Skip to Content
MIT Climate Grand Challenge

Predicting—and preparing for—the worst

Climate change is making extreme weather events more frequent and intense. So improving local risk prediction—and studying how extreme weather could affect renewable energy systems—will be critical to building resilience into vulnerable communities and the growing green energy sector.

August 24, 2022
group of trees bending in high tropical winds
Hurricane Irma seen striking Miami, Florida with 100+ mph winds and destructive storm surge.iStock

Climate Grand Challenges Flagship Project:

  • Preparing for a new world of weather and climate extremes

    Part of a special report on MIT’s Climate Grand Challenges Initiative

After a massive earthquake struck near Kobe, Japan, in 1995, Miho Mazereeuw went to the city to visit her parents, who had just moved there.  

“There was so much chaos,” she remembers. “Had there been some thought in the design of the city in anticipation of this kind of event, that could have saved a lot of lives.”

The event set Mazereeuw, who was then an undergraduate at Wesleyan University, on a path to becoming an expert in urban disaster resilience. She now runs the Urban Risk Lab at MIT and is a tenured associate professor of architecture and urbanism. 

The particular disaster that informed Mazereeuw’s professional path was seismic. But over the past 30 years, it’s become increasingly clear that more of the natural disasters confronting the world will be related to climate change—and that cities and countries need to do a much better job bracing for them.

Mazereeuw is one of three co-leads of a Climate Grand Challenges project that aims to help communities, especially those vulnerable to climate-induced extreme weather, build resilience to catastrophes such as devastating hurricanes, heat waves, or floods.

“One of the things we know about climate change is that it’s affecting extreme events like heavy rainfall events, windstorms, and also heat waves,” says co-lead Paul O’Gorman, a professor in the Program in Atmospheres, Oceans, and Climate in the Department of Earth, Atmospheric, and Planetary Sciences (EAPS). Rainfall of up to 60 inches in four days from Hurricane Harvey in 2017 and the unprecedented 2021 heat wave in western North America are recent examples. “We know that the major impacts of climate change are actually from extreme weather, not so much from changes in the average temperature,” he says. “We need to know, for adaptation, what will happen in a specific region or a city, and that’s much more difficult.”

Paul O’Gorman is a professor of atmospheric science in the Program in Atmospheres, Oceans, and Climate in the Department of Earth, Atmospheric, and Planetary Sciences.
Miho Mazereeuw is a tenured associate professor of architecture and urbanism in the Department of Architecture and director of the Urban Risk Lab.
Kerry Emanuel is the Cecil and Ida Green Professor of Atmospheric Science in the Department of Earth, Atmospheric, and Planetary Sciences and co-director of the MIT Lorenz Center.

The three-pronged effort will tackle related problems: uncertainties in current climate modeling projections (and the difficulty of using global models to assess local risks), challenges that under-resourced communities face in adapting to climate change, and questions about whether a new, carbon-free energy system can withstand extreme weather.

“One thing that is exciting about our approach, I think, is that we’re going to be focusing on the science of extreme weather and engineering societal solutions to decarbonization in one integrated project,” said team member Jessika Trancik, a professor at the Institute for Data, Systems, and Society, at the public unveiling of the project in April. “This means that we may identify extreme weather events that we might have otherwise overlooked, where we need new science to understand them, and we’re also going to be able to identify risks to infrastructure that can inform the design of our decarbonized energy system.”

Big, hugely destructive weather events have historically happened just once every 100 or 200 years, and historical records don’t go back far enough to provide meaningful information about their impacts. “One needs to have about 1,000 years of historical record, and we don’t have 1,000 years of record,” says co-lead Kerry Emanuel, the Cecil and Ida Green Professor of Atmospheric Science in EAPS. That’s where modeling comes in. “The purpose of our project is to use advanced numerical modeling of climate and weather to [tackle] this problem of estimating natural hazard risk,” he says. “We think we can get much better estimates of risk through models than historical statistics.” 

But the models present a challenge too, Emanuel explains. Running computer simulations of these rare events takes time and a vast amount of costly computational power. They also don’t adequately capture the risk that a once-in-a-decade or once-in-a-century event will strike a particular location. 

So the team intends to work on improving a process known as “downscaling,” in which climate scientists modify a complex global model with calculations that account for the impact of localized features like water bodies and vegetation. This offers a better sense of the projected risk of an extreme event—the type that is clearly becoming more frequent and more dramatic as a result of climate change.  

“Models are based on equations informed by physics and chemistry, but the issue is that in most cases—not all, but most—they’re too coarse, meaning their grid spacing is too large to be used for, say, the risk of flooding in a given city,” O’Gorman explains. “The way to deal with that is to use statistical relationships from the current climate to relate what’s happening in the climate model at the larger scale at a much finer grid spacing.” Machine learning plays a big role in making that possible. 

The next key step will be getting that understanding into the communities most at risk.  

“Adaptation has been occurring, but it’s relatively slow and works best in communities with existing resources,” O’Gorman says. “There’s also a challenge of communicating and facilitating efforts to adapt in these communities.”

To address that challenge, the researchers intend to build what they’re calling a “climate adaptation and preparedness toolkit” that will help communities draw on the improved information from the models to integrate communication, planning, development, and design. The team is already creating prototypes of these toolkits through projects in Massachusetts, Florida, Puerto Rico, and South Africa. 

Mazereeuw, who heads this part of the project, explains that the kits will consist of educational modules, interactive platforms, and technologically innovative building and infrastructure designs, all customized for a location through input from community workshops.

“Working with emergency officers, city planners, real estate developers, local organizations, and residents, we will create more accessible and equitable ways to prepare for extreme weather events,” she says. “It’s our role to listen.”

The final piece of the project will be to analyze vulnerabilities in the fast-growing renewable energy system, which feeds solar, wind, and hydroelectric power into newly built grids. Critics of renewable energy have long said it’s less reliable than a system based on fossil fuel.

Trancik says otherwise. 

“We know that a carbon-free energy system can be much more resilient than the energy system that we have today,” she said in her presentation. “It will provide pollution-free, highly reliable power to consumers, but in order to achieve it, we will have to be deliberate in our design.”

A guiding principle for the project is the need to acknowledge that climate change disproportionately affects developing countries, even though they have historically emitted far less greenhouse-gas pollution than the wealthier countries and can’t as readily afford to adapt. And many poor communities, even in the US, could be disproportionately affected by climate change and may face similar adaptation challenges.

“We know, through all the research we’ve done on disasters, that the suffering isn’t equally distributed,” Mazereeuw says—natural disasters expose the glaring inequalities between the rich and the poor. “We want to get ahead of that,” she says.

This story is part of a series, MIT Climate Grand Challenges: Hacking climate change.

Keep Reading

Most Popular

Large language models can do jaw-dropping things. But nobody knows exactly why.

And that's a problem. Figuring it out is one of the biggest scientific puzzles of our time and a crucial step towards controlling more powerful future models.

OpenAI teases an amazing new generative video model called Sora

The firm is sharing Sora with a small group of safety testers but the rest of us will have to wait to learn more.

Google’s Gemini is now in everything. Here’s how you can try it out.

Gmail, Docs, and more will now come with Gemini baked in. But Europeans will have to wait before they can download the app.

This baby with a head camera helped teach an AI how kids learn language

A neural network trained on the experiences of a single young child managed to learn one of the core components of language: how to match words to the objects they represent.

Stay connected

Illustration by Rose Wong

Get the latest updates from
MIT Technology Review

Discover special offers, top stories, upcoming events, and more.

Thank you for submitting your email!

Explore more newsletters

It looks like something went wrong.

We’re having trouble saving your preferences. Try refreshing this page and updating them one more time. If you continue to get this message, reach out to us at customer-service@technologyreview.com with a list of newsletters you’d like to receive.