Skip to Content
MIT Climate Grand Challenge

Sounding the alarm, bolstering defenses

An early warning system to help mitigate the impact of climate disasters in the world’s most vulnerable regions.

August 24, 2022
aerial view of flooding in residential area
iStock

Climate Grand Challenges Flagship Project:

  • Climate Resilience Early Warning System (CREWSnet)

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

At the public unveiling of the flagship Climate Grand Challenges project that he’s co-leading, Elfatih Eltahir, SM ’93, ScD ’93, showed a photograph of a small house in coastal Bangladesh, made of rough-hewn branches.

The house, Eltahir explained, had been destroyed by a cyclone and then rebuilt—in the same place, using the same design and the same materials. “That kind of reaction to climate change,” he said, “is not enough.” 

As one of the countries most vulnerable to climate change, Bangladesh is already facing an onslaught of cyclones, droughts, heat waves, rising sea levels, and pervasive flooding—conditions that are expected to intensify as the atmosphere continues to warm. By 2050, it’s predicted, climate change will displace one in seven Bangladeshis. 

Even if people stop pumping greenhouse gases into the atmosphere, we will need to adapt in preparation for the changes already on the way. In rich, developed countries, that adaptation—the climate-proofing of infrastructure, health-care systems, and agriculture—will be a lot easier. But many poor countries, like Bangladesh, will need help. “Despite all these significant risks, the state of the art in how adaptation is done, especially in vulnerable communities around the world, is not satisfactory,” Eltahir said.

The global modeling systems used to project how climate change will play out operate on large scales—too large to provide useful information to communities or even countries. But Eltahir, the H.M. King Bhumibol Professor in Civil and Environ­mental Engineering, has been working for nearly 25 years on a modeling system that better reflects the impacts of coastlines, topographies, and differences in soil and land-use patterns. He believes that system is now ready to be used to inform communities around the world. Armed with projections from its models, social development organizations with experience on the ground can help them build more resilient homes, schools, infrastructure, and food systems. 

Elfatih Eltahir is the H.M. King Bhumibol Professor of Hydrology and Climate in Civil and Environmental Engineering.
John Aldridge is assistant leader of the Humanitarian Assistance and Disaster Relief (HADR) Systems Group at MIT’s Lincoln Lab.
Deborah Campbell leads MIT Lincoln Lab’s Climate Change Initiative and is a senior staff scientist in the lab’s HADR Group.

Eltahir and co-leads John Aldridge, assistant leader of the Humanitarian Assistance and Disaster Relief Systems Group at MIT Lincoln Laboratory, and Deborah Campbell, who leads Lincoln Lab’s Climate Change Initiative, call their project CREWSnet—the Climate Resilience Early Warning System. The name is a nod to another project, the Famine Early Warning System, or FEWS NET, which the US government developed after major famines in Africa in the 1980s.  

“The basic idea behind CREWSnet is to reinvent the process of climate change adaptation,” Eltahir says. “It integrates a lot of expertise and knowledge and tools that we have developed in different corners of the Institute, pulling them together to help those vulnerable communities.”

Like FEWS NET, it will attempt to predict disasters. But rather than looking at weather events and providing information to guide an emergency response, as FEWS NET does, CREWSnet will anticipate climate change events, incorporate socioeconomic impact forecasting, and inform communities’ efforts to build resilience, in everything from structures to irrigation systems, so they can withstand the changes projected over the next three decades. FEWS NET has evolved into a robust network that relies on local and regional weather forecasting systems and emergency responders. The CREWSnet team hopes to foster a similar network.

“FEWS NET was initially created by the US and has been run here and supported here, but it’s transitioned to more local and regional-based leadership. That transition is important,” Aldridge says. “It’s also become a globally accessible tool. The vision is for [CREWSnet] to become a global public good.”

In its first pilot project, CREWSnet will partner with the Bangladesh-based social development organization BRAC. One of the largest nongovernmental organizations in the world, it has decades of experience running humanitarian, climate-
related, and urban development programs in Bangladesh. 

“What we intend to do is not pretend we’re the local experts,” Aldridge says. “We’re bringing in local knowledge.” Those local experts will team up with all the MIT partners involved in the project, including the Center for Global Change Science, the Joint Program on the Science and Policy of Global Change, and the Abdul Latif Jameel Poverty Action Lab (J-PAL).

“We’re trying—I think in a way that’s never been done before—to stitch together the various facets of MIT’s climate science and impact modeling capacities into one continuous thread, and that thread continues past MIT into the hands of an organization like BRAC,” Aldridge says.  

Like other Climate Grand Challenges flagship projects, CREWSnet aims to serve communities, groups, and governments that might not otherwise have world-class models or research at their disposal; the goal is to translate the outputs from the models into easily accessible, usable information.   

“When a climate model is produced or a numerical simulation is run, oftentimes that information will go into an academic paper—into Science or Nature or a specialty climate journal. And that’s very important work. That’s how science advances,” Aldridge says. “But there’s an asymmetry of access.”

To address that, the CREWSnet team is tapping into Lincoln Laboratory’s information management expertise. “We’re good at decision support systems—how you distill information so that it’s useful,” Aldridge says. With CREWSnet, countries, communities, and families will be able to use the information generated by the models to understand local risks and take action to make their livelihoods, agriculture, health care, education, housing, and infrastructure more resilient.

So instead of simply rebuilding the destroyed houses on the coast of Bangladesh, for example, CREWSnet will look at the predictions for climate change in that area over the next few decades and then provide data that decision makers, including city officials or regional planners, can use to build structures that will last.

“We’re saying in the next 30 years, the storms are going to be 50% more destructive,” Eltahir says. On the basis of such a projection, CREWSnet would advise people to build homes using stronger materials or in less vulnerable locations.

CREWSnet can also be used to help farmers understand the risks of flooding and salinity intrusion so they can consider growing crops better suited to the new conditions. And communities in high-risk areas can use its highly localized forecasting to devise their own strategies for relocation.

CREWSnet will attempt to track just how effective any adaptive intervention is, using a methodology developed at J-PAL. (J-PAL cofounders Esther Duflo, PhD ’99, and Abhijit Banerjee shared the 2019 Nobel Prize in economics with Harvard’s Michael Kremer for this approach, which designs field experiments akin to randomized, controlled trials.) That impact data will be fed into an iterative, feedback-based process to monitor and refine resilience programs and innovations. If this is able to demonstrate CREWSnet’s value in Bangladesh, the researchers plan to expand the project to other high-risk places with the help of local public, private, and nonprofit partners. 

“I’ve done the kind of modeling in which we improve the models and we write papers and publish them. This is different,” Eltahir says. “This is a model tailored for a specific region, and then projecting the information and providing the support and evidence that it works. This is very real.”

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.