FORE­CAST­ING PER­FECT STORMS

Science is giv­ing us ad­vance warn­ing of ‘com­pound events’. MICHAEL LUCY reports.

Cosmos - - Digest -

THE RUS­SIAN SPRING of 2010 was un­sea­son­ably dry. Then the sum­mer ar­rived with a blis­ter­ing heat­wave. Hun­dreds of fires sprang up in forests across the coun­try’s west. Vast tracts of the coun­try were clouded with smoke; in Moscow the air be­came al­most un­breath­able.

All told, the sum­mer left 55,000 dead and sent af­ter­shocks across the planet. Dev­as­tated farms meant wheat ex­ports were shut down, which drove up world food prices and – in some analy­ses – spurred on the Arab Spring up­ris­ings the fol­low­ing year.

Noth­ing like it had been seen be­fore. A year later the same words were used to de­scribe the cat­a­strophic Bris­bane floods that killed 35 and forced a whole town to be re­built on higher ground. Six years later we heard the same re­frain in the wake of Hur­ri­cane Har­vey, which dev­as­tated Hous­ton.

Th­ese ‘never be­fore seen events’ now have a tech­ni­cal name: ‘com­pound events’.

They are per­fect storms blown up by cas­cad­ing events that am­plify one an­other’s ef­fects.

The prob­lem is we’re do­ing a poor job of pre­dict­ing them. In the past, long-term weather records pro­vided a fair guide to fu­ture weather. We could pre­dict a lo­ca­tion’s po­ten­tial for floods, droughts, storms and heat­waves.

As the cli­mate changes, all that is go­ing out the win­dow. “His­tor­i­cal records of nat­u­ral haz­ards can no longer be our sole source of in­for­ma­tion on how fu­ture haz­ards will play out,” says Seth Wes­tra, an en­gi­neer at the Univer­sity of Ade­laide who stud­ies cli­mate risk.

Wes­tra is one of a small but grow­ing group of re­searchers who are us­ing the tools of risk anal­y­sis and phys­i­cal weather mod­el­ling to pre­dict and pre­pare for the com­pound events of the fu­ture.

The key in­sights are that global warm­ing means cli­matic events will cas­cade in new and un­prece­dented ways.

In the Rus­sian heat­wave, for ex­am­ple, the lack of rain­fall com­bined with the heat and the fires and the smog to deadly ef­fect. Hur­ri­cane Har­vey, for an­other ex­am­ple, killed more than 100 peo­ple and in­flicted US $125 bil­lion in dam­age, not be­cause it was a par­tic­u­larly big storm but be­cause it hit a high-pres­sure sys­tem that stopped it moving in­land and dis­si­pat­ing. In­stead, it looped around along the coast, spend­ing more time over Hous­ton and de­liv­er­ing more rain.

If you imag­ine nat­u­ral catas­tro­phes as be­ing doled out by the roll of in­vis­i­ble dice, study­ing his­tory has al­lowed sci­en­tists to build ta­bles of how of­ten each num­ber is likely to come up. As hu­man car­bon emis­sions warm the planet, how­ever, the weight­ing of the dice them­selves is shift­ing.

“We can’t just look at the big­gest flood we’ve ever had any more and use that as the ba­sis for plan­ning,” says Wes­tra.

In­stead, we will need to un­der­stand not only the phys­i­cal pro­cesses that drive catas­tro­phes – the rain­fall that causes flood­ing, the heat­waves that set the scene for fire – but also how they are con­nected. Till now, long-range cli­mate pro­jec­tions usu­ally fo­cussed on sin­gle vari­ables, per­haps pre­dict­ing heat­waves or rain­fall with­out study­ing the in­ter­ac­tions be­tween them and the phys­i­cal pro­cesses that tie them to­gether.

Th­ese connections – how the roll of one dice will af­fect the next – are also chang­ing. Take ex­treme heat: it is more likely to oc­cur than in the past, but it is also more likely to oc­cur at the same time as ex­tremely high or low rain­fall.

Cli­mate sci­en­tists have be­gun com­ing to grips with the prob­lem of com­pound events in the past six years or so. As re­cently as 2012, a 500-page re­port by the In­ter­gov­ern­men­tal Panel on Cli­mate Change (IPCC) con­tained barely half a page on the topic.

In 2014, how­ever, Wes­tra, along with his Univer­sity of Ade­laide col­league Michael Leonard and oth­ers, put to­gether a frame­work for think­ing about com­pound events and the risks they pose.

Ear­lier think­ing about fu­ture catas­tro­phes would of­ten take a top­down ap­proach, Leonard says, be­gin­ning with a large-scale cli­mate pro­jec­tion and then try­ing to work out its ef­fects on the ground. A bet­ter an­gle of at­tack, he says, is to think in terms of net­works: start with a cat­a­strophic re­sult like fires or floods and map the events that could lead to it.

“You work back from there,” he says, us­ing cli­mate mod­els to find the odds of the spe­cific com­bi­na­tions of con­di­tions that could trig­ger dis­as­ter.

Andy Pit­man, the head of the Cen­tre of Ex­cel­lence for Cli­mate Ex­tremes at the Univer­sity of New South Wales, says it’s tricky work.

“It’s not the ma­jor­ity of ex­treme events that cat­a­stroph­i­cally af­fect peo­ple,” says Pit­man. He cites Aus­tralia’s east coast low­pres­sure sys­tems as an ex­am­ple. If one hits Syd­ney, it fills up the dams. If an­other

hits a few days later, there will be some flood­ing be­cause the ground is al­ready wet and the dams are full.

“If you have three in suc­ces­sion,” says Pit­man, “you get cat­a­strophic flood­ing and the dams might fail.” Wa­ter re­leased from over­full dams was re­spon­si­ble for the flood­ing of Bris­bane in 2011, and a sim­i­lar prospect might be­fall Syd­ney. “So we’re moving our at­ten­tion away from the prob­a­bil­ity of one east coast low to the prob­a­bil­ity of three east-coast lows hit­ting the coast in quick suc­ces­sion.”

The next steps, he says, are to im­prove fore­cast­ing – us­ing bet­ter dy­namic mod­els of the Earth’s at­mos­phere to see what’s com­ing next. The im­pact of Hur­ri­cane Har­vey was pre­dicted by weather fore­casts in the days be­fore it hit. The Rus­sian heat­wave, on the other hand, was too big to be seen in its en­tirety much in ad­vance.

As com­puter power in­creases, so does the scale of fore­casts in time and space. In fu­ture, Pit­man says, “there’s a good chance they could cap­ture th­ese kind of things”.

Changes in how ex­treme weather events com­bine are also a con­cern for the in­sur­ance in­dus­try. The tra­di­tional gold stan­dard for in­sur­ers has been to cal­cu­late risk from past records, but this is be­gin­ning to change.

Most cli­mate mod­els don’t zoom in to the level of in­di­vid­ual prop­er­ties that might be in­sured against flood or fire, or even that of large events like cy­clones, says Kate Sim­monds, a catas­tro­phe an­a­lyst at rein­sur­ance bro­ker Wil­lis Re in Syd­ney. Timescale is the other is­sue: in­sur­ance con­tracts typ­i­cally only last 12 months, so the longer view may not be rel­e­vant for cur­rent in­sur­ance pre­mi­ums.

“There is a com­pet­i­tive ad­van­tage in think­ing fur­ther into the fu­ture,” Sim­monds says, “but you don’t want to in­cor­po­rate changes that might hap­pen af­ter 30 years be­cause then you’re go­ing to be of­fer­ing much more ex­pen­sive poli­cies.”

David Bresch, a for­mer head of the At­mo­spheric Per­ils team at rein­sur­ance gi­ant Swiss Re who is now at ETH Zurich, says in­sur­ers should in­crease prices where ap­pro­pri­ate. This would give cus­tomers an in­cen­tive to pre­pare for the fu­ture, to keep their prop­er­ties in­sur­able. In­sur­ers, he says, “would be bet­ter in­formed by tak­ing a long-term view into ac­count.”

Even in purely sci­en­tific terms, the study of com­pound events still has a long way to go. Jakob Zscheis­chler, a cli­mate sci­en­tist at ETH Zurich, is work­ing on an in­ter­na­tional project to bring to­gether cli­mate sci­en­tists, im­pact mod­ellers and statis­ti­cians.

“We need to es­tab­lish meth­ods to study dif­fer­ent types of com­pound events. We also don’t re­ally know how to eval­u­ate how well cli­mate mod­els sim­u­late com­pound events.”

Com­pound events may play an in­creased role in the IPCC’S Sixth Assess­ment Re­port – the of­fi­cial roundup of the state of cli­mate science – due to be pub­lished in 2022. So­nia Senevi­ratne, a co­or­di­nat­ing lead au­thor on the re­port’s chap­ter on ex­treme events, is pro­hib­ited from dis­cussing its con­tent but re­gards com­pound events as “a very promis­ing new area of re­search”.

Science is gear­ing up to give us bet­ter predictions of the ‘never be­fore events’ that are com­ing our way. The chal­lenge will be to pre­pare for them.

Global warm­ing means cli­matic events will cas­cade in new and un­prece­dented ways.

Newspapers in English

Newspapers from Australia

© PressReader. All rights reserved.