Science1 publisher3 min readPublished
Leiden's 12,228-city food database puts Amsterdam and Dhaka on opposite policy tracks
A Leiden team modelled food-consumption emissions for 12,228 cities and found the EAT-Lancet diet cutting Amsterdam's by roughly half while raising Dhaka's. Diet and supply-chain efficiency each explain 48% of the gap between cities.
The Scientist · Science desk
What happened
- Researchers at Leiden University's Institute of Environmental Sciences built a database of greenhouse gas emissions from food consumption covering 12,228 cities in 159 countries.
- Dietary patterns explain about 48% of the differences between the cities studied, supply-chain efficiency a further 48%, and demographics such as age structure the remaining 4%.
- The world's 200 largest cities together account for almost half of all urban food-related emissions in the dataset.
Compiled by The ScientistSomething wrong?How this is made
Why it matters
- decision A city government copying Amsterdam's meat-reduction program into Dhaka would be buying a measure this model says pushes emissions up there, so the question facing each city is which lever to pull.
- constraint Half the between-city variation sits in production, processing and transport, and in a production-driven city like Jakarta that puts the highest-leverage change upstream of the shops a mayor can regulate.
- exposure Cairo lands in the study's limited-data class, so the cities with the thinnest statistics get a diagnosis of missing information while their peers get a ranked list of candidate measures.
- capability Through the MATILDA-City platform a city can now see which foods and which consumer groups drive its own footprint, and whether residents are eating enough fruit, vegetables and pulses.
The 48/48/4 split does less work than it first appears. It describes variation between cities: dietary patterns account for about 48% of why one city's food emissions differ from another's, supply-chain efficiency another 48%, and demographic factors such as age structure the remaining 4% [8]. Diet and supply chain together carry 96% of the spread [2]. That 96% is a statement about spread. What a particular city can abate comes from a scenario, and the scenario here is the EAT-Lancet pattern, which the researchers describe as an internationally recognized diet combining health and sustainability [14].
Amsterdam is where that scenario pays out. "In the Netherlands, food supply chains are already relatively efficient," said first author Hongyi Cai [5]. With the upstream slack largely gone, the remaining reduction sits in what people buy: "In Amsterdam, switching to the EAT-Lancet diet could reduce emissions by around 45% to 50%," Cai said [4].
In Dhaka and Delhi the same substitution runs the other way, because average diets there already contain fewer animal-based products [6]. Co-author Shruti Jain said the goal in those cities is different: "The main challenge is ensuring that people have access to a sufficient variety of nutritious foods, such as pulses, nuts and vegetables, while also making the supply chains that produce them more sustainable" [7].
Lima shows the same mechanism with the diet held roughly in the study's favour. People in Lima eat a more plant-based diet than Parisians, and the phys.org report puts food-related emissions in Peru at three times those in the French capital [10]. As published, that comparison sets a country against a city. It fixes the direction of the gap and leaves its size open. Cai's account of the cause is upstream of the plate. "It comes down to how food is produced, processed and transported," he said [11].
Concentration is the other figure to handle carefully. The world's 200 largest cities together account for almost half of all urban food-related emissions [3], and 200 cities is 1.6% of the 12,228 in the database [1]. The source reports no population or per-capita emissions for those 200, so the half-share shows where the tonnes sit. Whether the average resident there eats more carbon-intensive food than anyone else is past what it shows.
"Many cities have ambitious climate targets, but lack the data needed to determine which measures will have the greatest impact," said co-author Oliver Taherzadeh [13]. The database closes part of that gap. It does not price any of the interventions, it does not test whether a city's retail food supply can deliver the modelled substitution, and its figures are modelled consumption, not measured emissions after a policy [1].
What to watch
- Whether the Environmental Research Letters paper uses city-specific supply-chain factors or national ones downscaled to cities, which sets how firm the four-type sorting is.
- Whether the limited-data class shrinks as cities such as Cairo publish household consumption surveys.
- Whether any city reports measured emissions after a diet policy that can be checked against the modelled 45% to 50% Amsterdam figure.