A new app details where your food comes from — and just how fragile the global food system really is

After founding the Better Planet Laboratory at the University of Colorado Boulder in 2021, Zia Mehrabi, one of a handful of scientists studying the intersection of food insecurity and climate change, soon found himself fielding a steady stream of calls from policymakers and peers. Everyone wanted more quantitative insight into how extreme weather events affect food supply chains and contribute to hunger around the world. But Mehrabi found the economic puzzle difficult to solve due to the limited public information available. What he could readily find mostly analyzed each disruption in isolation, focusing on one specific part of the world. It failed to account for the expansive flow of goods in global markets or the compounding effects of climate change on the supply chain — and it had to be laboriously mined from reports and one-off case studies. 

So when the nonprofit Earth Genome, which builds data-driven tools and resources for a more sustainable planet, approached Mehrabi to collaborate on developing his vision for a digital food supply map, he leapt at the chance. When their U.S. prototype proved successful, they went global.

The resulting app, which launched Thursday and was shared exclusively with Grist, identifies food flows through just about every major port, road, rail, and shipping lane across the world and traces goods to where they are ultimately consumed. The developers have crowned it a “digital twin of the global food system” and hope it will be used by policymakers and researchers working to better adapt to an increasingly fragile supply chain beleaguered by climate change. The model pinpoints critical global transportation chokepoints where disruptions, such as extreme weather, would have domino effects on food security and, in doing so, identifies opportunities for local and regional agricultural producers to gain a forward-thinking market foothold.

“Food is so important to us,” said Mehrabi. “There’s a need for building these systems, these digital food twins that can be used in decision-making contexts. The first step to doing that is building the data.”

The model is a “first of its kind,” according to Alla Semenova, an economist at St. Mary’s College of Maryland who was not involved with the development of the project. The tool makes the interconnected nature of the global food supply system clear and “underlines the importance of government policies aimed at supporting diversified and localized food production and distribution systems,” she said.

Food systems don’t operate independently. From seeds sprouting to life in fallow fields to the very moment a shopper buys a packaged good from a local vendor, the supply chain links producers, consumers, laborers, processors, regulators, analysts, drivers, and retailers together in a complex web. It’s a network that stretches beyond borders and bodies of water, connecting people and places across the globe. That complexity also makes our understanding of the ripple effect of climate disruptions across the planet’s food system inherently fragmented.

The map attempts to make sense of the tangled maze of food supply chains across the world. It provides a detailed view of the amount of the most common agricultural food groups — from grains and oils to dairy, eggs, and meat — exported outside states, districts, and municipalities. Other elements embedded into its data repository measure the total economic impact of the supply chain on people and food accessibility in a region, tallying the size of its agricultural sector, the average annual economic output per person, population size, and measures of human health, standard of living, and education. The tool also calculates the total mass, calories, and macronutrient content of all crop, aquatic, and livestock commodities flowing in and out of a place. It illustrates trade data, too, for nearly 3,800 regions across 240 countries. 

The model also visualizes critical choke points where disruptions, such as extreme weather, would have cascading effects on these commodity flows. In the data, the Suez Canal, the Panama Canal, the Turkish Straits, the Strait of Malacca, the Black Sea, and a relatively small number of ports, inland waterways and railway networks in the U.S. and Brazil all stand out as bottlenecks — key maritime passages and coastal and island choke points handling considerable portions of the world’s food trade.

It can even be proactively used to assess how a corresponding series of climate shocks on a trade route is measured in calories, protein, or critical micronutrients — all prime food insecurity benchmarks, said Mehrabi. Roughly 9 percent of the world’s supply chain routes — fewer than 350 — account for 80 percent of global caloric flows.

The U.S. is not insulated from the effects of extreme weather shocks on the food system. It imports about 128 megatonnes of food from roughly 154 countries around the world, which represents about a third of the nation’s food supply, according to an analysis by Better Planet Laboratory data scientist Ginni Braich. Some of its top imports, including bananas, coffee, olive oil, cocoa beans, and oranges, face the most imminent climate-related risks.

Similarly, if a series of simultaneous and destabilizing climate shocks hit one of the leading wheat exporters in Western Australia, India’s rice powerhouse in Uttar Pradesh, and Paraná, Brazil, which is among the planet’s biggest exporters of soybeans, it could disrupt food supplies and affect food energy requirements for tens of millions of people. These regions have already experienced severe extreme weather in recent years. In 2023, parts of the state of Western Australia confronted the lowest annual rainfall on record since 1900, above-average temperatures, and everything from severe heat waves to catastrophic fire danger conditions and significant blazes. Uttar Pradesh experienced extreme weather events on 167 days of 2024 — up from 119 days in the year before — while periods of heavy rainfall flooded swaths of Paraná and droughts dried up rivers throughout Brazil.

According to the open-source data the team released with the map, severe disruptions to food exports from these three regions could affect the calories that support more than a million people in the U.S. and Mexico and 55 million people in China for a year. The cascading effects would be most acutely felt by low-income households in these locations that are already struggling with food access.

Given the implications for food security, Mehrabi’s team has heard from several groups interested in understanding how the tool might be used to help governments prepare emergency food reserves. The initial U.S. prototype garnered interest from officials at the State Department and Department of Homeland Security during the Biden administration, including former Special Envoy for Global Food Security Cary Fowler.

Fowler told Grist that when he was at the State Department, his office had “a number of interactions” with the team behind the map while they were developing it. “I thought then and think now that this approach holds much promise in helping us understand and analyze large amounts of data and complex relationships,” said Fowler. “As these tools are improved, I can imagine that they will catalyze new insights and help with program and policy development. They could potentially provide us with an ‘early warning’ of where food system problems are set to erupt into crisis.”

Despite its clear benefits, the map does have some limitations. It doesn’t display what specific agricultural goods a place may import or where residents’ food comes from. (Though the developers say that can be mined from the data.) The map shows where the food is flowing based on estimates of the cheapest route to transport the food and satellite data on known routes — and not, say, the precise numbers of trucks or rail cars, or port capacities. And unlike its U.S.-geared predecessor, the tool does not have an embedded model of what different climate shocks and extreme weather events might do to food availability in an area.

“We’re not directly competing with a very specific use case for something like ‘How much do you stock a warehouse?’” said Mehrabi of the model’s limitations. “That’s not what we’re trying to do … our aim is from a humanitarian perspective.”

And that shows up in how the tool visualizes the brittleness of the current food system, according to Earth Genome’s creative technologist, Cameron Kruse. While their initial U.S. model showed that just 5.5 percent of the nation’s total counties produce half the country’s food, this global picture is even more concentrated, he said. Just 1.2 percent of the world’s countries are responsible for half of all domestic wheat exports, exposing critical vulnerabilities in the global food supply. It also sounds the alarm about the global effects of localized transport disruptions and provides a framework for future simulations that could predict the effects of climate shocks.

“As long as these models stay siloed and isolated, they continue growing siloed and isolated,” said Kruse. “If you hear about a drought in the news, or you hear about certain hurricanes impacting a region, go to that region in Food Twin, and see where that region is producing food. And then check out the news and see if global leaders are talking about it,” he said. “Use this as almost a gut-check of like, ‘Are we focusing on the right issues?’”

Editor’s note: Cary Fowler is a former Grist donor. Funders have no role in Grist’s editorial decisions.