Sustainability & Environment (Agri-Food Systems) · Sustainability Assessment, Carbon & Governance · also GHG Accounting, Corporate Carbon Footprinting, National GHG Inventories, Carbon Management
Counting greenhouse gases for nations, organisations, farms and products: IPCC inventory guidelines, the GHG Protocol corporate and Scope 3 standards, land-sector guidance, ISO 14064, emission factors and tiers, and science-based targets.
In the exams: GARP SCR 'Transition Planning and Carbon Reporting'; ISO 14064 verifier courses; sustainability reporting roles; GATE ES Section 9
Carbon accounting is the discipline of counting greenhouse gases (GHGs) released to and removed from the atmosphere by a defined entity over a defined period, and reporting the result so that someone else can check it. The entity may be a country, a company, a farm, a single product or a project. The same physical events -- a cow belching methane, a tractor burning diesel, a field releasing nitrous oxide after fertiliser, a hedge growing and storing carbon -- are counted in all of these inventories, but each inventory draws its boundary differently, uses different rules and answers a different question. A national inventory asks what happened on the country's territory; a corporate inventory asks what the company is responsible for, directly and through its value chain; a product footprint asks what one kilogram of milk or wheat carried with it to the farm gate or the shelf.
The subject matters because nearly every climate commitment now rests on a number produced by these methods. Countries report to the United Nations Framework Convention on Climate Change (UNFCCC) using the Intergovernmental Panel on Climate Change (IPCC) Guidelines. Companies report under the Greenhouse Gas Protocol (GHG Protocol) and ISO 14064, set targets validated against science-based pathways, and increasingly have those numbers assured by an independent party, as financial statements are. In agriculture and food the stakes are unusually high: methane and nitrous oxide from livestock, soils and rice dominate farm emissions, land-use change can dwarf everything else, and land can also remove carbon -- which brings difficult questions about permanence, measurement and double counting.
This manual follows the twelve units of the curriculum: why and how GHGs are counted, the IPCC Guidelines and their Agriculture, Forestry and Other Land Use (AFOLU) volume, tiers and emission factors, the GHG Protocol Corporate and Scope 3 standards, land-sector and removals accounting for companies, farm carbon calculators, the ISO 14064 series, data quality and uncertainty, base years and targets, science-based and net-zero targets including the Forest, Land and Agriculture (FLAG) guidance, and verification and assurance. Worked examples use farms and food companies. Frameworks change: the GHG Protocol standards and the Science Based Targets initiative (SBTi) net-zero standard are under revision, so the version described is stated wherever it matters.
Worldwide, carbon accounting is examined in professional certificates such as the GARP Sustainability and Climate Risk certificate (its 'Transition Planning and Carbon Reporting' area), in ISO 14064 verifier and GHG-auditor training, and in sustainability-professional credentials; in India it appears in GATE Environmental Science and Engineering (Sections 8 and 9), UGC NET Environmental Sciences and the ICAR PhD paper in Environmental Sciences. Expect conversion to CO2-equivalent, scope classification, Tier 1 arithmetic, base-year recalculation and target arithmetic -- and expect the examiner to state which GWP values to use.
A GHG inventory is a list of emission sources and removal sinks with the quantity of each gas emitted or removed over a reporting period, usually a calendar or financial year. Inventories serve four broad purposes. Compliance: emissions trading systems and mandatory reporting laws require named installations to report, often with verification. Disclosure: investors, lenders and buyers ask companies for emissions data, and sustainability-reporting standards now require it in many jurisdictions. Management: a company cannot reduce what it has not measured, and an inventory shows where the large sources lie. Policy: national inventories track progress against Nationally Determined Contributions (NDCs) under the Paris Agreement.
Five scales are common. National inventories follow the IPCC Guidelines and cover all emissions and removals on a country's territory. Corporate or organisational inventories follow the GHG Protocol Corporate Standard or ISO 14064-1 and cover what the organisation controls or owns, plus its value chain. Product footprints follow the GHG Protocol Product Standard, ISO 14067 or sector rules, and divide emissions across the life cycle of a unit of product. Project accounting follows ISO 14064-2 or a carbon-crediting programme and measures reductions against a baseline. Farm inventories borrow from all of them: IPCC equations for the biology, a corporate-style boundary, and a product intensity for buyers.
The principles are shared. The GHG Protocol names five -- relevance, completeness, consistency, transparency and accuracy. The IPCC asks national inventories to be transparent, accurate, complete, consistent and comparable (often abbreviated TACCC). ISO 14064-1 adds conservativeness for some uses. Learn these lists; examiners ask them directly and use them to judge cases.
The Kyoto Protocol basket, used by the GHG Protocol and most national rules, has seven gases or groups: carbon dioxide (CO2), methane (CH4), nitrous oxide (N2O), hydrofluorocarbons (HFCs), perfluorocarbons (PFCs), sulphur hexafluoride (SF6) and nitrogen trifluoride (NF3), the last added for the second commitment period. Ozone-depleting gases controlled by the Montreal Protocol, such as CFCs and HCFCs, are not part of the scope totals; the GHG Protocol allows them to be reported separately.
To add gases together, each mass is multiplied by its global warming potential (GWP): the time-integrated radiative forcing of a pulse emission of one kilogram of the gas, relative to one kilogram of CO2, over a chosen horizon -- almost always 100 years (GWP100). The result is CO2-equivalent (CO2e). GWP values are revised in each IPCC assessment report, so a stem must say which set to use. The 100-year values most often met are: Second Assessment Report (SAR), CH4 21 and N2O 310; Fourth (AR4), CH4 25 and N2O 298; Fifth (AR5), CH4 28 and N2O 265 without climate-carbon feedbacks (34 and 298 with them); Sixth (AR6), CH4 27.0 for non-fossil methane, 29.8 for fossil methane, and N2O 273. Reporting under the Paris Agreement's enhanced transparency framework uses AR5 GWP100 values; companies are generally asked to use the most recent IPCC values, or those their programme specifies, and to state the source.
GWP100 is a convention, not a law of nature. Methane is short-lived (an atmospheric lifetime of about a decade), so over 20 years its GWP is far higher (AR5 GWP20: 84 without feedbacks). The global temperature change potential (GTP) and the 'warming-equivalent' metric GWP* treat short-lived gases differently and are argued over for livestock. Inventories nevertheless report in GWP100; good practice is to also report each gas in tonnes so readers can apply other metrics.
A final conversion recurs everywhere: one tonne of carbon is 44/12, or 3.667, tonnes of CO2, because the molar mass of CO2 (44) divided by that of carbon (12) is 3.667. Similarly, nitrous oxide reported as N2O-N is multiplied by 44/28 to give N2O.
CO2 from burning or decomposing biomass (crop residues, wood fuel, biogas) is biogenic: the carbon was recently taken from the air by plants. The GHG Protocol requires biogenic CO2 from combustion to be reported separately, outside the scopes, while CH4 and N2O from the same combustion stay inside them. Enteric methane is counted as CH4 even though its carbon is biogenic, because methane warms far more than the CO2 the plant absorbed. Net changes in carbon stocks on land are handled separately, as land-use emissions and removals (chapter 6).
For a farm, the practical consequences are clear. Diesel and petrol burned in machinery are fossil CO2 and sit in Scope 1. Straw or wood burned for heat or drying releases biogenic CO2, which is reported outside the scopes, but its CH4 and N2O are counted in Scope 1. Methane captured from a covered manure store and flared becomes biogenic CO2, which is why flaring or using biogas cuts the CO2e total so sharply: a tonne of methane is replaced by about 2.75 tonnes of biogenic CO2 that is not added to the scope totals.
The IPCC Task Force on National Greenhouse Gas Inventories (TFI) writes the methods countries use. The current basis is the 2006 IPCC Guidelines for National Greenhouse Gas Inventories, updated by the 2019 Refinement, with supplements such as the 2013 Wetlands Supplement. The 2006 Guidelines have five volumes: Volume 1 General Guidance and Reporting; Volume 2 Energy; Volume 3 Industrial Processes and Product Use (IPPU); Volume 4 Agriculture, Forestry and Other Land Use (AFOLU); and Volume 5 Waste. The 2019 Refinement does not replace the 2006 Guidelines; it updates, supplements and elaborates them, and countries choose which parts to apply as their reporting rules allow.
The guidelines are territorial: a country reports emissions and removals that occur within its territory and offshore areas over which it has jurisdiction. Emissions from international aviation and shipping bunkers are reported as memo items, not in national totals. Emissions are reported for the year in which they occur. Each source is estimated in a category of the IPCC structure, such as 3.A Enteric fermentation in the 2006 structure.
Under the UNFCCC, developed (Annex I) countries have long submitted annual inventories -- a National Inventory Report and common tables -- that undergo expert review. Under the Paris Agreement, all Parties report through Biennial Transparency Reports under common modalities, procedures and guidelines, with flexibility for developing countries that need it. A national inventory system includes institutional arrangements, a quality assurance/quality control (QA/QC) plan, archiving and a key category analysis, repeated every cycle.
Notation keys fill cells where no number is reported: NO (not occurring), NE (not estimated), NA (not applicable), IE (included elsewhere) and C (confidential). Using them correctly is part of transparency: an empty cell tells a reviewer nothing.
A key category is one that has a significant influence on the total inventory in level, trend or uncertainty. In Approach 1, categories are ranked by their absolute contribution and the key categories are those which, added in descending order, reach 95% of the total level (and a parallel test for trend). Approach 2 weights each contribution by its uncertainty. Key categories are where an inventory compiler should spend effort: good practice is to use higher-tier methods for them.
Quality control (QC) is the routine technical checking done by the compilers themselves -- arithmetic, units, transcription, comparison with previous years. Quality assurance (QA) is review by people not directly involved in compiling the inventory, such as peer reviewers or an independent body.
A time series must be consistent: the same method and data sources for every year, so that a trend reflects real change, not a change in method. When a method improves, the whole series is recalculated. Where data for early years are missing, the guidelines give splicing techniques: overlap, surrogate data, interpolation and trend extrapolation.
Volume 4 brings agriculture and land use together because the same land is involved. Its chapters cover generic methods (chapter 2), consistent representation of lands (3), the six land-use categories -- forest land, cropland, grassland, wetlands, settlements and other land (chapters 4 to 9) -- emissions from livestock and manure management (10), N2O from managed soils and CO2 from liming and urea application (11), and harvested wood products (12).
Land is represented in one of three approaches: Approach 1 gives total areas per land category without tracking conversions; Approach 2 tracks conversions between categories in a non-spatial matrix; Approach 3 tracks land-use change spatially, plot by plot, usually from remote sensing. Each category is split into land remaining in that use and land converted to it; converted land is conventionally kept in the 'converted' class for 20 years.
Five carbon pools are recognised: aboveground biomass, belowground biomass, dead wood, litter and soil organic matter (harvested wood products are treated as a further pool). Stock changes can be estimated by the gain-loss method (annual growth minus annual losses) or the stock-difference method (two inventories of stock, divided by the years between them). Only managed land is reported; the managed-land proxy treats emissions and removals on managed land as anthropogenic.
In the 2006 structure agriculture's non-CO2 sources include enteric fermentation, manure management (CH4 and N2O), rice cultivation, agricultural soils (direct and indirect N2O), field burning of residues, liming and urea application. Agricultural energy use, such as tractor diesel, is reported in the Energy sector, not in AFOLU -- a frequent source of confusion when national and farm figures are compared.
Almost every IPCC method has the form: emissions = activity data x emission factor. Activity data (AD) describe the size of the activity -- head of cattle, tonnes of nitrogen applied, hectares of rice, litres of diesel. The emission factor (EF) gives emissions per unit of activity. Tiers describe how specific the method and data are.
Tier 1 uses the simplest equations with IPCC default emission factors, usually from global or regional tables, and activity data often from national statistics. Tier 2 uses the same general structure but with country- or region-specific emission factors and more disaggregated activity data, such as cattle split by productivity class. Tier 3 uses higher-order methods: process-based models, detailed inventory measurement systems or direct measurement, repeated over time and space and validated. Higher tiers are expected to be more accurate and less uncertain, but only if they are properly documented and checked. Decision trees in each chapter tell the compiler which tier is good practice, usually depending on whether the category is key.
Default emission factors vary by region, livestock productivity, climate and management, and they changed between the 2006 Guidelines and the 2019 Refinement. That is why this manual quotes the structure of the equations but leaves the number in each worked question to the stem: look up the table that matches your country and the guidelines version your programme requires.
Enteric methane, Tier 1: CH4 = number of animals x EF (kg CH4 per head per year), summed over animal categories. Tier 2 builds the factor from feed energy: EF = (GE x Ym/100 x 365) / 55.65, where GE is gross energy intake in MJ per head per day, Ym is the methane conversion factor (the percentage of gross energy converted to methane) and 55.65 MJ/kg is the energy content of methane. Gross energy intake in turn is estimated from the animal's net energy requirements for maintenance, activity, growth, lactation and pregnancy and the digestibility of the diet. Ym depends on diet; values for cattle differ by feeding system and between guideline versions.
Manure methane, Tier 2: EF = VS x 365 x B0 x 0.67 x sum over systems (MCF x MS), where VS is volatile solids excreted per day, B0 the maximum methane-producing capacity (m3 CH4 per kg VS), 0.67 the density of methane in kg per m3, MCF the methane conversion factor of each manure management system and MS the fraction of manure handled in it. MCF is high for liquid systems such as lagoons in warm climates and low for pasture or daily spread, which is why manure management is often the first mitigation lever examined.
Manure N2O has direct emissions from storage (nitrogen excreted x fraction in each system x EF3 x 44/28) and indirect emissions from volatilised and leached nitrogen.
Direct N2O from managed soils, Tier 1: N2O-N = (F_SN + F_ON + F_CR + F_SOM) x EF1 + flooded-rice and organic-soil terms + grazing deposition x EF3PRP. F_SN is synthetic fertiliser N, F_ON organic N (manure, compost), F_CR crop residue N and F_SOM N mineralised from soil organic matter lost through land-use or management change. The 2006 default EF1 is 0.01 kg N2O-N per kg N; the 2019 Refinement keeps 0.01 as an aggregated default but also gives disaggregated factors by climate and fertiliser type. Multiply N2O-N by 44/28 to get N2O.
Indirect N2O follows two pathways. Volatilisation: a fraction of applied N is lost as ammonia and nitrogen oxides (Frac_GASF for synthetic, Frac_GASM for organic N) and later redeposited, emitting N2O with factor EF4. Leaching and runoff: in regions where it occurs, a fraction Frac_LEACH is lost to water, emitting N2O with factor EF5. These fractions and factors also changed between versions, so stems give them.
CO2 from liming uses 0.12 t C per tonne of limestone (CaCO3) and 0.13 t C per tonne of dolomite (CaMg(CO3)2); CO2 from urea uses 0.20 t C per tonne of urea. Each is multiplied by 44/12. These factors come from the chemistry, which is why they are the same everywhere at Tier 1.
Rice methane: CH4 = daily emission factor x cultivation period (days) x area, where the daily factor is a baseline for continuously flooded fields without organic amendment scaled by factors for water regime during and before the season and for organic amendments. Drainage (as in alternate wetting and drying) lowers the water-regime scaling factor.
The GHG Protocol Corporate Accounting and Reporting Standard (revised edition, 2004), developed by the World Resources Institute (WRI) and the World Business Council for Sustainable Development (WBCSD), is the most widely used basis for company inventories. Its first decision is the organisational boundary: which operations belong to the company. Two approaches are allowed. Under the equity share approach the company accounts for emissions in proportion to its economic interest in each operation. Under the control approach it accounts for 100% of emissions from operations it controls and none from operations it has an interest in but does not control. Control can be defined as financial control (the ability to direct financial and operating policies with a view to gaining economic benefits) or operational control (full authority to introduce and implement operating policies). The company must apply one approach consistently across its inventory.
The choice matters in agri-food, where joint ventures, cooperatives, contract growers and leased farms are common. A food processor with a 40% stake in a joint-venture dairy plant that it operates reports 100% of the plant's emissions under operational control but 40% under equity share. A cooperative that neither owns nor operates its members' farms reports their emissions in Scope 3, not Scope 1.
Operational boundaries classify emissions by scope. Scope 1 covers direct emissions from sources owned or controlled by the company: fuel combustion in boilers, dryers, tractors and trucks; process emissions; fugitive emissions such as refrigerant leaks; and, for a farm, enteric methane, manure emissions and soil N2O on land it manages. Scope 2 covers indirect emissions from the generation of purchased or acquired electricity, steam, heat and cooling consumed by the company. Scope 3 covers all other indirect emissions in the value chain. Scopes 1 and 2 are required in the Corporate Standard; Scope 3 was optional there and now has its own standard.
The Scope 2 Guidance (2015) requires dual reporting where contractual instruments exist. The location-based method uses average grid emission factors for the grids where electricity is consumed. The market-based method uses emission factors from contractual instruments -- energy attribute certificates, direct power purchase agreements, supplier-specific rates -- and, for consumption not covered by any instrument, a residual mix factor, falling back to grid average only where no residual mix is available. Instruments must meet the Scope 2 quality criteria (for example, they must convey the attribute claim, be retired or cancelled for the user, be issued close to the period of consumption and come from the same market).
Common farm and food-company traps: diesel burned in owned tractors is Scope 1, not Scope 2; electricity generated by the company's own solar panels and consumed on site has no Scope 2 emissions; electricity sold to the grid is not subtracted from Scope 2; carbon credits bought by the company are not deducted from any scope and are reported separately; and avoided emissions (for example, from a product that saves energy for its users) are not part of the inventory.
The GHG Protocol Agricultural Guidance (2014) interprets the Corporate Standard for farms. It recommends reporting agricultural emissions by source category -- mechanical sources, enteric fermentation, manure, soils, rice, burning, land use -- as well as by scope, separating CO2 from land-use change and carbon stock changes from other emissions, and reporting non-CO2 gases in tonnes as well as CO2e. Its message for exam answers is that agricultural inventories follow the same scope logic as any company, with IPCC equations supplying the numbers for biological sources.
Two boundary questions recur on farms. First, land rented in and managed by the farmer is normally inside the farmer's operational control, so its soil and livestock emissions are Scope 1 of the farm business; land rented out to another operator is not. Second, contract-rearing arrangements -- for example, a poultry integrator that owns the birds and feed while a grower owns the sheds and manages the flock -- must be analysed case by case under the chosen control definition, and the same allocation must be used every year. When a boundary is unclear, the inventory report should state the rule applied so that a verifier and a reader can follow it.
The Corporate Value Chain (Scope 3) Standard (2011) divides Scope 3 into fifteen categories. Upstream: 1 purchased goods and services; 2 capital goods; 3 fuel- and energy-related activities not included in Scopes 1 or 2 (upstream emissions of fuels and electricity, and transmission and distribution losses); 4 upstream transportation and distribution; 5 waste generated in operations; 6 business travel; 7 employee commuting; 8 upstream leased assets. Downstream: 9 downstream transportation and distribution; 10 processing of sold products; 11 use of sold products; 12 end-of-life treatment of sold products; 13 downstream leased assets; 14 franchises; 15 investments.
For a food manufacturer or retailer, category 1 -- the farm and processing emissions embedded in purchased ingredients -- is usually the largest single item, often most of the total footprint. For a fertiliser company, category 11 (use of sold products) includes the N2O released when its fertiliser is applied in fields, which is the farmer's Scope 1. One company's Scope 1 is another's Scope 3: double counting across companies in a chain is expected and intended, because each reports its own responsibility. Double counting within one inventory, however, is an error.
The Scope 3 technical guidance describes a hierarchy of methods. The supplier-specific method uses cradle-to-gate data from the supplier for the product bought. The hybrid method combines supplier data on its own Scope 1 and 2 with secondary data for the rest. The average-data method multiplies the mass or quantity bought by a secondary emission factor, such as kg CO2e per kg of wheat from a life-cycle database. The spend-based method multiplies money spent by an environmentally extended input-output (EEIO) factor in kg CO2e per currency unit. Spend-based results are quick but crude: they move with prices, not with farming practice, so a supplier's real reduction cannot show up in them. Mature programmes start spend-based to find hotspots and move the largest categories to activity-based and supplier-specific data.
For transport, emissions are often estimated as tonne-kilometres x a factor per tonne-km for the mode. For waste, the quantity sent to each treatment route is multiplied by a route-specific factor. For employee commuting, survey data on distance and mode are scaled to the workforce.
Companies should account for all relevant Scope 3 emissions and justify exclusions. Relevance is judged with criteria in the standard: size, influence, risk, stakeholder interest, outsourcing and sector guidance. A screening exercise -- often spend-based -- ranks categories, after which data collection effort follows the size of each category. Emissions from purchased agricultural raw materials include land-use change where it occurs, which can make sourcing region decisive; soy or palm oil from recently converted land carries a much higher footprint than the same commodity from long-established farmland. Supplier engagement -- asking farms and processors for primary data, sharing calculation tools and paying for practice change -- is the main way food companies reduce Scope 3. The emission factors and the methods used for each category should be disclosed so a reader can compare companies.
Land both emits and removes CO2, and the same hectare may do both in different years. Three groups of land-related flows matter to a company. Land-use change (LUC) emissions arise when land is converted from one use to another -- above all, natural forest or grassland to cropland or pasture -- and carbon stored in vegetation and soil is lost. Land-management emissions and removals arise on land remaining in the same use, from changes in soil carbon or trees under different practices. Non-CO2 emissions from agriculture (enteric, manure, soils, rice) are counted as in the IPCC methods. Removals are CO2 taken from the atmosphere and stored in biomass, soils or products.
The GHG Protocol has developed land-sector and removals rules (released as draft guidance in 2022 for pilot testing and being finalised as a standard; check the current text before applying it). Key ideas that have been stable through drafts are: land-use change emissions are spread over an assessment period, commonly 20 years, so a conversion counts against products from that land for 20 years; land emissions and removals are reported separately from each other and from other emissions, never simply netted; and a company reporting removals must show ongoing storage, traceability to the land where the carbon is stored, quantification with stated uncertainty, and an obligation to account for reversals if the carbon is later lost.
Where a company can trace a commodity to the plot of origin and knows when that plot was converted, it can calculate direct land-use change (dLUC). Where it cannot, statistical land-use change (sLUC) attributes the recent expansion of a crop in a country or region to all the production from that area. sLUC spreads emissions across all producers in a region, so a company sourcing from a deforestation-free supplier in a high-deforestation country may still carry a large sLUC figure until it can show traceable data. The 20-year assessment period follows earlier product standards and the IPCC convention for converted land.
Indirect land-use change -- displacement of production elsewhere through markets -- is a modelling concept used in biofuel policy; corporate inventories do not include it.
Removals on farms come from trees (agroforestry, hedges, woodlots), from soil organic carbon gains and, in some products, from durable wood. A removal is only credible while the carbon stays stored. Soil and biomass carbon are reversible: ploughing, drought, fire or a change of land manager can return the carbon to the air. That is why removal claims need monitoring for as long as they are reported, and why reversals must be counted as emissions. Soil carbon also saturates: after a practice change, gains slow and approach a new equilibrium, often over a couple of decades, so a farm's sequestration rate is not a permanent annual figure.
Removals inside a company's value chain (insetting) are different from carbon credits bought from outside it (offsetting). Removals that meet the requirements may be reported separately in the inventory, while bought credits are never included in scope totals. Conflating the two is one of the commonest errors in food-company net-zero claims.
Farm carbon calculators turn farm records into an emissions estimate. The user enters activity data: herd numbers by class, liveweights, milk yield, feed rations, fertiliser and manure applications, crop areas and yields, residue management, fuel and electricity use, and sometimes land-use history and tree cover. The calculator applies IPCC Tier 1 or Tier 2 equations, sometimes with country-specific factors, and for soil carbon may apply an empirical factor approach or a process-based (Tier 3) model. Outputs are typically total emissions by source and gas, an intensity per unit of product (kg CO2e per kg of milk, per kg of liveweight, per tonne of grain) and a comparison of mitigation scenarios.
Public tools include the Food and Agriculture Organization's EX-ACT (ex-ante carbon-balance tool for projects and policies) and its GLEAM models for livestock, and the United States Department of Agriculture's COMET-Farm; many countries and buyer programmes run their own. Tools differ in boundary (farm gate or field), in whether upstream emissions of feed and fertiliser are included, in default factors, in soil-carbon methods and in allocation rules. Studies that run the same farm through several calculators have found large differences in the result, so a figure is meaningful only with its tool, version and settings.
A dairy farm sells milk and also culled cows and calves, so its emissions must be shared between products. Milk is first standardised: the International Dairy Federation's fat and protein corrected milk (FPCM) formula is FPCM (kg) = milk (kg) x (0.1226 x fat% + 0.0776 x true protein% + 0.2534). The IDF also gives a biophysical allocation for milk: AF_milk = 1 - 6.04 x BMR, where BMR is the ratio of meat (liveweight sold) to milk (FPCM) mass. Economic allocation (by revenue share) and mass allocation are alternatives; product category rules usually prescribe one. Crop farms face similar choices between grain and straw.
Intensity metrics can fall while absolute emissions rise -- a herd producing more milk per cow may have lower kg CO2e per kg FPCM but more total emissions if it grows. Buyers using farm data for Scope 3 need intensity; national targets and the atmosphere respond to absolute tonnes. Report both.
Good practice for an adviser: record the tool and version, the GWP set and the guidelines version; check activity data against invoices and herd records; never compare intensities from different tools; use the same tool across years for trend; document any farm-specific factor (for example a measured manure nitrogen content); and treat soil-carbon outputs as estimates with wide uncertainty unless they are backed by repeated soil sampling on an equivalent-soil-mass basis. When farm data feed a buyer's Scope 3 inventory, data-sharing agreements should settle who owns the data and whether reductions may also be sold as credits, to avoid the same tonne being claimed twice.
ISO 14064-1:2018 specifies principles and requirements for quantifying and reporting GHG emissions and removals at the organisation level. It is consistent with the GHG Protocol in spirit but uses its own structure. The organisation sets its organisational boundary by control or equity share. Instead of three scopes, emissions are reported in six categories: (1) direct GHG emissions and removals; (2) indirect emissions from imported energy; (3) indirect emissions from transportation; (4) indirect emissions from products used by the organisation; (5) indirect emissions associated with the use of products from the organisation; and (6) indirect emissions from other sources. The organisation must define and apply criteria to decide which indirect emissions are significant and must justify excluding any. It must also establish a base year, describe its quantification approaches, assess uncertainty and prepare a GHG inventory report. Biogenic CO2 and removals are reported separately.
The 2018 revision moved from three 'scopes' in the 2006 edition to the six categories above, which map approximately to Scope 1 (category 1), Scope 2 (category 2) and Scope 3 (categories 3 to 6).
ISO 14064-2:2019 covers GHG projects -- activities designed to reduce emissions or increase removals, such as a biogas digester on a pig farm or a reforestation scheme. The project proponent identifies the GHG sources, sinks and reservoirs relevant to the project, selects and justifies a baseline scenario (what would have happened without the project), quantifies emissions and removals in the project and baseline, monitors and reports. The standard is policy-neutral: it does not itself define additionality tests, crediting periods or eligibility; those are set by the GHG programme that issues credits.
A well-documented project plan under ISO 14064-2 also identifies which sources, sinks and reservoirs are controlled by the proponent, related to the project or affected by it -- the last group being where leakage arises, such as displaced grazing when pasture is planted with trees. It sets a monitoring plan stating parameters, frequency, equipment and responsibilities, and it applies conservativeness where assumptions are uncertain, so that reductions are not overstated. The resulting GHG statement can then be validated before the project starts and verified after each monitoring period under ISO 14064-3.
ISO 14064-3:2019 gives the specification with guidance for verification and validation of GHG statements -- for organisations, projects and products. Verification evaluates a statement about historical data; validation evaluates the reasonableness of assumptions, limitations and methods behind a statement about future activities, such as a project plan's projected reductions. The verifier agrees the level of assurance (limited or reasonable), objectives, criteria, scope and materiality with the client, plans the engagement with a risk assessment, gathers evidence, evaluates the statement and issues an opinion.
Related standards: ISO 14065:2020 sets requirements for bodies that validate and verify environmental information, building on ISO/IEC 17029, and is the basis for their accreditation; ISO 14066 sets competence requirements for validation and verification teams; ISO 14067:2018 covers the carbon footprint of products; ISO 14068-1:2023 covers carbon neutrality. Accreditation bodies assess verification bodies against ISO 14065, and GHG programmes such as emissions trading schemes add their own rules.
Data quality is judged on representativeness and reliability. The Scope 3 Standard lists five data quality indicators: technological representativeness (does the factor describe the actual technology or practice?), temporal representativeness (the right year?), geographical representativeness (the right place?), completeness (does the data set cover all relevant sites and processes?) and reliability (how were the data obtained and checked?). Primary data are collected from the specific activity in the value chain; secondary data come from databases, literature or industry averages. A dairy processor using its own milk suppliers' herd records has primary data; one using a national average footprint for milk has secondary data.
Practical controls include a data management plan, documented sources for every factor, version control, reconciliation of activity data to financial records (litres of fuel to fuel invoices, tonnes of fertiliser to purchase ledgers), and review by someone other than the preparer. These mirror IPCC QA/QC.
Uncertainty describes the range within which the true value is likely to lie, usually stated as a 95% confidence interval in percent of the estimate. It has two sources: parameter uncertainty in activity data and emission factors, and model or scientific uncertainty in the equations themselves. Uncertainty is distinct from error -- a transcription mistake is an error and should be corrected, not added to the uncertainty.
IPCC Approach 1 combines uncertainties by error propagation, valid for roughly independent quantities with moderate uncertainties. For a product (emissions = AD x EF), the percentage uncertainty is the square root of the sum of the squared percentage uncertainties: U = sqrt(U_AD^2 + U_EF^2). For a sum of sources, absolute uncertainties are combined: U_total = sqrt(sum of (U_i x x_i)^2) / |sum of x_i|. Approach 2 uses Monte Carlo simulation, which handles large, asymmetric or correlated uncertainties. Agricultural N2O typically has a large emission-factor uncertainty, so a farm total dominated by soil N2O carries a wide range even with precise fertiliser records.
Uncertainty matters for decisions: a reported reduction smaller than the uncertainty of the trend may not be real, and a removal claim with a wide uncertainty may need a conservative discount.
Materiality asks whether an error, omission or misrepresentation could change the decisions of the intended users. In verification, a materiality threshold is set -- often a percentage of the total -- and misstatements are judged individually and in aggregate against it, together with qualitative considerations (an error in a headline claim may be material even if small). Thresholds differ by programme and engagement; they are agreed in the verification plan, not chosen after the evidence is in. In reporting standards, 'materiality' also refers to which topics a company must report (financial or impact materiality); in GHG work the word usually means the verification sense.
A base year is the reference against which emissions are tracked over time. The GHG Protocol asks companies to choose the earliest year for which verifiable data exist, to explain the choice and to set a recalculation policy. A company may use a single fixed base year, an average of several years (useful where one year was unusual, as with drought on a farm) or, less commonly, a rolling base year. ISO 14064-1 also requires a base year and recalculation procedures.
For farms and agri-food companies, weather makes single years noisy: a dry year lowers yields and raises emissions per tonne, while a wet year changes soil N2O and rice methane. An average base year, such as the mean of three consecutive years, smooths this. Whatever is chosen, the base-year inventory must be held to the same standard as the current year: same boundary, same gases, documented factors and, ideally, verified data. Targets set against a poorly documented base year are hard to defend later, because every recalculation and every progress claim depends on it.
Base-year emissions are recalculated when the company's structure or methods change so that comparison would otherwise mislead. Triggers under the GHG Protocol are: structural changes -- mergers, acquisitions and divestments, and outsourcing or insourcing of emitting activities; changes in calculation methodology or improvements in the accuracy of emission factors or activity data; and discovery of significant errors or a number of cumulative errors that together are significant. Each company sets a significance threshold for recalculation and discloses it.
Base-year emissions are not recalculated for organic growth or decline -- changes in production output, product mix, or opening and closing of units owned or controlled by the company. They are also not recalculated for the acquisition or divestment of a facility that did not exist in the base year. When a facility that existed in the base year is acquired, its base-year emissions are added; when one is divested, its base-year emissions are removed. The aim is like-for-like comparison: reductions from selling a high-emitting plant are not real reductions for the atmosphere.
An absolute target is a reduction in total tonnes of CO2e by a target year relative to the base year. An intensity target is a reduction in emissions per unit of output (per tonne of product, per litre of milk, per unit of revenue). Absolute targets map directly onto atmospheric outcomes; intensity targets allow growth and suit commodity sectors, but absolute emissions can rise under them. A target should state its boundary (scopes, gases, sites), base year, target year, percentage reduction, whether removals and credits are excluded, and the recalculation policy.
Progress = (base-year emissions - current emissions) / base-year emissions, using the recalculated base year where recalculation applies. Annual linear reduction rates are applied to the base year: a 4.2% linear annual reduction over ten years means a 42% reduction, not compounding.
A target is 'science-based' when it is consistent with the reduction pathways needed to meet the Paris Agreement temperature goal. The Science Based Targets initiative (SBTi), founded by CDP, the UN Global Compact, WRI and WWF, sets criteria and validates company targets. Under the criteria in force at the time of writing, near-term targets cover 5 to 10 years from submission and must be aligned with 1.5 degrees C for Scopes 1 and 2; they must cover at least 95% of Scope 1 and 2 emissions; a Scope 3 target is required when Scope 3 is 40% or more of total Scope 1, 2 and 3 emissions, and near-term Scope 3 targets must cover at least two-thirds (67%) of Scope 3. Base years should be no earlier than 2015. Bought carbon credits cannot be counted towards reductions.
Two methods dominate. The absolute contraction approach requires all companies to cut absolute emissions at the same rate; the cross-sector 1.5 degrees C rate is a minimum 4.2% linear annual reduction. The sectoral decarbonisation approach sets intensity pathways for homogeneous sectors such as power or cement.
The SBTi Corporate Net-Zero Standard (version 1, 2021; a second version is under development) defines corporate net zero as reducing Scope 1, 2 and 3 emissions to a residual level consistent with 1.5 degrees C pathways -- for most sectors at least a 90% reduction by 2050 at the latest -- and neutralising the residual with permanent carbon removals. Long-term targets must cover at least 90% of Scope 3. Companies are encouraged to invest in mitigation beyond their value chain (BVCM) in the meantime, but this does not count towards their targets. ISO 14068-1:2023 deals with carbon neutrality claims, which can include credits, and must not be confused with net zero. Other net-zero criteria exist, such as the UN High-Level Expert Group's recommendations on net-zero pledges by non-state entities (2022), which stress no new fossil supply, integrity of credits and transparent transition plans.
The SBTi Forest, Land and Agriculture (FLAG) Guidance (2022) applies to land-intensive companies. A company must set a separate FLAG target if it belongs to a FLAG-designated sector (such as food production, food and beverage processing, food and staples retailing, forest and paper products, or tobacco) or if its FLAG-related emissions are 20% or more of its total emissions across scopes. FLAG targets cover land-use change, land-management emissions and land-based removals; unlike energy and industry targets, removals within the value chain can count towards FLAG targets if they meet the guidance's requirements. Companies setting FLAG targets must commit to no deforestation across their primary deforestation-linked commodities, with a target date no later than 31 December 2025.
The FLAG sector pathway requires an absolute reduction of at least 3.03% per year (linear) for near-term targets, and long-term FLAG targets require at least a 72% reduction by 2050. Commodity pathways (for beef, dairy, poultry, pork, rice, soy, palm oil, wheat, maize and others) give intensity alternatives. These numbers belong to the 2022 guidance and may be revised; quote them with the version.
Assurance is an independent practitioner's conclusion about whether a GHG statement is free of material misstatement against stated criteria. Reasonable assurance is a high but not absolute level; the practitioner gathers enough evidence to give a positive-form opinion ('in our opinion, the statement is prepared, in all material respects, in accordance with...'). Limited assurance involves fewer and less extensive procedures, mainly inquiry and analytical review, and gives a negative-form conclusion ('nothing has come to our attention that causes us to believe the statement is materially misstated'). Many disclosure regimes begin with limited assurance and move towards reasonable assurance over time.
Two families of standards are used. Accountants and other assurance practitioners use the International Auditing and Assurance Standards Board's ISAE 3410 (Assurance Engagements on Greenhouse Gas Statements), under ISAE 3000, and the newer ISSA 5000 for sustainability information generally (issued in 2024). Environmental verification bodies use ISO 14064-3, with accreditation to ISO 14065. Emissions trading schemes have their own verification regulations, which set programme-specific materiality thresholds and site-visit rules.
A verification follows a predictable path. Pre-engagement: independence and competence checks, agreement on the level of assurance, criteria (such as the GHG Protocol or ISO 14064-1), scope, boundary, period and materiality. Planning: understanding the organisation and its data flows, a strategic analysis and risk assessment that identifies where misstatement is likely, and a sampling plan concentrated on high-risk, high-emission sources. Evidence gathering: site visits, tracing reported figures back to source records (fuel invoices, meter readings, herd registers, fertiliser purchase records), recalculation, checking emission factors and GWP values against the stated sources, and testing controls. Evaluation: misstatements are listed, the client is asked to correct them, and uncorrected misstatements are aggregated and compared with materiality. Reporting: the opinion or conclusion, with any qualification, and a management letter of findings.
In agriculture the high-risk areas are typically livestock numbers and classes, feed and productivity assumptions behind Tier 2 factors, fertiliser nitrogen and manure applications, land-use change history and any claimed soil-carbon removal. A verifier will ask whether factors match the region and guideline version stated, and whether removals and emissions have been netted against the rules.
An unmodified opinion states that the statement is fairly presented. A qualified opinion states that it is fairly presented except for a specified matter that is material but not pervasive. An adverse opinion states that the statement is materially misstated. A disclaimer is given when the practitioner cannot obtain sufficient evidence to form a conclusion. Independence is essential: a body that prepared the inventory, or designed the client's calculation system, should not verify it. Assurance also does not make a number exact -- it tests that the reported number follows the criteria within materiality, which is why the criteria, boundary and level of assurance should always be read along with the opinion.
Learn the gas basket, the GWP sets and which one a question asks for, and the 44/12 and 44/28 conversions until they are automatic. Be able to classify any emission into Scope 1, 2 or 3 and into the right Scope 3 category, and to explain equity share versus control in one sentence. Practise Tier 1 arithmetic for enteric methane, soil N2O, urea and lime, and the two uncertainty-propagation rules. Know the five volumes of the 2006 IPCC Guidelines, the six land categories, the five carbon pools and the three ISO 14064 parts. For interviews, be ready to explain why spend-based Scope 3 figures cannot show a farmer's practice change and why removals are reported separately from emissions.
As a carbon analyst at a food company, consultancy or farm advisory service, you build and maintain inventories: set the boundary, collect activity data from sites and suppliers, choose factors, document every source, and prepare for verification. Typical decisions are which consolidation approach to use for joint ventures, how to move the largest Scope 3 categories from spend-based to supplier data, which calculator to use for a supplier programme and how to keep it consistent, when a base-year recalculation is triggered, and how to answer a verifier's request for evidence. You also translate the numbers into reduction plans with operations and procurement teams.
As head of sustainability, lead verifier or senior adviser, you own the targets and the claims. You decide whether to set an SBTi target including FLAG, what share of the footprint can be reduced with suppliers, how to treat land-use change and removals in public statements, and whether the company is ready for reasonable assurance. You manage greenwashing and legal risk -- ensuring offsets are not presented as reductions and intensity gains are not sold as absolute cuts -- and you brief boards, auditors and investors on what the numbers can and cannot support. You also decide how much to invest in primary data versus buying better secondary data.
As a researcher, IPCC author or standard-setter, you work on the weak points of the system: emission factors for tropical livestock and smallholder systems, measurement-based Tier 3 methods, soil-carbon monitoring and verification, metrics for short-lived gases, allocation in mixed farming, and the consistency between national inventories, company inventories and carbon markets. You design the studies that turn default factors into country-specific ones, review standards in public consultation, supervise students who will build national systems, and write the exam questions and course material that train the next cohort.
This manual describes the GHG Protocol standards (2004, 2011, 2015), ISO 14064 (2018-2019) and the SBTi Net-Zero Standard v1 and FLAG Guidance (2022) as published; all are being revised, and the GHG Protocol land-sector and removals rules were still being finalised from the 2022 draft -- check the current versions. It does not tabulate IPCC default emission factors, which differ by region, livestock productivity, climate and guideline version; look them up in the tables that apply to you. It does not cover emissions-trading-scheme monitoring rules country by country, financed-emissions methods for banks (PCAF) in depth, product category rules for individual foods, or carbon-credit methodologies, which belong to the carbon-markets subject. GWP20, GTP and GWP* are introduced but not developed.
Original practice questions written to the syllabus and pattern of the exams named in each question's examTags. They are not past papers and do not reproduce any board's questions. Every numerical answer was computed by script; GWP sets, emission factors and fractions are stated in each stem because they differ between IPCC reports, guideline versions and regions.
These notes are free to read here, with no account.
What the Laisha app adds:
Free for ten days, then a subscription. We would rather write that here than have you find it later.