5
he elegantly dressed presenters walk to the lectern, looking out over the glittering auditorium filled with celebrities, stars in their fields each and every one. The presenters read the list of five nominees, pausing between the names as a film clip features the work of each . "The envelope, please," one presenter says . He peels open the sealed flap and pulls out the card while the audience waits with bated breath . Who will take home the prestigious award? Who will walk out having won the stamp of approval of his or her esteemed colleagues? "And the winner is . . . Director Kelvin Droegemeier for ARPS, developed at the University of Oklahoma's CAPS." What? Did he mean to say Kevin Costner? Was that supposed to be Universal Studios? Could he have meant Raiders of the Lost ARPS or CAPS Courageous? No . These are not movie stars, studios or feature films ; they are scientists, re- search centers and computer software . This is a mistake, right? Wrong . Here is the story . CAPS is the Center for Analysis and Prediction of Storms, OU's Science and Technology Center funded in 1989 with an 11- year National Science Foundation grant worth ap- proximately $18 million . ARPS is its award-win- ning production : the Advanced Regional Prediction System, a storm prediction software that has the potential of revolutionizing the ability to forecast severe storms and, thus, forewarn those endan- gered physically or economically by them . Kelvin Droegemeier, associate professor of meteorology the Envelope Please ... Kelvin Droegemeier and his intrepid band of storm predictors are caught in a whirlwind of national prize recognition . BY KATHRYN JENSON WHITE and the director of CAPS, has overseen this epic production . Meteorologically speaking, ARPS has taken the weather world by storm . On May 31 at Epcot, the Disney theme park in Florida devoted to science and exploration, Dis-couer MagazinenamedARPS the 1997recipientof' its Award for Technological Innovation in computer software design . A few days later, ARPS struck again, this time in Washington, D .C ., by snaring ; the Computerworld Smithsonian Award for visionary use of information technology in science . Contrary to weather myth, lightning can strike twice in the same place . Winning the scientific equivalents of the Golden Globe and the Academy Award has brought Droegemeier, his colleagues and their achievement a whirlwind of attention and praise . 1997 SUMMER 19

the Envelope Please....of `The envelope, please.' Wedidn't knowuntiljustthatmomentwhohad won." The Smithsonian event was, as might be expected, a bit less glitzy, butglamourwasstill

  • Upload
    others

  • View
    1

  • Download
    0

Embed Size (px)

Citation preview

Page 1: the Envelope Please....of `The envelope, please.' Wedidn't knowuntiljustthatmomentwhohad won." The Smithsonian event was, as might be expected, a bit less glitzy, butglamourwasstill

he elegantly dressed presenters walk to the lectern, looking out over theglittering auditorium filled with celebrities, stars in their fields each and every one.The presenters read the list of five nominees, pausing between the names as a filmclip features the work of each .

"The envelope, please," one presenter says . He peels open the sealed flap andpulls out the card while the audience waits with bated breath . Who will take homethe prestigious award? Who will walk out having won the stamp of approval of hisor her esteemed colleagues?

"And the winner is . . . Director Kelvin Droegemeier for ARPS, developed at theUniversity of Oklahoma's CAPS."

What? Did he mean to say Kevin Costner? Wasthat supposed to be Universal Studios? Could hehave meant Raiders of the Lost ARPS or CAPSCourageous? No . These are not movie stars,studios or feature films ; they are scientists, re-search centers and computer software .

This is a mistake, right? Wrong .Here is the story . CAPS is the Center for

Analysis and Prediction of Storms, OU's Scienceand Technology Center funded in 1989 with an 11-year National Science Foundation grant worth ap-proximately $18 million . ARPS is its award-win-ning production : the Advanced Regional PredictionSystem, a storm prediction software that has thepotential of revolutionizing the ability to forecastsevere storms and, thus, forewarn those endan-gered physically or economically by them . KelvinDroegemeier, associate professor of meteorology

the EnvelopePlease . . .

Kelvin Droegemeier and his intrepid bandof storm predictors are caught in a whirlwind

of national prize recognition.

BY KATHRYN JENSON WHITE

and the director of CAPS, has overseen this epicproduction .

Meteorologically speaking, ARPS has taken theweather world by storm .On May 31 at Epcot, the Disney theme park in

Florida devoted to science and exploration,Dis-couer MagazinenamedARPS the 1997recipient of'

its Award for Technological Innovationin computersoftware design . A few days later, ARPS struckagain, this time in Washington, D.C ., by snaring;theComputerworld Smithsonian Award for visionaryuse of information technology in science . Contraryto weather myth, lightning can strike twice in thesame place .

Winning the scientific equivalents ofthe GoldenGlobe and the Academy Award has broughtDroegemeier, his colleagues and their achievementa whirlwind of attention and praise .

1997 SUMMER

19

Page 2: the Envelope Please....of `The envelope, please.' Wedidn't knowuntiljustthatmomentwhohad won." The Smithsonian event was, as might be expected, a bit less glitzy, butglamourwasstill

"It's been very exciting," he says ."The Discover Magazine competitionwas billed as the `Academy Awards ofScience,' and, in fact, that's exactlywhat it looked like the way they had itstaged . There were big video screensand many celebrities : gymnast Shan-non Miller, Miss America Tara Hol-land, actorLeVar Burton, SecretaryofTransportation Federico Pena andJohn H. Gibbons, the president's sci-ence adviser . The ceremony followedthe same setup as the Oscars in termsof `The envelope, please .' We didn'tknow untiljust that moment who hadwon."

The Smithsonian event was, asmight be expected, a bit less glitzy,but glamour was still a strong force .

"It was another surprise, but in atotally different kind of situation,"Droegemeier says . "The Smithsonianhad a formal, black-tie dinner with1,200 people and an awards ceremonyMondaynight . The audience consistedof CEOs of many major corporations,which makes sense given that this isthe technology gala of the Northeast,maybe of the whole country . ThenTuesday morning we had a formalawarding at the Enid A. Haupt Gar-den at the Smithsonian Castle ."

As a result ofthis double honor, theARPS code and related materials willbecome a permanent part of theSmithsonian's collection

of key Ameri-can science and technology items andbe featured for a year in DiscoverMagazine's impressive Epcot displayarea highlightingthe magazine's tech-nology awards .

Just as winning an AcademyAwardcan have a significant impact on boxoffice receipts and reputation in thefilm community, winning these twoscience awards will bring more thanprestige to CAPS and, by extension,the University .

"We've been given two strongstamps of recognition and approval,and that will help us and our otherpotential partners in developingARPS," Droegemeier says . "Now weare at the point at which we can beginto return to the University some ofwhat it has given us as we earn rev-enues fromjointventure partnerships

20

SOONER MAGAZINE

Director Kelvin Droegemeier, above, and the meteorological team that composeOU's Center for the Analysis and Predication of Storms are receiving nationalacclaim for the development of ARPS, a storm prediction software with the potentialof revolutionizing forecasting .

with private industry . That's whatthese centers were created to do, andI think we should see a big return oninvestment in the next five years .

"These awards are independentassessments of the quality and thevision and the usefulness of our tech-nology . Both awards are geared to-ward recognizing innovative technol-ogy with a practical use for society andpotential strong economic impact .They attempt to recognize the unsungheroes . Movie stars get their due allthe time, but much is going on in sci-ence that isn't recognized . The realheroes are those who aren't recognizedon the street but who work to developthe cellular phone, the Internet andother technological wonders."

That category would include, ofcourse, ARPS, which has as its ulti-mate goal providing accurate informa-tion about the course and intensity ofsevere storms so that those in theirpathshave six to nine hours ofwarningrather than the 30 minutes now thenorm . With that much warning, livesand money in significant numberscould well be saved .

"Lightning is the most dangerousaspect ofsevere storms," Droegemeiersays . "More people are killed by light-ning than by tornadoes, but totalweather-related deaths number in thehundreds each year . The economicaspects boggle the mind . I'm writinga paper on that now . The insuranceindustry alone pays losses ofover $250million each year to commercial air-lines . I imagine that, overall, lossesrun in the neighborhood of $2 billion-$3 billion each year . Agriculture loses$40 billion-$50 billion a year . Truck-ing companies and commercial powerutilities can lose millions and mil-lions .

"American Airlines entered into athree-year research and developmentpartnership with CAPS because ofthepotential savings to the airline indus-try. We want to be able to tell them at3:30 p .m . that by 9 o'clock this stormwe're tracking will be over Dallas-Fort Worth Airport, so plan on cancel-ing flights in order to avoid costlydiversions . It costs roughly $40,000 tocancel a flight and about $150,000 todivert one . The other day, one airline

Page 3: the Envelope Please....of `The envelope, please.' Wedidn't knowuntiljustthatmomentwhohad won." The Smithsonian event was, as might be expected, a bit less glitzy, butglamourwasstill

had 70 diversions . That'sjust one day,one airport . Think of 365 days a yearand the many airports aroundthe coun-try . It adds up very, very fast ."

American Airlines did the math .Its substantial funding bought thecompany first commercial use rights .At present, the governments ofSouthKorea, Thailand and Japan are talk-ing with CAPS, as are Lockheed-Mar-

This

is going on in

tin and a variety of other privatesector and governmental entities .They are wise to do so . Statisticsfrom CAPS show that annual flood-related damage has increased in thelast 30 years by 75 percent, whenadjusted for inflation, and deaths by85 percent . The insurance industryreports that 88 percent of all cata-strophic losses industrywide from

1984 to 1993-about $45 billion incurrent dollars-resulted directlyfrom weather . The Federal Emer-gency Management Agency declared492 disasters from 1985 throughMarch 1996, ofwhich 80 percent wereweather related .

ARPS-which Droegemeier says isabout 25 percent of the way towardbeing all he and his colleagues thinkit

Movie stars get their due all the time, but muchscience that isn't recognized .

image, created with ARPS storm forecast data and a Cray T3D super computer,

shows development of a thunderstorm over Oklahoma . It covers a surface area 67

kilometers square extending 17 kilometers in altitude . The white areas show high

concentrations of combined cloud and rain water. The red areas represent columns

of high vorticity, which have the potential to spawn tornadoes. The blue ribbons

show wind velocity .

can be and doing all they think it cando-exists as a result ofthe combiningof many disciplines in pursuit of acommon goal . Mathematicians, com-puter scientists and meteorologistspooled their resources to save livesand part of the total $14 billion thatstorms cost U.S . businesses each year .

Understanding how ARPS func-tions is not easy for those unfamiliar

with supercomputing and mathemati-cal models . The problem ARPS ismeant to solve is that current radarprovides scientists only limited infor-mation about a given storm in a givenarea . Radar observes, for example,the speed of the wind parallel to theradar beam and the intensity of therain . Radar does not tell stormpredic-tors the perpendicular components ofwind speed, temperature, air pres-sure or approximately 10 other vari-ables theyrequire topredict the courseand intensity of a storm . That iswhere ARPS comes in .

This computationally superchargedprogram draws upon Doppler radarinformation, Mesonet data and thecyberpower of supercomputing equip-ment at OU's Sarkeys Energy Centerand the federally funded PittsburgSupercomputing Center . ARPS's cre-ators place the information they haveinto millions of mathematical equa-tions, then run the equations back-ward and forward up to 50 times . Eachtime they run the equations, they fine-tune data . What results is a fairlyaccurate prediction of where a stormis going to be some six hours in ad-vance of its arrival and of how it willbehave once it gets there . In essence,that is ARPS.

Behind that essence is approxi-mately eight years of work under theleadership of CAPS co-foundersDroegemeier and Douglas Lilly, now aGeorge L . Cross Research Professoremeritus of meteorology . Lilly servedas the Center's director untilDroegemeier assumed the position inJuly 1994 . Droegemeier earned a 1980OU undergraduate degree in meteo-rology and went to the University ofIllinois at Urbana-Champaign for hismaster's of science and doctoral de-grees in atmospheric science . In 1982,Lilly came to OU from the NationalCenter for Atmospheric Research . In1985, Droegemeier returned to hisalma mater as assistant professor ofmeteorology . Together, they wrotethe grant request to the National Sci-ence Foundation asking for the fundsto create CAPS and begin work devel-oping ARPS.

"Doug was an imminent theoreti-

1997 SUMMER

21

Page 4: the Envelope Please....of `The envelope, please.' Wedidn't knowuntiljustthatmomentwhohad won." The Smithsonian event was, as might be expected, a bit less glitzy, butglamourwasstill

cal meteorologist who had a visionabout the same thing as I did, butcoming from a different direction,"Droegemeier says . "He was comingfrom the theoretical side and focusingon the difficulty of data retrieval . Myexpertise was in numerical modelingand high performance computing ap-plied to meteorology . Together we

ing to predict . Ten years ago we hadthe modeling capabilities and some ofthe computing capabilities, but wedidn't have the radar data ."

In talking about the honors rainingdownon ARPS, Droegemeier uses lan-guage that recalls that of the accep-tance speeches on Oscar night .

"There's one thing I want to be sure

Droegemeier points to a stellarquintet of co-project investigators :

" Fred Carr, professor of meteorol-ogy and director of the school, is anexpert in synoptic meteorology andnumerical weather prediction and oneof the country's leading authorities onoperational meteorology .

" Alan Shapiro, who rose through

Posting the latest research contributions of CAPS scientists is a constant and satisfying activity for Center Director Kelvin

Droegemeier. The core of the CAPS team includes 15 Ph.D.-level veteran investigators as well as 15-20 graduate students .

could deal with both of the significantproblems associated with storm scopeprediction : the theory behind it andthe theory of retrieval as well as thecreation of a model forecast systemthat had the computational power andability to use parallel computers .

"Then, what really pushedthis to apractical reality was the NationalWeather Service's installation of anetwork ofDoppler radar around thecountry, NEXRAD . You can't predictthe future of the atmosphere unlessyou observe it on the scale you're try-

22

SOONERMAGAZINE

doesn't get lost," he says . "ARPS is ateam effort; we're a center . I happentobe director of the center, but a team ofabout 15 or so Ph.D . level scientists,some ofwhomhave been here for sixtoeight years, and 15-20 graduate stu-dents form a core group that does thedevelopment work that leads to thecreation of all the new science . Theyhave a vision . These individuals havedeveloped the software and tested itoperationally . It's a phenomenal groupofpeople . Our success is a reflection ofthe quality of our people ."

the ranks from post-doctoral fellow toresearch scientist to assistant profes-sor of meteorology, is a theoreticianwho has developed many of the tech-niques now used to retrieveunobservedinformation from Doppler radars .

" Ming Xue, a senior research scien-tist, is one of the world's leading nu-merical modelers and leads the modeldevelopment effort at CAPS. He alonewrote amajorportion oftheARPS modeland has been central to its success.

" Jerry Straka, associate professorof meteorology, is a cloud physicist

Page 5: the Envelope Please....of `The envelope, please.' Wedidn't knowuntiljustthatmomentwhohad won." The Smithsonian event was, as might be expected, a bit less glitzy, butglamourwasstill

0

aUa

0U

Environmental forecasting and research at OU received a big boost with the recentacquisition of a newsupercomputer funded by a National Science Foundation grant

and University resources. Formally dedicating the Cray J90 are CAPS Director KelvinDroegemeier, left, Cray account manager Don Crow and OU President David Boren.

APS vs. Reality : June 8, 1995The ARPS forecast (left) created at 1 p.m . for conditions at 7 p.m . compares well with

an actual radar image (right) at 7 p.m . Color indicates rainfall intensity, increasing

from light blue to pink . "The model did a remarkably good job," says KelvinDroegemeier, "getting these storms in just about the right location, especially with most

intensely rotating storms in the northeast Texas panhandle. And it predicted the north-

south extent of the storm line up to Kansas . This represents a tremendous success."

and numerical modeler who has beenconductingresearch into severe stormsfor several years . He helped lead theVORTEX '94 and '95 field programsin Oklahoma, which were among themost successful ever in capturingdetailedinformation ofsevere stormsand tornadoes .

" Keith Brewster, a research sci-entist and doctoral student, is re-sponsible for the acquisition, qualitycontrol and analysis ofobservationaldata . An exceptionally talentedscientist, he is using the ARPS for hisPh.D . to develop better waysofinitializ-ing storm-scale numerical models .

"We also must thank the Univer-sity for its support," Droegemeier in-sists . "When we competed to get thefunding for CAPS, we went up against322 other universities, many of thembig names like Rice, Berkeley, North-western . Our proposal was scientifi-cally competitive to begin with, butthe reviewers were looking for whatthe University would provide . OUcreated three new faculty positionsfor the Center, one each in math-ematics, computer science and me-teorology . The school raised moneyforan endowed chair. I thinkEddie Smith,deanoftheGraduate College, has givenus halfa million dollars over the yearsin matching support forhardware, dol-lar for dollar . OU put up $400,000 sowe could buy the Cray computer, andnow it's putting up $60,000 for an up-grade that will allow us to do dailyforecastruns . The Universityin ahugeway has supported us . That was cru-cial ."

Like the acceptance speeches at thereal Oscars, this list of credits and ku-dos does go on a bit, but Droegemeierwill not be silenced by cut-to-the-com-mercial music. He should not be . Heand his team are not creating specialeffects for disaster movies aboutfloods and tornadoes ; instead theyare working on methods to reducethe damage done by these destruc-tive acts of nature and, by so doing,save lives and property .

The lives they will save are realones, and the rewards they will reapare much more significant than agolden statue .

1997 SUMMER

23