It Came From the Ocean

Sci­en­tists in­ves­ti­gate how global warm­ing is tur­bocharg­ing mon­ster storms like Hur­ri­cane Florence.


AS hur­ri­cane florence ap­proached the tiny port town of Wilm­ing­ton, North Carolina, Steven Pfaff took small so­lace in the knowl­edge that there hadn’t been much rain­fall in Au­gust.

He hoped the ab­sorbent ground would soak up what­ever pun­ish­ment Florence might in­flict; when Hur­ri­cane Matthew ar­rived in 2016, the ground was al­ready soggy, and the winds eas­ily top­pled trees in the loose earth.

Pfaff hun­kered down in the of­fices of the Na­tional Weather Ser­vice (NWS), where he is a warn­ing co­or­di­na­tion me­te­o­rol­o­gist. There were shut­ters on the win­dows and a tor­nado shel­ter. But the 100-miles-per-hour wind was not the big prob­lem. As Florence made land­fall, it slowed to a crawl and dumped tor­ren­tial rains. From his desk, he could hear the sounds of crack­ing wood as the trees gave way. The of­fice smelled like a wood­shop.

“The ground quickly sat­u­rated,” he says. “That buf­fer was im­me­di­ately wiped away. We went from rivers well below flood stage to now we’re talk­ing his­toric lev­els in some places. That’s how in­tense this rain was.”

An­thony Nor­ris watched the down­pour from nearby El­iz­a­beth­town. As deputy fire chief, he’d seen his share of bad weather, and Florence didn’t im­press him at first. “It seemed like the nor­mal storm that comes through,” he says. Then one road started to wash out after an­other, un­til the town was iso­lated—no one could come in or out.

The num­bers re­leased later by the NWS tell the story: more than 26 inches of rain­fall in Wilm­ing­ton, nearly 36 inches in El­iz­a­beth­town—about half a year’s worth of pre­cip­i­ta­tion. In a week, more than 8 tril­lion gal­lons of rain fell on North Carolina. The storm put about 10,000 peo­ple in shel­ters and left hun­dreds of thou­sands with­out power. At least 37 died in the storm.

Storms that drop that much rain are rare, but not as rare as they used to be. Even in just the past few weeks, sev­eral big storms have de­fied any no­tion of or­di­nary. As Florence struck North Carolina, a trop­i­cal dis­tur­bance dumped nearly as much rain on Texas as Florence did, al­beit over a much larger area. And Typhoon Mangkhut struck South­east Asia, killing at least 81 in the Philip­pines, many of them in land­slides that buried homes and shel­ters.

This re­cent march of storms is part of a trend that has reignited par­ti­san di­vides over cli­mate change (see Page 29). Big storms with mas­sively heavy rain­fall are on the rise. Data from the Na­tional Cli­mate Assess­ment show that be­tween 1958 and 2012 rain­fall from “heavy pre­cip­i­ta­tion events” in­creased ev­ery­where in the United States, but the east­ern half of the na­tion has ab­sorbed the brunt. Rain­fall in the North­east rose 71 per­cent, fol­lowed by the Mid­west (37 per­cent) and the South­east (27 per­cent).

The fre­quency of more po­tent storms is grow­ing, but the storms are also slow­ing down in speed, in­un­dat­ing the ground below. Last year, Hur­ri­cane Har­vey dropped nearly 50 inches of rain on some parts of Hous­ton, caus­ing $125 bil­lion in dam­age. (Hur­ri­canes Irma and Maria ar­rived shortly there­after, in­flict­ing in ex­cess of $100 bil­lion in dam­age on Florida, Puerto Rico and the Caribbean.) Rain, not wind, was the main de­struc­tive force as the storm stalled over Texas for days. Florence fol­lowed the same script. A Cat­e­gory 5 hur­ri­cane when push­ing across the At­lantic, it de­cel­er­ated as it struck land. “Just after land­fall, it was go­ing 2 miles per hour—you could run faster than the for­ward speed,” says Scott Weaver, di­rec­tor of the Na­tional Wind­storm Im­pact Re­duc­tion Pro­gram in Gaithers­burg, Mary­land. The weather ser­vice down­graded it to a trop­i­cal cy­clone in recog­ni­tion of the milder-than-ex­pected winds, but its slow pace gave the rains more time to wreak havoc.

What is Mother Na­ture try­ing to tell us? Prob­a­bly the same thing cli­mate sci­en­tists have been say­ing for years: that emis­sions of green­house

gases, mainly car­bon diox­ide from burn­ing fos­sil fu­els, will cause an in­crease in the in­ten­sity of big, wet storms. Their “cli­mate mod­els”—vast com­puter pro­grams used to pre­dict what the cli­mate is go­ing to do in 50 or 100 years—tell them this. These mod­els in­cor­po­rate our knowl­edge of how storms be­have and the sur­round­ing en­vi­ron­ment, in­clud­ing how much green­house gas the at­mos­phere con­tains, how much po­lar ice sheets have melted, how much heat the oceans hold and so forth. These mod­els are fab­u­lously com­plex, but the ba­sic physics is rel­a­tively straight­for­ward: Warmer air holds more mois­ture, which spells more rain.

The mod­els take all this data and spit out pro­jec­tions, in the same way weather mod­els pre­dict whether it will rain over the week­end, but they give their an­swers couched in sta­tis­tics rather than di­rect state­ments of causal­ity. We know that cig­a­rette smok­ing in­creases the like­li­hood of get­ting lung can­cer, but whether Un­cle Harry would have got­ten sick if he hadn’t smoked two packs of cig­a­rettes a day is im­pos­si­ble to know.

In the same way, sci­en­tists can­not de­clare that cli­mate change cre­ated Florence. Prob­a­bil­ity, in the highly po­lar­ized pol­i­tics of cli­mate change, puts sci­en­tists into a rhetor­i­cal bind. It’s hard to con­vince the pub­lic or politi­cians with pro­jec­tions about the long-term fu­ture couched as prob­a­bil­i­ties. For in­stance, the In­ter­gov­ern­men­tal Panel on Cli­mate Change says av­er­age tem­per­a­tures in the U.S. will rise any­where from 3 de­grees to 12 de­grees Fahren­heit by 2100. We’ve heard this state­ment many times be­fore.

Kevin Reed was pon­der­ing this co­nun­drum as Florence bore down on the East Coast. Reed, a cli­mate sci­en­tist at Stony Brook Univer­sity, got to­gether with Michael Wehner of Lawrence Berke­ley Lab­o­ra­tory and other col­leagues and de­cided to find out how much cli­mate change may be in­flu­enc­ing the in­ten­sity of the ap­proach­ing storm. It was a bit like try­ing to imag­ine what Un­cle Harry might have been like if he’d never cracked that first pack of Marl­boros.

What al­lowed Reed to even con­sider such an en­deavor is the grow­ing power of the com­put­ers that run these cli­mate mod­els. Un­til re­cently, cli­mate

mod­els and weather mod­els were two com­pletely dif­fer­ent beasts. Weather mod­els could look a week or so into the fu­ture, but they would choke mak­ing pre­dic­tions a year or 50 years out. Cli­mate mod­els had the op­po­site prob­lem: They could see the big pic­ture but didn’t have the res­o­lu­tion to fo­cus on spe­cific events, such as a hur­ri­cane. In re­cent years, how­ever, sci­en­tists have been able to bridge this gap. For his pur­poses, Reed took a cli­mate model and tweaked it to make pre­dic­tions about Florence.

On Septem­ber 10, while Florence was still over the At­lantic Ocean head­ing to­ward North Carolina, Reed and his col­leagues went to the U.S. Na­tional Oceanic and At­mo­spheric Ad­min­is­tra­tion (NOAA) and ob­tained reams of data de­scrib­ing ex­actly what Florence looked like at pre­cisely 8 p.m. East­ern Day­light Time—thou­sands of read­ings of tem­per­a­ture, hu­mid­ity, baro­met­ric pres­sure, wind ve­loc­ity and so forth gath­ered by satel­lites, weather bal­loons, light­houses and ships.

After plug­ging these bits of data—what com­puter geeks call “ini­tial con­di­tions”—into his cli­mate model, Reed ran the pro­gram and got a pre­dic­tion of how the storm would de­velop over the next seven days. It showed pretty much what the weather re­ports were say­ing: Florence would make land­fall on the North Carolina coast near El­iz­a­beth­town and Wilm­ing­ton, slow down and drop up­ward of 20 inches of rain.

Then they went back and ran the model again, but they used a dif­fer­ent set of ini­tial con­di­tions— ones that sim­u­lated what the cli­mate would have been like in prein­dus­trial times, circa 1850, be­fore Wil­liam Blake’s “dark Sa­tanic mills” filled Eng­land with soot, kick­ing off the In­dus­trial Revo­lu­tion that has caused so much cli­mate trou­ble since. That year, Mil­lard Fillmore be­came pres­i­dent, the U.S. pop­u­la­tion stood at 23 mil­lion, and green­house gases in the at­mos­phere amounted to about 284 parts per mil­lion (they are now at 407 ppm).

This time, the model spit out a much dif­fer­ent fore­cast. Although it showed the storm fol­low­ing pretty much the same path, rain­fall es­ti­mates were far lighter—50 per­cent less in the worst-hit parts of North Carolina. In El­iz­a­beth­town, that im­plied 17.5 inches of rain­fall rather than 35. That’s still a lot of

rain, but the dam­age would have been far less se­vere.

Reed isn’t the first one to cal­cu­late how mod­ern-day hur­ri­canes would have de­vel­oped with­out the green­house gases from in­dus­trial ac­tiv­ity that we’ve emit­ted over the years. Sci­en­tists have done sim­i­lar work for other trop­i­cal storms, with sim­i­lar re­sults. This is an emerg­ing field in cli­mate science called “at­tri­bu­tion” be­cause it is an ef­fort to iso­late spe­cific ef­fects of cli­mate change. It bor­rows ideas from epi­demi­ol­o­gists, who study risk fac­tors as­so­ci­ated with dis­eases—lead poi­son­ing’s im­pact on cog­ni­tive skills or the ef­fects of air pol­lu­tion on lung can­cer. But cli­mate at­tri­bu­tion as­signs risks to spe­cific weather events.

Cli­mate sci­en­tists started at­tri­bu­tion stud­ies after the 2003 Eu­ro­pean heat wave, which killed 35,000 peo­ple by some es­ti­mates, many of them el­derly peo­ple con­fined to apart­ments with no air con­di­tion­ing. Sci­en­tists found that cli­mate change made the heat wave twice as likely to oc­cur that year than in prein­dus­trial times. Stud­ies of Hur­ri­cane Har­vey showed that pre­cip­i­ta­tion in the Greater Hous­ton area was el­e­vated by nearly 40 per­cent be­cause of cli­mate change.

The ex­per­i­ment that the sci­en­tists con­ducted has lim­i­ta­tions. For one thing, the study has not yet run the gant­let of “peer re­view” that all stud­ies must sur­vive to be ac­cepted by the sci­en­tific com­mu­nity. Although sci­en­tists are con­fi­dent of their re­sults, they have not yet been able to quan­tify this con­fi­dence in rig­or­ous ways. Work­ing with mod­els, rather than ob­ser­va­tions, is also in­her­ently less pre­cise. For in­stance, when Reed set the ini­tial con­di­tions for the model, there was a cer­tain amount of “ad­just­ing” that goes on to get the mod­els to come out with a re­sult that makes sense. Cli­mate sci­en­tists are re­luc­tant to talk pub­licly about this kind of thing, be­cause cli­mate skep­tics have jumped on phrases like “tweak­ing the data” to cast doubt on le­git­i­mate cli­mate re­search. And since the model starts after Florence was al­ready a hur­ri­cane, it doesn’t ad­dress the more fun­da­men­tal ques­tion of whether the storm would have arisen in a world with­out global warm­ing.

The Florence ex­er­cise is in­struc­tive, how­ever. It sug­gests that cli­mate change is al­ready mak­ing the world vul­ner­a­ble to flood­ing from ex­ces­sive rain. The next step for Reed and his col­leagues is to study how well their model matched up with the ac­tual course of Florence and fig­ure out how to im­prove it for fu­ture stud­ies of this sort.

Some crit­ics have pointed out that Reed’s and other “at­tribu­tive” cli­mate mod­els are mis­lead­ing be­cause they un­der­es­ti­mate the con­tri­bu­tion of cli­mate change to ex­treme weather. Although they prob­a­bly cap­ture things like wind and rain­fall with some ac­cu­racy, im­por­tant cli­mate fac­tors such as storm surges and sea-level rise are be­yond the ken of the mod­els. As ice melts at the poles, sea lev­els have al­ready risen be­tween 4 and 8 inches in the past 100 years, and they are ex­pected to rise be­tween 1 and 4 feet by the end of the cen­tury. The east­ern United States is par­tic­u­larly vul­ner­a­ble in the com­ing decades. Due to the land sub­si­dence— yes, the ground is lit­er­ally sink­ing—the East Coast is likely to suf­fer an ad­di­tional 4 to 9 inches of sea-level rise in the same pe­riod.

These de­vel­op­ments prom­ise to make a bad sit­u­a­tion worse. As hur­ri­canes move in from the At­lantic, they tend to push wa­ter up against the East Coast and into es­tu­ar­ies, caus­ing ex­treme tem­po­rary rises in sea lev­els. Dur­ing Florence, storm surges caused spikes in sea level of as much as 20 feet in some places. Su­per­storm Sandy in 2012 caused 11-foot storm surges in New York City, the largest ever recorded in the area, caus­ing tens of bil­lions of dol­lars in dam­age that the city is still ad­dress­ing. Look­ing to­ward the end of the cen­tury, says Sci­en­tific Amer­i­can, pro­tect­ing the en­tire

East Coast from storm surges would re­quire a sea wall 16 feet high from Maine to Mi­ami. Build­ing a bor­der wall with Mex­ico would be easy by com­par­i­son.

It doesn’t help that de­vel­op­ers con­tinue to put con­dos along the coasts and build high-rises in Mi­ami and New York and res­i­den­tial neigh­bor­hoods along es­tu­ar­ies. De­vel­op­ment is a big rea­son why hur­ri­cane dam­age is so ex­pen­sive. U.S. poli­cies have “demon­stra­bly caused a huge in­crease in hur­ri­cane dam­age,” says Kerry Emanuel, a cli­mate sci­en­tist at MIT. They “have strongly sub­si­dized coastal de­vel­op­ment and pop­u­la­tion, lead­ing more peo­ple to move to risky places and place in­fra­struc­ture in harm’s way.” Un­less poli­cies are changed, he says, hur­ri­cane dam­age will con­tinue “for the fore­see­able fu­ture.”

In the mean­time, sci­en­tists are work­ing to es­tab­lish a firmer link be­tween what the­ory (cli­mate mod­els) tells them about ex­treme weather and what they ob­serve in the real world. We know for a fact that oceans are ab­sorb­ing the lion’s share of heat from cli­mate change. Mea­sure­ments show that oceans con­tain 90 per­cent of the heat that has ac­cu­mu­lated be­tween 1971 and 2010. Ac­cord­ing to the most re­cent In­ter­gov­ern­men­tal Panel on Cli­mate Change re­port, most of that heat is in the top layer of wa­ter, which is ris­ing about a 10th of a de­gree Cel­sius each decade. The en­ergy in that top layer is a big driver of storms.

The more heat the oceans ab­sorb, the more pow­er­ful these storms are go­ing to get. “We see in mod­els that it is likely that weak hur­ri­canes will be­come less fre­quent while strong hur­ri­canes will be­come more fre­quent,” says Emanuel. Adam So­bel, a cli­mate sci­en­tist at the Earth In­sti­tute at Co­lum­bia Univer­sity, agrees that cli­mate change is con­tribut­ing to these storms. “We know there’s cli­mate change, so we ex­pect to see cer­tain changes in hur­ri­canes,” he says. “And we are start­ing to see hints of those hap­pen­ing al­ready.”

How that plays out in the real world over the com­ing decades is not clear. Weather is far more com­plex than any mod­els sci­en­tists can de­vise, and nat­u­ral vari­abil­ity cre­ates a lot of noise that sci­en­tists have to sort through. The El Niño cy­cle,

for in­stance, which sends warm wa­ter slosh­ing around the Pa­cific Ocean, has a huge ef­fect on rain­fall and weather pat­terns. The fate of the Ama­zon rain­for­est, the bo­real forests of Canada and Siberia, the mon­soons of West Africa and In­dia, the Gulf Stream that car­ries warmth up the At­lantic to Europe—all of these nat­u­ral fea­tures of the cli­mate sys­tem are in­ter­re­lated, and a change in any one of them af­fects the oth­ers and how fu­ture storms are go­ing to play out.

An­other prob­lem is that weather records be­fore the 1950s are un­re­li­able and spotty—“too short and too flawed and not up to the task,” says Emanuel. Much of what we know about weather back then comes from ship records, which were rife with in­ac­cu­ra­cies and omis­sions. The lack of good his­tor­i­cal data has ham­pered the abil­ity of sci­en­tists to dis­cern trends. For in­stance, Gabriel Vec­chi and Thomas Knut­son, cli­mate sci­en­tists at the NOAA, ar­gue that the hur­ri­cane count may have gone up in re­cent decades, rather than down, as most sci­en­tists be­lieve.

The prac­tice of at­tribut­ing spe­cific weather events to cli­mate change also makes some sci­en­tists un­easy. “I think it is dan­ger­ous,” says Emanuel. Bet­ter to con­tinue to talk in the sci­en­tific lan­guage of prob­a­bil­i­ties. “Har­vey’s rain would have been pos­si­ble 50 years ago,” he says, “but far less likely.”

The main ben­e­fit of the at­tri­bu­tion stud­ies may be their pub­lic re­la­tions value. Reed and Wehner cer­tainly don’t act the way sci­en­tists are sup­posed to act. Most sci­en­tists are loath to dis­cuss the re­sults of their re­search un­til it is pub­lished in a pres­ti­gious peer-re­viewed jour­nal. Pub­lic­ity, in fact, can ac­tu­ally have a detri­men­tal ef­fect on an au­thor’s chance of get­ting pub­lished. Reed, by con­trast, spent much of the week of the storm giv­ing in­ter­views to the me­dia.

“Our view is that this was a way to en­gage the pub­lic,” he says. “Cli­mate change is of­ten viewed as a dis­tant threat. These at­tri­bu­tion stud­ies show that cli­mate change is here and is hav­ing im­pacts now.”

In their study of Florence, science was not the only ob­jec­tive of the ex­er­cise. The pri­mary goal was to get the pub­lic’s at­ten­tion. By that stan­dard, the re­search is a suc­cess.

FROM BET­TER TIMES A pho­to­graph from the Queen's Point con­dos, along the Neuse River in New Bern, North Carolina, winds up in the ʀot­sam after Hur­ri­cane Florence. Heavy rain, storm surges and high tides caused e[ten­sive ʀood­ing along the North Carolina coast.

THE AF­TER­MATH Boatswain's mate Dim­itri Ge­or­goulopou­los, a mem­ber of the U.S. Coast Guard’s punt team, searches the ʀood­wa­ters of the May­fair neigh­bor­hood of Lum­ber­ton, North Carolina, on Septem­ber 17; op­po­site page, two days later, Don­ald Trump vis­ited the scene of the disas­ter and took a walk­ing tour of re­cov­ery ef­forts in New Bern, North Carolina.

DISAS­TER SCE­NAR­IOS Rain on Myr­tle Beach, South Carolina, left; op­po­site page, clock­wise from top left: a branch falls on a Cadil­lac; store owner as­sesses dam­age in New Bern, North Carolina; High­way 70 bar­ri­cades; a col­lapsed road atop Pa­tri­cia Lake’s dam; mem­bers of Colorado Task Force 1 rest; a vol­un­teer res­cues a pet; Rock­ɿsh Creek overʀows onto a Fayet­teville road; a boat smashes a New Bern garage.

HIGHER GROUND Jimmy Shack­le­ford of Bur­gaw, North Carolina, left, trans­ports son Jim, wife Lisa and pets Izzy, Bella and Nala (in the cage) in the bucket of his trac­tor; right, vol­un­teers help res­i­dents of James City, North Carolina, es­cape their ʀooded homes.

Newspapers in English

Newspapers from USA

© PressReader. All rights reserved.