How the World tab works

Growing food takes land and releases greenhouse gases. The World tab asks two questions about your diet: does it need more or less land than the average person in the Netherlands, and does it release more or less greenhouse gas? It then applies both answers to the whole world. The first answer redraws the world map, and the second projects a hundred years of emissions. This page explains step by step how that works, so you know what you are looking at.

The result is a comparison. The tab does not measure how much land you use yourself, and the map is not a forecast. The rest of this page explains why it is set up this way.

The comparison everything rests on

your land use per day ÷ 3.09 m²·year per day

Your figure comes from your own log: the average land use of the days you filled in, calculated with the same figures the app shows everywhere else.

3.09 m²·year per day is what the average person in the Netherlands needs. The unit looks unusual but is simple: the area of land used for your food, multiplied by the time that land is in use. Over a whole year this comes to about 1,100 m² of farmland per person.

Where 3.09 comes from

RIVM, the Dutch National Institute for Public Health and the Environment, recorded what 3,570 people in the Netherlands actually eat. The group was weighted to represent the population by age, sex, region, education, season, and weekday or weekend. RIVM combined this with its own measurements of how much land each food needs.

Those are the same measurements this app uses for your food. That is why this figure was chosen. Both sides of the comparison were made in the same way, so dividing one by the other makes sense. A land figure from another study would have been measured differently. The comparison would still look convincing, but it would mean nothing.

Meat alone accounts for more than a quarter. Women use 2.74 on average and men 3.45, a difference of 26%.

Because of that difference, the World tab lets you choose who you are compared with. A single average for the whole population makes one half look better and the other half worse. That is why you can choose between Everyone, Women and Men below the headline figure. Your choice changes the land comparison and the greenhouse gas comparison together, because both come from the same study and the same set of tables. Comparing your land use with one group and your emissions with another would mean making two claims about two different groups of people.

Why a comparison, and not a number of hectares

Because the measurements describe the Dutch food system. In the Netherlands most cattle eat feed grown on farmland. In much of the world, cattle graze on open land instead. The same steak really does take a different amount of land depending on where the animal was raised, and only the Dutch situation has been measured here.

An absolute number would apply Dutch farming to wherever you live, without you noticing. What does hold up across that difference is the comparison: more land than average, or less. That is all this tab claims. The more your food culture differs from the Dutch one, the more carefully you should read even that.

The map starts from the world as it is

The first column of the table is the world today, so nothing changes there. Every other column is compared with it. This is good to know, because the calculation hides a real mismatch. The Dutch figure comes to about 1,100 m² of land per person, while the world actually has about 5,900 m² of farmland per person.

That gap is real. Farmland also grows cotton, tobacco and crops for fuel. Much of the world's grazing land is dry, poor land that feeds very few animals per hectare. And different studies count land in different ways. The World tab does not try to make these fit. It takes today's world as the reference point for average and shows how the picture would move away from it.

The four kinds of land

The world's land is divided into four kinds, on a grid of squares of about 9 km. All figures are in ×10⁶ km², which means millions of square kilometres.

• Farmland: 47.97. Crop fields and grazing land together. Grazing land is included because it explains much of why meat needs so much land, and the figures for food already count it.
• Wild land: 66.18. Forest, grassland, scrub, wetland and tundra. This figure is what remains after subtracting the other three from the world total, so it is not counted directly.
• Built-up land: 0.86. Smaller than most people expect. Further down this page you can read why.
• Desert and ice: 16.79. Bare sand, rock and gravel, plus glaciers and ice sheets.

What changes, and what does not

When your diet changes, only two of the four move: farmland and wild land swap places. A growing population adds built-up land, and that also comes out of wild land. Desert and ice never change, not with what people eat and not with how many people there are. They have the same figure in every column, because no way of eating turns the Sahara into a field.

Why grazing land counts as farmland

Of the 47.97, about 32 is grazing land rather than crop fields, and much of it is open rangeland that has never been ploughed. You might almost call that wild.

It still counts as farmland, because the food figures already include it: the land figure for a steak includes the pasture the animal grazed on. Keeping it in the food figures but leaving it out of the world total would make a diet look as if it needs less land than it does. The fair caveat is about ecology. Giving rangeland back to nature restores less than giving back a crop field, because rangeland was never as far removed from nature in the first place.

Why built-up land looks so small

0.86 ×10⁶ km² is well under 1% of the world's land, much less than most people expect. The figure is correct. The difference lies in what is counted. This is the ground that is actually built on: buildings, streets and the paved surfaces around them. It is not the whole area within a city's boundary, which also includes parks, gardens, verges and empty plots and is several times larger. The well-known claim that cities cover about 3% of the planet refers to that wider area.

Satellite surveys of built-up land agree closely with this figure. Rural roads, mines and reservoirs are not counted separately. If the figure is off at all, it is slightly too low for all the land people have physically changed.

Nothing is counted twice

One 9 km square can be partly bare and partly farmed. Subtracting both from the total would count the same ground twice and make wild land look smaller than it is. So where farming and bare ground overlap, the farming counts.

Antarctica is missing from the underlying data altogether. It is left out of the map and out of every figure on this page, instead of being added to desert and ice without saying so.

Crops and grazing are shown together

The map shows a single farmland category. Your diet produces one number for land in total, and that number says nothing about which kind of land. Splitting the two on the map would suggest a shift the figures cannot support. A vegan diet would then seem to need pasture.

Where land changes from one kind to another

Farmland does not appear or disappear evenly across the world. It grows from an edge: land next to an existing field is cleared long before land a hundred kilometres into the forest. That is why every square on the map knows how far it is from the nearest farmland. Growth fills the squares in that order, nearest first and then further out.

More wild land lies close to farming than you might think. About 71% of it is in squares that are already partly farmed. Moderate growth therefore mostly fills in patchy countryside, and opens up little new land.

Shrinking works the other way round. Wild land grows back from its own edges, so the first farmland to be given up is farmland that already borders wild land. About 89% of farmland is in that position. In the most extreme case, the last remaining farmland lies in the Nile valley, Mesopotamia and the Indus valley. These are the dry regions that have been converted most completely, with no wild land nearby to grow back from.

Both are simplifications. They look only at distance and know nothing about soil, water, climate or economics. The distance itself is measured on the real land map.

How many people

With the buttons under World population you can ask the same questions for a world with more people. There are now about 8.2 billion people, and the UN expects a peak of about 10.3 billion in the 2080s. Farmland and built-up land both grow along with that population.

Choose At peak and a slider appears that sets where the population stops growing. It does not draw a new curve. It keeps the shape of the UN projection, including when the growth happens and the slight decline afterwards, and only stretches or squeezes how far above today it rises. Today's figure never moves, because it is measured and not projected. Set the slider below today's 8.2 billion and you get a shrinking world instead of a growing one.

The whole population curve is drawn in the strip below the emissions chart, on the same years. It is worth a look, because population is the only thing that moves the top line of that chart. Nobody's diet changes on that line. It rises only because there are more people to feed. On a running total that shows up as a gentle bend that is easy to miss, so the population is drawn separately, where its shape is clear.

This assumes that food demand per person and farm yields both stay as they are. Neither is likely. Higher yields in particular would reduce the land needed. So read these figures as what happens if nothing else improves.

If a scenario needs more farmland than the planet has, the page says so and shows the shortfall. It does not quietly cap the number, because that would make an impossible world look merely crowded.

The hundred-year emissions chart

The chart What your diet would do to the atmosphere, below the table, shows what all of this would mean for the air. It draws four lines that all measure the same thing: the total amount of greenhouse gas released from today onward, added up year after year.

1. The world as it eats now. What the world's food actually emits today, carried forward as the population grows. Nobody's diet changes on this line. It rises only because there are more people to feed.
2. If everyone ate like you. The same, adjusted by how your meals compare with those of the average person in the Netherlands.
3. Including the carbon that farmland releases or takes back. Line 2, plus the two things land does with carbon: filling up again when it goes back to nature, and continuing to absorb carbon after that.
4. If all the land freed up is grazing land. The same again, but assuming every hectare given back is pasture rather than a mix of pasture and cropland. The real answer lies between these two lines.

Where the world figure comes from

The UN Food and Agriculture Organization (FAO) publishes how much the world's food system emits every year. In 2023 that was 16.5 billion tonnes of CO₂-equivalent, about a third of everything people emit. It is made up of three parts.

The third row covers fertiliser production, processing, packaging, transport, shops, cooking at home and food that is thrown away. That is why this figure is used and not a narrower one. It covers the same part of the food chain as the figures per food everywhere else in this app.

Divided by the number of people alive that year, this is about 2.0 tonnes per person per year. Everything in the chart is this figure multiplied by a number of people and, for your line, by your comparison.

Your comparison, and where 3.78 comes from

your greenhouse gas per day ÷ 3.78 kg CO₂-eq per day

This figure comes from the same report, the same survey and the same food database as the 3.09 for land: RIVM's, based on what 3,570 people in the Netherlands reported eating. Taking both from one source matters more than it seems. A land comparison with one group next to an emissions comparison with another group would be two claims about two different groups of people, shown on one screen as if they were about the same person.

Kilograms of CO₂-equivalent per person per day. Here too, meat alone accounts for more than a quarter. Women average 3.34 and men 4.23.

Here too the comparison is used instead of the absolute figure, for the same reason as with land. The measurements were made for the Dutch food system, so what carries over to another country is more or less than average, and not the number itself.

What the third line adds

Land stores carbon in trees, roots, grass and above all in the soil. Farmland stores much less of it than the forest, savanna or scrub that grew there before. So when farmland makes way for nature, carbon moves from the air back into the ground. When farming takes over wild land, that carbon ends up in the air.

How much depends greatly on the place. A hectare of cleared tropical forest holds many times more than a hectare of dry rangeland. So instead of one world average, this calculation uses a published map of that difference, measured place by place at the same resolution as everything else on this page. It follows exactly the same land, ring by ring, as the world map. The land that changes colour on the map is the land whose carbon changes.

The third line is compared with what the land would do anyway as the population grows, and not with today's world. That is on purpose. A growing population clears land under every diet, including the average one. Blaming that on your diet would make a completely ordinary eater look like someone who cuts down forests. Measured this way, an exactly average diet puts the third line right on top of the second, which is the only answer that makes sense.

The part of the land effect that never stops

Filling up again is not the only thing land does with carbon. The world's plants currently absorb about 11.7 billion tonnes of CO₂ a year, roughly a quarter of everything people emit. They do so in response to the extra CO₂ that is already in the air. That is something different from land growing back after farming, and it does not stop once the regrowth is finished.

The green lines include it in the simplest possible way. It is a measured global figure for the amount of wild land the world has now, so 10% more wild land absorbs 10% more. Spread over today's wild land, that comes to about 1.8 tonnes of CO₂ per hectare per year.

The two can safely be added together, because the people who publish them keep them apart. The Global Carbon Budget books regrowth on abandoned farmland under land-use change, calculated with one set of models. It books this ongoing absorption separately as the land sink, calculated with twenty other models. A published total for the forest sink would have mixed the two and counted the same trees twice.

A one-off deposit and a steady income

This is the most important thing to take from the chart. The three effects each have a different shape. That is why the chart adds everything up, instead of showing how much is released each year.

Eating differently pays off every year. Less greenhouse gas this year, and again next year, and the year after that.

Filling up the soil pays off once. A hectare of returning forest takes up carbon quickly and is then full: after about twenty-five years in the tropics and thirty in temperate regions. After that the ground holds what it can hold. Over a century this is the largest single effect, and it is over before a third of that century has passed.

Wild land keeps paying off, slowly. Less per year than filling up, but it never ends. The further ahead you look, the more it counts. Over a century it adds about half as much again on top of filling up. Over three centuries it would be the larger of the two.

Why there are two green lines

The world has about 16 million km² of cropland and 32 million km² of grazing land, and they are not worth the same. On this grid a hectare of cropland holds 80 tonnes of carbon that can come back, and a hectare of grazing land 37 tonnes. That is more than twice as much. Much grazing land is dry rangeland where the natural vegetation was never forest, so there is simply less carbon to come back.

A diet without meat frees up much more land per kilogram of food than a diet without bread. But it mostly frees up the kind of land that holds less carbon. Instead of picking one and calling that the answer, the chart shows both. The solid line gives land back in the same mix the world farms today. The dashed line gives back only grazing land. For any real change, the truth lies between them.

When the two lines lie on top of each other, the diet no longer frees up any land, because population growth has caught up with it. There is then nothing left to choose. When farmland grows instead, the two lines are the same by definition. What is lost when wild land is cleared is what grew on it, whatever the field is used for afterwards.

The second chart: food as a share of all emissions

The first chart answers how much of the world's food emissions a diet could change. On its own, that invites you to read the headline figure as a statement about the climate as a whole, which it is not. So the second chart draws the same century again, with everything else people emit included.

The total comes from the same FAOSTAT table as the food figure: item 6825, All sectors with LULUCF. LULUCF stands for land use, land-use change and forestry, the category that includes deforestation and regrowth. Both figures weigh greenhouse gases the same way, following the fifth assessment report (AR5) of the IPCC, the UN climate panel. That weighting is used everywhere on this page. Both also come from the same row for the whole world. This matters for more than convenience. FAO's own statement that agrifood systems cause about a third of global emissions is exactly these two items divided by each other. Pairing them here repeats a figure the source itself stands behind, instead of dividing one study by another. In 2023 the total was 52.1 billion tonnes, of which 16.5 billion (32%) came from food.

Both figures include land-use change, which is why the total used is the one with LULUCF. Measuring a food total that includes deforestation against a world total that leaves it out would make food's share look about 3 billion tonnes a year larger than FAO says it is.

Only the food part changes. The other roughly 4.3 tonnes per person per year, from electricity and heating to cement, steel, industry and flights, stays per person exactly as it is today, in all four lines. That is not a forecast that those emissions stay the same. Nobody expects that. It means that diet does not affect them. That is exactly what lets you read the gap between the lines as the effect of eating differently, and of nothing else.

The saving is the same number of tonnes in both charts. Only what it is measured against changes. A saving worth several tens of percent of the world's food emissions comes closer to a tenth when measured against everything people emit. Both figures are correct. They answer different questions. A figure above 100% is not a mistake either. Land that is given back absorbs carbon as well as avoiding emissions, so the food system on its own can end the century taking more carbon out of the air than it puts in. The economy as a whole cannot, and the second chart shows that most clearly.

What this deliberately leaves out

• It does not model the atmosphere. It counts tonnes released, not degrees of warming. Turning one into the other needs a climate model.
• It assumes everyone's food changes at the same moment and then stays that way. That never happens in reality. Read the totals as the size of what could be gained, and not as a forecast.
• The carbon map covers land used for animal feed and grazing. That is most farmland, but not all of it. Cropland for food that people eat directly takes its figure from the nearest place the map does cover.
• It measures how much carbon land could hold if it were left alone. Much wild land is already grazed or logged and holds less. So the carbon lost when farmland expands is an upper limit rather than a best estimate.
• Carbon is assumed to leave the land as slowly as it comes back. In reality, clearing releases it much faster than regrowth takes it up. So the first decades of a scenario with growing farmland look milder here than they really would be. Over the total of a century it makes little difference.
• The ongoing absorption is kept the same for a hundred years. In reality it varies: in the last ten years alone it moved between 8.3 and 14.2 billion tonnes, mainly because of El Niño. That is why a ten-year average is used rather than a single year. The absorption is also expected to weaken as the extra CO₂ that drives it stops being extra. That part of the green lines is therefore the optimistic reading.
• It assumes farmland adds nothing to that ongoing absorption. Crops do take up carbon, but they are harvested and eaten within the year, so little of it stays behind.
• It counts carbon going into and out of the land, and greenhouse gases from the food system. It counts nothing else that people do.

Why the tab sometimes shows nothing at all

Only about one in five foods in this app has land-use figures. On an ordinary day the figure therefore covers only part of what you ate. A map built on that would show too little farmland, for the wrong reason.

That is why the World tab only uses days on which every food you entered had land-use figures. If there is no such day, the tab says so and draws nothing. Choose All logged days to see an estimate based on incomplete days instead. That estimate is marked as an underestimate everywhere it appears.

The emissions chart applies the same rule to greenhouse gas separately. A day can have complete land figures and incomplete emissions figures. The map can then appear while the chart still says it has nothing to go on. The chart does not borrow the land figure to fill that gap.

Where the figures come from

• Dutch food figures: RIVM report 2024-0101, on the environmental impact of food consumption in the Netherlands from 2019 to 2021.
• Farmland and built-up land: HYDE 3.5, a long-running dataset on land use by PBL Netherlands Environmental Assessment Agency and Utrecht University.
• Bare ground: satellite land cover data from the European Space Agency, 2022.
• Glaciers and ice: Natural Earth.
• Carbon held by land: Hayek, Harwatt, Ripple and Mueller, The carbon opportunity cost of animal-sourced food production on land, Nature Sustainability, 2021.
• World food emissions and total human emissions: FAOSTAT Emissions Totals, the yearly overview by the UN Food and Agriculture Organization. Both figures come from the same table and weigh greenhouse gases the same way, which is what makes one a share of the other.
• How much carbon land absorbs: the Global Carbon Budget 2024, the yearly overview of the world's carbon flows by the Global Carbon Project.
• Population: United Nations World Population Prospects 2024, medium variant.

The map uses the Mollweide projection, which draws every area at its true relative size. The more familiar rectangular world map stretches land near the poles, which makes Greenland look much bigger than it is. A page about how much land there is needs a map that gets area right.

Want to see this with your own figures? Go back to the app, fill in a few days of meals and open the World tab.