New tool predicts flood risk from hurricanes in a warming climate
Using New York as a test case, the model predicts flooding at the level experienced during Hurricane Sandy will occur roughly every 30 years by the end of this century.
Coastal cities and communities will face more frequent major hurricanes with climate change in the coming years. To help prepare coastal cities against future storms, MIT scientists have developed a method to predict how much flooding a coastal community is likely to experience as hurricanes evolve over the next decades.
When hurricanes make landfall, strong winds whip up salty ocean waters that generate storm surge in coastal regions. As the storms move over land, torrential rainfall can induce further flooding inland. When multiple flood sources such as storm surge and rainfall interact, they can compound a hurricane’s hazards, leading to significantly more flooding than would result from any one source alone. The new study introduces a physics-based method for predicting how the risk of such complex, compound flooding may evolve under a warming climate in coastal cities.
One example of compound flooding’s impact is the aftermath from Hurricane Sandy in 2012. The storm made landfall on the East Coast of the United States as heavy winds whipped up a towering storm surge that combined with rainfall-driven flooding in some areas to cause historic and devastating floods across New York and New Jersey.
In their study, the MIT team applied the new compound flood-modeling method to New York City to predict how climate change may influence the risk of compound flooding from Sandy-like hurricanes over the next decades.
They found that, in today’s climate, a Sandy-level compound flooding event will likely hit New York City every 150 years. By midcentury, a warmer climate will drive up the frequency of such flooding, to every 60 years. At the end of the century, destructive Sandy-like floods will deluge the city every 30 years — a fivefold increase compared to the present climate.
“Long-term average damages from weather hazards are usually dominated by the rare, intense events like Hurricane Sandy,” says study co-author Kerry Emanuel, professor emeritus of atmospheric science at MIT. “It is important to get these right.”
While these are sobering projections, the researchers hope the flood forecasts can help city planners prepare and protect against future disasters. “Our methodology equips coastal city authorities and policymakers with essential tools to conduct compound flooding risk assessments from hurricanes in coastal cities at a detailed, granular level, extending to each street or building, in both current and future decades,” says study author Ali Sarhadi, a postdoc in MIT’s Department of Earth, Atmospheric and Planetary Sciences.
The team’s open-access study appears online today in the Bulletin of the American Meteorological Society. Co-authors include Raphaël Rousseau-Rizzi at MIT’s Lorenz Center, Kyle Mandli at Columbia University, Jeffrey Neal at the University of Bristol, Michael Wiper at the Charles III University of Madrid, and Monika Feldmann at the Swiss Federal Institute of Technology Lausanne.
The seeds of floods
To forecast a region’s flood risk, weather modelers typically look to the past. Historical records contain measurements of previous hurricanes’ wind speeds, rainfall, and spatial extent, which scientists use to predict where and how much flooding may occur with coming storms. But Sarhadi believes that the limitations and brevity of these historical records are insufficient for predicting future hurricanes’ risks.
“Even if we had lengthy historical records, they wouldn’t be a good guide for future risks because of climate change,” he says. “Climate change is changing the structural characteristics, frequency, intensity, and movement of hurricanes, and we cannot rely on the past.”
Sarhadi and his colleagues instead looked to predict a region’s risk of hurricane flooding in a changing climate using a physics-based risk assessment methodology. They first paired simulations of hurricane activity with coupled ocean and atmospheric models over time. With the hurricane simulations, developed originally by Emanuel, the researchers virtually scatter tens of thousands of “seeds” of hurricanes into a simulated climate. Most seeds dissipate, while a few grow into category-level storms, depending on the conditions of the ocean and atmosphere.
When the team drives these hurricane simulations with climate models of ocean and atmospheric conditions under certain global temperature projections, they can see how hurricanes change, for instance in terms of intensity, frequency, and size, under past, current, and future climate conditions.
The team then sought to precisely predict the level and degree of compound flooding from future hurricanes in coastal cities. The researchers first used rainfall models to simulate rain intensity for a large number of simulated hurricanes, then applied numerical models to hydraulically translate that rainfall intensity into flooding on the ground during landfalling of hurricanes, given information about a region such as its surface and topography characteristics. They also simulated the same hurricanes’ storm surges, using hydrodynamic models to translate hurricanes’ maximum wind speed and sea level pressure into surge height in coastal areas. The simulation further assessed the propagation of ocean waters into coastal areas, causing coastal flooding.
Then, the team developed a numerical hydrodynamic model to predict how two sources of hurricane-induced flooding, such as storm surge and rain-driven flooding, would simultaneously interact through time and space, as simulated hurricanes make landfall in coastal regions such as New York City, in both current and future climates.
“There’s a complex, nonlinear hydrodynamic interaction between saltwater surge-driven flooding and freshwater rainfall-driven flooding, that forms compound flooding that a lot of existing methods ignore,” Sarhadi says. “As a result, they underestimate the risk of compound flooding.”
Amplified risk
With their flood-forecasting method in place, the team applied it to a specific test case: New York City. They used the multipronged method to predict the city’s risk of compound flooding from hurricanes, and more specifically from Sandy-like hurricanes, in present and future climates. Their simulations showed that the city’s odds of experiencing Sandy-like flooding will increase significantly over the next decades as the climate warms, from once every 150 years in the current climate, to every 60 years by 2050, and every 30 years by 2099.
Interestingly, they found that much of this increase in risk has less to do with how hurricanes themselves will change with warming climates, but with how sea levels will increase around the world.
“In future decades, we will experience sea level rise in coastal areas, and we also incorporated that effect into our models to see how much that would increase the risk of compound flooding,” Sarhadi explains. “And in fact, we see sea level rise is playing a major role in amplifying the risk of compound flooding from hurricanes in New York City.”
The team’s methodology can be applied to any coastal city to assess the risk of compound flooding from hurricanes and extratropical storms. With this approach, Sarhadi hopes decision-makers can make informed decisions regarding the implementation of adaptive measures, such as reinforcing coastal defenses to enhance infrastructure and community resilience.
“Another aspect highlighting the urgency of our research is the projected 25 percent increase in coastal populations by midcentury, leading to heightened exposure to damaging storms,” Sarhadi says. “Additionally, we have trillions of dollars in assets situated in coastal flood-prone areas, necessitating proactive strategies to reduce damages from compound flooding from hurricanes under a warming climate.”
This research was supported, in part, by Homesite Insurance.
###
Written by Jennifer Chu, MIT News
Paper: “Climate change contributions to increasing compound flooding risk in New York City”
https://journals.ametsoc.org/doi/10.1175/BAMS-D-23-0177.1
JOURNAL
Bulletin of the American Meteorological Society
ARTICLE TITLE
“Climate change contributions to increasing compound flooding risk in New York City”
Paper provides a clearer picture of severe hydro hazards
Understanding the likelihood of a major rain event, or pluvial, following a drought and vice versa will aid in planning and preparation for ‘climate hazard flip.’
IMAGE:
LINYIN CHENG (LEFT) AND YICHAN LI
view moreCREDIT: UNIVERSITY RELATIONS
Over the last two decades an estimated three billion people have been affected by water-related natural disasters such as droughts and floods. Climate change is expected to increase the frequency of these hydro hazards, with some prognosticators estimating there will be upwards of $3.7 trillion in water-related damage over the next 30 years in the U.S. alone. Beyond damaging homes and infrastructure, severe wet and dry spells will also devastate crops and deplete water reservoirs.
An increasing area of interest to researchers is the frequency of compound drought and pluvial flooding (caused by quick, heavy rainfall or sustained rainfall beyond the norm), which is when both occur in succession in the same area within a year of each other. Historically, this level of coincidence has been under-examined.
Of similar interest is when the reverse happens: extreme rainfall followed by a meteorological drought. Meteorological drought is when dry weather patterns prevail, which can eventually trigger hydrological drought, leading to dry streams and plunging reservoir levels, such as what happened at Lake Mead in 2022.
A new study co-authored by researchers in the University of Arkansas Department of Geosciences, as well as colleagues in China, now provides a global examination of drought-pluvial volatility — or the tendency to shift from one extreme to another (from dry to wet or wet to dry) in a short period of time.
Yichan Li, a Ph.D. candidate at the U of A, was the first of four authors on the paper, “Observational Uncertainty for Global Drought-Pluvial Volatility,” published in Water Resources Research, while Linyin Cheng, an assistant professor of geosciences, was second author.
The study looks at extreme dry-to-wet and wet-to-dry transitions over the past seven decades through event coincidence analysis, a method of quantifying the number of consecutive extreme events that also considers instantaneous or lagged responses within an uncertain period between them. The study used three widely used climate data sets to provide evidence of increased drought-pluvial volatility on time scales of less than a year.
The team also evaluated the accuracy of these data sets, finding varying strengths and weaknesses of each due to observational uncertainties in data collection. For instance, the remoteness of a region may play a role in collecting accurate data.
Averaged out at the global scale, the team found that 15.46% of all meteorological droughts were succeeded by a pluvial the following season. The wet-to-dry transition percentage proved remarkably similar: 15.49%. However, prominent differences exist when looking at particular regions.
Toward that end, the study provides a map demonstrating how incidents of these two phenomena are distributed globally. Overall, the spatial pattern of extreme dry-to-wet and wet-to-dry events’ coincidence rates is largely in agreement among the three data sets, though there is prominent regional variability.
For instance, in Eurasia since the mid-20th century, there is a relatively low probability for meteorological droughts transitioning to pluvials, but a higher chance for the opposite scenario, rapid shifts from wet to dry events. A similar pattern also exists over western North America, which sees severe wet to dry transitions at a frequency greater than 17% on average. Conversely, South Asia and Australia are more prone to immediate transitions from meteorological droughts to pluvials.
The authors noted: “Our findings indicate that differences associated with drought-pluvial volatility among the considered observations are in many regions larger than that of their single events [droughts or pluvials alone], highlighting a need of to use multiple independent observation-based data sets for more robust examinations when studying such compound extreme events.”
Ultimately, the authors stress the need to use multiple independent observation-based data sets when analyzing extreme, compound dry-to-wet events. This will provide clearer guidelines for climate-related decision making, especially water resources planning, as well as ensure better accuracy when modeling future weather events.
Additional co-authors in the paper included Chiyun Miao, with Beijing Normal University and Zhiyong Liu, with Sun Yat-sen University.
JOURNAL
Water Resources Research
METHOD OF RESEARCH
Computational simulation/modeling
SUBJECT OF RESEARCH
Not applicable
ARTICLE TITLE
Observational Uncertainty for Global Drought-Pluvial Volatility
