Where the emissions sit, why they are hard to abate, and how the sector reaches net zero. Paper 1 sets the baseline the whole series builds on.
Agriculture and the food system feed eight billion people and support livelihoods across most of the developing world. They also account for close to a third of human-caused greenhouse gas emissions. Most of those emissions are biological rather than industrial. They come from cattle, from nitrogen applied to fields, and from forests cleared for farmland, not from a smokestack that a cleaner fuel can fix. That is what makes them hard to measure and hard to cut, and it is part of why the sector has attracted less climate investment than its footprint warrants.
Agrifood emissions reached 16.5 billion tonnes of CO₂ equivalent in 2023, 21 percent higher than in 2001. The sector’s share of the global total fell over the same period, from 38 percent to 32 percent, but that is because other sectors grew faster, not because agriculture shrank. Output is getting cleaner per unit — emissions per dollar of agricultural production dropped 25 percent since 2001, and per-person agrifood emissions edged down to 2.0 tonnes — but the absolute total keeps rising.
| Component | 2023 (Gt CO₂e) | Share of agrifood (2001 → 2023) | Change since 2001 |
|---|---|---|---|
| Total agrifood systems | 16.5 | 100% | +21% |
| Farm-gate (crop + livestock) | 8.1 | 49% → 49% | +17% |
| Pre- & post-production (supply chain) | 5.2 | 26% → 32% | +33% |
| Land-use change | 3.2 | 25% → 19% | −6% |
Share gives each component’s portion of agrifood emissions in 2001 and 2023; the final column is the change in that component’s own emissions. Agrifood is 32% of all global emissions (52.1 Gt), down from 38% in 2001.
Each subsector sits at a stage in the value chain and is tagged with the gases behind its emissions. The value-chain view shows who operates at each stage of production; the emissions view shows where the carbon sits and which levers will move it. The series takes these on one at a time.
| Subsector | Covers | Dominant emissions (gas) |
|---|---|---|
| Primary production — farm gate · 49% of sector emissions | ||
| Cattle & ruminant livestock | Beef, dairy, sheep, goats | Enteric methane, manure (CH₄, N₂O) |
| Non-ruminant livestock | Pigs, poultry | Manure management, feed (CH₄, N₂O) |
| Cereals & staple crops | Rice, wheat, maize | Rice-paddy methane, soil nitrous oxide (CH₄, N₂O) |
| Oilseeds, fruit, vegetables, other crops | Soy, palm, sugar, coffee, cocoa | Soil N₂O; land-use change (N₂O, CO₂) |
| Inputs and upstream | ||
| Fertilizer & agrochemicals | Synthetic N/P/K, pesticides, ammonia | Ammonia production, field N₂O (CO₂, N₂O) |
| Farm machinery, energy, fuel | Tractors, irrigation, on-farm fuel | Fuel combustion (CO₂) |
| Processing, manufacturing, packaging — part of the 32% supply chain | ||
| Food & beverage processing | Manufacturing, refrigeration, dairy/meat, sugar | Process heat, refrigerants (CO₂, F-gases) |
| Pulp, paper & packaging | Pulp mills, paper, board, packaging | Process heat and energy — largely biogenic (CO₂) |
| Downstream, land use and waste | ||
| Land use & land-use change | Deforestation, peatland, land clearing | Carbon released from cleared land (CO₂) |
| Distribution, retail, consumption | Cold chain, transport, retail | Transport, refrigerants (CO₂, F-gases) |
| Food loss, waste, end-of-life | Food waste, landfill, wastewater | Landfill methane (CH₄) |
Agrifood emissions come from three different gases, and each behaves differently. Most industrial sectors deal mainly with carbon dioxide, which clean energy can address. Agriculture adds two more — methane and nitrous oxide — both biological, both far more potent than CO₂ tonne for tonne, and neither responsive to a fuel switch. Once you know which gas a subsector emits, you know most of what you need about its levers and the carbon market it can reach.
| Methane (CH₄) | Nitrous oxide (N₂O) | Carbon dioxide (CO₂) | |
|---|---|---|---|
| Main sources | Cattle & ruminants, rice paddies, manure, food waste | Over-applied nitrogen on soils; synthetic fertilizer & manure | Land-use change (~2.8 Gt/yr from deforestation); post-farm energy |
| Potency vs CO₂ | ~80× over 20 years | ~265× over a century | 1× — the baseline |
| Lifetime | ~Two decades — cuts buy fast relief | More than a century | Centuries to millennia |
| Key levers | Feed additives (3-NOP, red seaweed), paddy drainage, anaerobic digesters | Enhanced-efficiency fertilizers, nitrification inhibitors, precise dosing; green ammonia upstream | Halt clearing & restore land; electrify heat, clean power, CCS on biomass |
| Why it matters | IPCC: cut 24–47% below 2010 by 2050; measurable cuts make strong credits | Synthetic-N chain alone is ~1.13 Gt CO₂e/yr — a tenth of agricultural emissions | Where the largest nature-based removal opportunities sit |
One table, three abatement logics — the gas tells you the levers, the measurement, and the market.
Read a subsector by its gas
Livestock and rice are methane stories. Fertilizer and intensively farmed soils are nitrous-oxide stories. Land use and processing are carbon-dioxide stories. The gas points to the levers, to the way they are measured, and to the market that will pay for them.
Climate Decode builds the emissions baseline, the lever screen, and the financed plan that agrifood decarbonization needs — subsector by subsector.
Agriculture is the largest user of land on the planet. The most thorough global study to date, covering 38,700 farms across 119 countries, finds that about 43 percent of the world’s ice- and desert-free land is farmed. Roughly 83 percent of that land goes to animal-sourced food, which returns only about 18 percent of calories and 37 percent of protein.
Want this baseline built for your portfolio or supply shed?
Speak to an Expert →Agriculture’s footprint reaches well beyond greenhouse gases. The food system is the leading driver of deforestation and biodiversity loss, and farming is listed as a threat to most of the species judged to be at risk of extinction. Clearing land for crops and pasture is the main way food production destroys habitat.
Water is the other large pressure. Agriculture is usually said to take about 70 percent of the world’s freshwater withdrawals — a figure repeated everywhere but resting on thinner evidence than its popularity suggests, so we treat it as an indicator of scale rather than a precise number. These pressures matter to the climate case because the measures that cut emissions — restoring soil carbon, applying less fertilizer, halting deforestation — also ease the strain on water, soil, and wildlife.
No single measure decarbonizes agriculture. Each subsector needs its own set of levers, matched to the gas it emits. They group into five families:
| Lever family | What it does | Where it applies |
|---|---|---|
| 1 · Livestock methane | Feed additives, productivity, manure capture & biogas | Cattle, dairy, pigs, poultry |
| 2 · Efficient nitrogen | Enhanced-efficiency & low-carbon fertilizer, precise dosing | Crops, fertilizer production |
| 3 · Soil & land carbon | Soil carbon, halted conversion, restoration — opens NBS removals | Land use, crops, forestry |
| 4 · Clean processing | Electrified heat, heat pumps, clean power, CCS on biomass | Food & beverage, pulp & paper |
| 5 · Loss & waste | Cold-chain efficiency, waste-to-value, landfill methane capture | Distribution, retail, end-of-life |
Ranking levers by cost per tonne alone sends capital to the wrong projects. Efficiency measures such as heat recovery abate little carbon on their own, so they look expensive per tonne, but they have to come first: they shrink the energy load before the larger electrification and fuel-switching projects are sized. Skip them, and the new equipment is built for an oversized baseline and the capital bill climbs.
The sequencing rule
A workable sequence weighs four things at once — the strategic role of each measure, the cost of saved energy, the simple payback, and the net present value — rather than a single ratio. That is the analysis TerraNova runs automatically.
TerraNova is Climate Decode’s decarbonization platform for industrial and agrifood companies. It takes the workflow this paper describes — operations data, hotspots, levers, the policy and carbon-market backdrop, and financing — and turns it into a decision-ready plan for a specific site. For each facility it builds a greenhouse-gas inventory on the right regulatory basis, assigns emissions to the process steps that cause them, screens the available levers against local carbon prices and incentives, and returns a ranked, sequenced investment plan with the net present value, return, and payback of each project and the year-by-year compliance position.
This paper is the first in the series and sets the baseline the rest build on. Later papers carry the same analysis into individual subsectors, with the numbers and framing kept consistent from one to the next. The case is the same throughout: the agrifood transition is large, it can be measured, and it can be financed.
SBTi FLAG for Agrifood · CBAM & Fertilizer · Supplier Engagement & Scope 3
Notes and sources for the figures in this paper.
Climate Decode turns agrifood decarbonization into a measured, sequenced, financeable program — TerraNova for the reductions, Canopy for the residuals and removals.