This year has been record breaking for extreme hurricane activity. First came聽Harvey, which dropped a record-breaking 50-plus inches of rain across parts of southeast Texas, leading to widespread flooding. Less than two weeks later, Irma聽pummeled parts of the Caribbean before聽slamming into聽Florida. And most recently, Maria destroyed livelihoods across Puerto Rico, wiping out the island鈥檚 power infrastructure.
September 2017 now holds the record for most active month of聽any Atlantic hurricane season. Of the 13 named storms so far this season, eight have been hurricanes, with five of the eight 鈥� Harvey,聽Irma,聽Jose, Lee and Maria 鈥� reaching Category 3 or higher.
Supporting the need for increased understanding of natural disasters through improved modeling and forecasting, the National Science Foundation awarded a team of University of 91吃瓜 engineers nearly $1 million to advance accuracy in forecasting storm surge.
Storm surge 鈥� how high ocean waters rise and where flooding occurs 鈥� 聽is often the greatest threat to life and property during a tropical cyclone. A single storm can devastate livelihoods and cause tens of billions of dollars in damage.
As a hurricane approaches land, forecasters currently determine storm surge using complex computer models that account for uncertainties in storm size, track and intensity. Existing models run for real-time forecasts use simplified representations of coastlines and oceans, which are faster for computers to process, yet less accurate than models that use more detailed geographic data.
鈥淐urrently, models are able to predict storm surge reasonably well; however the problem remains that these models can take anywhere from days to weeks to run on a computer,鈥� said study co-investigator , Frank M. Freimann Collegiate Chair in Hydrology and associate professor of the at 91吃瓜. 鈥淰ice versa, there are current approaches that run in a reasonable time frame but aren鈥檛 accurate enough, thus compromising stakeholder鈥檚 abilities to make effective decisions.鈥�
During the four-year study, researchers will work to develop improved storm surge models that incorporate fine-scale data to increase the accuracy of forecasts, while also maintaining reduced computer time and reasonable computational costs.
鈥淏y the end of this study, we hope to bridge the speed-accuracy tradeoff聽 that now exists in surge prediction, said , co-investigator of the study and associate professor of the Department of Civil and Environmental Engineering and Earth Sciences at 91吃瓜. 鈥淭hese results will enable more accurate simulations of surge in real time to assist policymakers, emergency management personnel and coastal residents.鈥�
This award is one of 15 new grants from聽NSF's聽PREEVENTS (Prediction of and Resilience Against Extreme Events) program, which funded $18.7 million in awards this year. Research supported by PREEVENTS aims to improve predictability and risk assessments of natural hazards, increase resilience to these events, and reduce their effects on human lives, societies and economies.
The study, "Collaborative Research: Subgrid-Scale Corrections to Increase the Accuracy and Efficiency of Storm Surge Models" will collaborate with聽North Carolina State University聽and is affiliated with the聽. Research Assistant Professor聽, from the Department of Civil and Environmental Engineering and Earth Sciences at 91吃瓜,聽is also a co-investigator of the work.