Tropical cyclone risk research: Taking the world by STORM

Nadia Bloe­men­daal tells the sto­ry behind her RDNL Dutch Data Prize win­ning ‘STORM’ dataset that has been down­loaded more than 3,600 times since its pub­li­ca­tion in 4TU.ResearchData.

The roman­tic sto­ry of two storm chasers who devel­op a tor­na­do data-gath­er­ing device to pro­tect against one of the world’s most vio­lent tor­na­does trig­gered Nadia’s fas­ci­na­tion. 

“Ever since watch­ing the film, I want­ed to become a mete­o­rol­o­gist and study extreme weath­er. The film inspired me to cre­ate some kind of weath­er alert sys­tem that could help to pre­vent cat­a­stroph­ic storm dam­age, pro­tect peo­ple and save lives,” says Nadia.

To ful­fil her life­long ambi­tion, Nadia stud­ied a Bachelor’s degree in Applied Math­e­mat­ics at the Uni­ver­si­ty of Twente and a Master’s degree in Mete­o­rol­o­gy at Utrecht Uni­ver­si­ty before join­ing the VU’s Water and Cli­mate Risk group in 2016 to begin her PhD. 

Trop­i­cal cyclone risk  

Every year trop­i­cal cyclones (hur­ri­canes and typhoons) bring great dev­as­ta­tion to parts of our world, impact­ing peo­ple, economies and the coastal areas where they make land­fall. These dead­ly nat­ur­al dis­as­ters con­tin­ue to pose a major risk to soci­eties world­wide, and cal­cu­lat­ing their risk presents a sig­nif­i­cant chal­lenge due to the lack of his­tor­i­cal data.

“There’s only around 40 years of reli­able, glob­al-scale satel­lite data avail­able to study trop­i­cal cyclones,” explains Nadia. “What’s more, there are only around 90 cyclone for­ma­tions on aver­age each year. Of these, only a small pro­por­tion make land­fall and cause dam­age. So, there aren’t actu­al­ly that many trop­i­cal cyclones to study.”

She con­tin­ues, “Trop­i­cal cyclone dam­age typ­i­cal­ly occurs with­in 500 km of land­fall loca­tion which is a very small region rel­a­tive to the size of a country’s entire coast­line. This makes it even more chal­leng­ing to study risk as there’s an enor­mous num­ber of loca­tions where a trop­i­cal cyclone could hit.”

Mak­ing waves with ‘STORM’

In order to over­come the prob­lem of lim­it­ed data for trop­i­cal cyclone risk research, Nadia and her team of col­lab­o­ra­tors from IVM, Deltares, KNMI and the Uni­ver­si­ty of Southamp­ton, devel­oped a math­e­mat­i­cal mod­el that can sta­tis­ti­cal­ly extend the avail­able data to sim­u­late 10,000 years of trop­i­cal cyclone activ­i­ty. 

The ‘Synthet­ic Trop­i­cal cyclOne geneRation Model’, oth­er­wise known as ‘STORM’, can extract and ampli­fy data about trop­i­cal cyclones from any mete­o­ro­log­i­cal dataset to pro­vide infor­ma­tion about their pos­si­ble tracks and inten­si­ties. STORM sim­u­la­tions pro­vide the num­ber, geo­graph­ic loca­tion and wind speed of trop­i­cal cyclones. 

Nadia explains that cer­tain envi­ron­men­tal vari­ables must be con­sid­ered in order to ensure sim­u­la­tion accu­ra­cy. “STORM is a Python mod­el that inte­grates NetCDF mete­o­ro­log­i­cal data with the pub­licly avail­able ERA5 dataset (ECMWF) to adjust for sea sur­face tem­per­a­ture and atmos­pher­ic pres­sure. This allows STORM to accu­rate­ly and reli­ably sim­u­late trop­i­cal cyclone activ­i­ty in sus­cep­ti­ble coastal regions based on present-cli­mate con­di­tions.

The win­ning dataset 

Nadia applied STORM to extract 38 years of his­tor­i­cal trop­i­cal cyclone data from the Best Track Archive for Cli­mate Stew­ard­ship (IBTrACS). Since IBTrACS is the most com­pre­hen­sive dataset from the Nation­al Ocean­ic and Atmos­pher­ic Admin­is­tra­tion, Nadia gen­er­at­ed the first glob­al syn­thet­ic trop­i­cal cyclone risk dataset using STORM. 

Her resul­tant dataset, ‘STORM IBTrACS present cli­mate syn­thet­ic trop­i­cal cyclone tracks’, caused a big splash! 

It won last year’s RDNL Dutch Data Prize and has been down­loaded 3,670 times since it was pub­lished in 4TU.ResearchData only four months ago. The asso­ci­at­ed peer-reviewed pub­li­ca­tion is avail­able in Sci­en­tif­ic Data. 

Real world appli­ca­tions 

“STORM has high val­ue in pro­tect­ing people’s lives. I made my datasets pub­licly avail­able in the 4TU.ResearchData repos­i­to­ry so that they can ben­e­fit oth­ers in the field of cli­mate research,” says Nadia. 

A vari­ety of stake­hold­ers have shown inter­est in Nadia’s STORM data, from researchers study­ing trop­i­cal cyclone risk in Ton­ga in the South Pacif­ic to inter­na­tion­al insur­ance com­pa­ny, Allianz. 

“Insur­ance com­pa­nies can use STORM to mod­el trop­i­cal cyclone risk at high­er res­o­lu­tion and over longer-time scales, allow­ing them to per­form large-scale risk assess­ments with greater con­fi­dence than ever before.”

Nadia came third in the Allianz Cli­mate Risk Research Award last year. She was award­ed this prize for the best cli­mate risk relat­ed PhD research. 

Apart from improv­ing risk mit­i­ga­tion through insur­ances, Nadia hopes that her STORM datasets can help con­struc­tion com­pa­nies to build hous­es that can with­stand cer­tain wind speeds in trop­i­cal cyclone-prone coastal regions.This way, peo­ple will have a bet­ter chance at pro­tect­ing them­selves from a trop­i­cal cyclone, here­by reduc­ing impacts and ulti­mate­ly sav­ing more lives

Next steps 

As Nadia is set to com­plete her PhD in May, her final project is to apply STORM to the PRIMAVERA dataset to iden­ti­fy coastal regions that will be at increased risk of trop­i­cal cyclones under cli­mate change. Local pol­i­cy­mak­ers and oth­er stake­hold­ers can take appro­pri­ate mea­sures to reduce poten­tial future impacts.

And, with her RDNL Dutch Data Prize mon­ey, Nadia plans to design an inter­ac­tive online world map of pos­si­ble trop­i­cal cyclones based on her STORM datasets. Users will be able to search, view and select trop­i­cal cyclone tracks to learn about their risk, inten­si­ty and shift under cli­mate change. 

Nadia is most excit­ed about being able to pro­vide trop­i­cal cyclone infor­ma­tion at any loca­tion. “The web­site will allow users to nav­i­gate any place on earth, let’s say down­town Mia­mi, for exam­ple, and see its full trop­i­cal cyclone risk pro­file under present or cli­mate change con­di­tions!”

Nadia hopes that a sim­ple online tool will help the gen­er­al pub­lic to visu­alise com­plex mete­o­ro­log­i­cal data, under­stand trop­i­cal cyclone risk and act accord­ing­ly. 

The playlist behind STORM

To be ful­ly trans­par­ent about her research, Nadia wants to share the top four songs that she lis­tened to whilst devel­op­ing STORM:

1. Scor­pi­ons — Rock you like a hur­ri­cane 
2. The Doors — Rid­ers on the STORM
3. Bløf — Hard­er dan ik hebben kan 
4. Rammestein — Sonne 

Thank you, Nadia, for shar­ing the sto­ry behind ‘STORM’ with 4TU.ResearchData. We wish you all the best for your future career.

Acknowl­edge­ments to Nadia’s STORM col­lab­o­ra­tors; Hans de Moel and Jeroen Aerts (IVM), Sanne Muis (Deltares and IVM), Rein­dert Haars­ma (KNMI) and Ivan Haigh (Uni­ver­si­ty of Southamp­ton and IVM)

Writ­ten by Con­nie Clare (4TU.ResearchData)
Cov­er image illus­trat­ed by Con­nie Clare (4TU.ResearchData)