking the properties behind.”

Compound event analysis will also be incorporated into the IPCC’s sixth report, due to be published in 2022.

At the same time, Leonard and colleagues are calling for awareness of compound event risks beyond the scientific community, in practical planning for the future.

“People can compartmentalise things too easily, and rationalise that it’s too hard to think of everything that can happen with everything else,” he says.

“We’re calling for greater scrutiny. We should actually look at some of those dependencies and do a better job.”

Michael Lucy is features editor of Cosmos.
  1. http://nature.com/articles/doi:10.1038/s41558-018-0156-3
  2. https://wg1.ipcc.ch/srex/srexdownload.html
  3. https://onlinelibrary.wiley.com/doi/abs/10.1002/wcc.252
  4. https://journals.ametsoc.org/doi/10.1175/JHM-D-11-016.1
  5. https://www.ipcc.ch/report/ar5/
  6. https://www.ipcc.ch/report/srocc/
  7. https://wg1.ipcc.ch/AR6/AR6.html
Latest Stories
MoreMore Articles