Editing IPCC:AR6/WGIII/Chapter-12 (section)

=== 12.4.2 GHG Emissions from Food Systems ===

<div id="h2-14-siblings" class="h2-siblings"></div>

<div id="12.4.2.1" class="h3-container"></div>

<span id="sectoral-contribution-of-ghg-emissions-from-food-systems"></span>
==== 12.4.2.1 Sectoral Contribution of GHG Emissions from Food Systems ====

<div id="h3-5-siblings" class="h3-siblings"></div>

New calculations using the EDGAR v6.0 ( [[#Crippa--2021a|Crippa et al. 2021a]] ) and FAOSTAT ( [[#FAO--2021|FAO 2021]] ) databases provide territorial-based food system GHG emissions by country globally for the period 1990 to 2018 ( [[#Crippa--2021b|Crippa et al. 2021b]] ). The data are calculated based on a combination of country-specific data and aggregated information as described by [[#Crippa--2021b|Crippa et al. (2021b)]] and [[#Tubiello--2021|Tubiello et al. (2021)]] . The data show that, in 2018, 17 GtCO 2 -eq yr –1 (95% confidence range 13–23 GtCO 2 -eq yr –1 , calculated according to [[#Solazzo--2020|Solazzo et al. (2020)]] ) were associated with the production, processing, distribution, consumption of food and management of food system residues. This corresponded to 31% (range 23–42%) of total anthropogenic GHG emissions of 54 GtCO 2 -eq yr –1 . Based on the IPCC sectoral classification (Table 12.7 and Figure 12.5), the largest contribution of food systems GHG emissions in 2018 was from agriculture, that is, livestock and crop production systems (6.3 GtCO 2 -eq yr –1 , range 2.6–11.9) and land use, land use change and forestry (LULUCF) (4.0 GtCO 2 -eq yr –1 , range 2.1–5.9) (Figure 12.5). Emissions from energy use were 3.9 GtCO 2 -eq yr –1 (3.6–4.4 ''')''' , waste management 1.7 GtCO 2 -eq yr –1 (0.9–2.6), and industrial processes and product use 0.9 GtCO 2 -eq yr –1 (0.6–1.1). The share of GHG emissions from food systems generated outside the AFOLU (agriculture and LULUCF) sectors has increased over recent decades, from 28% in 1990 to 39% in 2018.

'''Table 12.7''' '''| GHG emissions from food systems by sector according to IPCC classification in Mt gas y''' '''r''' –1 '''and food systems’ share of total anthropogenic GHG emissions in''' '''1990 and 2015.'''

{| class="wikitable"
|-
! '''Sector'''

! '''CO''' 2

! '''CH''' 4

! '''N''' 2 '''O'''

! '''F-gases'''

! '''GHG'''

! '''CO''' 2

! '''CH''' 4

! '''N''' 2 '''O'''

! '''F-gases'''

! '''GHG'''

|-
!
! colspan="5"| '''Emissions (Mt gas y''' '''r''' –1 ''')'''

! colspan="5"| '''Share of total sectoral emissions (%)'''

|-
|
| colspan="10"| '''1990'''

|-
| 1 Energy

| 2212

| 10

| 0

| –

| 2583

| 10.5

| 10.2

| 26.7

| –

| 10.7

|-
| 2 Industrial processes

| 190

| 0

| 0

| 0

| 263

| 14.5

| 0

| 38

| 4.8

| 16.2

|-
| 3 Solvent and Other Product Use

| 0

| –

| –

| –

| 0

| 0.2

| –

| –

| –

| 0.2

|-
| 4 Agriculture

| 102

| 142

| 5

| –

| 5370

| 100

| 100

| 99.2

| –

| 99.8

|-
| 5 LULUCF

| 4946

| –

| 0

| –

| 5080

| 181

| –

| 194

| –

| 182

|-
| 6 Waste

| 3

| 40

| 0

| –

| 1155

| 29

| 72.4

| 99.1

| –

| 73.2

|-
| '''Total'''

| '''7453'''

| '''192'''

| '''6'''

| '''0'''

| '''14452'''

| '''29.3'''

| '''65.2'''

| '''84.5'''

| '''4.8'''

| '''40.3'''

|-
| '''Total (MtCO''' 2 '''-eq y''' '''r''' –1 ''')'''

| '''7453'''

| '''5243'''

| '''1755'''

| '''0'''

| '''14452'''

| '''29.3'''

| '''63.9'''

| '''84.5'''

| '''0.3'''

| '''40.3'''

|-
|
| colspan="10"| '''2015'''

|-
| 1 Energy

| 3449

| 13

| 0

| –

| 3927

| 10.1

| 9.5

| 24.1

| –

| 10.2

|-
| 2 Industrial processes

| 242

| 0

| 0

| 0

| 881

| 7.9

| 0

| 28.6

| 58

| 20.1

|-
| 3 Solvent and Other Product Use

| 7

| –

| –

| –

| 7

| 4.1

| –

| –

| –

| 3.6

|-
| 4 Agriculture

| 140

| 161

| 7

| –

| 6326

| 100

| 100

| 99.1

| –

| 99.7

|-
| 5 LULUCF

| 3823

| –

| 1

| –

| 3982

| 190

| –

| 229

| –

| 191

|-
| 6 Waste

| 5

| 58

| 0

| –

| 1699

| 30.6

| 71.8

| 99.1

| –

| 72.9

|-
| '''Total'''

| '''7666'''

| '''231'''

| '''8'''

| '''0'''

| '''16821'''

| '''19.3'''

| '''61.6'''

| '''83.7'''

| '''58'''

| '''31.1'''

|-
| '''Total (MtCO''' 2 '''-eq y''' '''r''' – 1 ''')'''

| '''7666'''

| '''6317'''

| '''2256'''

| '''581'''

| '''16821'''

| '''19.3'''

| '''60.2'''

| '''83.7'''

| '''53.6'''

| '''31.1'''

|}

Notes: Agricultural emissions include the emissions from the whole sector; biomass production for non-food use currently not differentiated. Non-food system AFOLU emissions are negative (that is, a net carbon sink), therefore the share of AFOLU food system emissions is &gt;100. Source: EDGARv6 ( [[#Crippa--2019|Crippa et al. 2019]] ; [[#Crippa--2021b|Crippa et al. 2021b]] ), and FAOSTAT ( [[#FAO--2021|FAO 2021]] ). LULUCF: land use, land-use change and forestry.

<div id="_idContainer105" class="_idGenObjectStyleOverride-1"></div>

[[File:bcbd3fbc20b5d7f8776af6a4ab091011 IPCC_AR6_WGIII_Figure_12_5.png]]

'''Figure 12.5 | Food system GHG emissions from the agriculture, LULUCF, waste, and energy &amp; industry sectors.''' Source: [[#Crippa--2021b|Crippa et al. (2021b)]] .

'''Energy:''' Emissions from energy use occur throughout the food supply chain. In 2018, the main contributions came from energy industries supplying electricity and heat (970 MtCO 2 -eq yr –1 ), manufacturing and construction (920 MtCO 2 -eq yr –1 , of which 29% was attributable to the food, beverage, and tobacco industry), and transport (760 MtCO 2 -eq yr –1 ). These emissions were almost entirely as CO 2 . Energy emissions from forestry and fisheries amounted to 480 MtCO 2 -eq yr –1 , with 91% of emissions as CO 2 . Emissions from residential and commercial fuel combustion contributed 250 MtCO 2 -eq yr –1 (79% of emissions as CO 2 , and with emissions of 1.7 MtCH 4 yr –1 ) and 130 MtCO 2 -eq yr –1 (with 98% of emissions as CO 2 ), respectively.

Refrigeration uses an estimated 43% of energy in the retail sector ( [[#Behfar--2018|Behfar et al. 2018]] ) and significantly increases fuel consumption during distribution. Besides being energy intensive, supermarket refrigeration also contributes to GHG emissions through leakage of refrigerants (fluorinated gases, or F-gases), although their contribution to food system GHG emissions is estimated to be minor ( [[#Crippa--2021b|Crippa et al. 2021b]] ). The cold chain accounts for approximately 1% of global GHG emissions, but as the volume of refrigerators per capita in developing countries is reported to be one order of magnitude lower than in developed countries (19 m 3 versus 200 m 3 refrigerated storage capacity per 1000 inhabitants), the importance of refrigeration to total GHG emissions is expected to increase ( [[#James--2010|James and James 2010]] ). Although refrigeration gives rise to GHG emissions, both household refrigeration and effective cold chains could contribute to a substantial reduction in losses of perishable food and thus in emissions associated with food provision ( [[#University%20of%20Birmingham--2018|University of Birmingham 2018]] ; [[#James--2010|James and James 2010]] ). A trade-off exists between reducing food waste and increased refrigeration emissions, with the benefits depending on type of produce, location and technologies used (Sustainable Cooling for All 2018; [[#Wu--2019|Wu et al. 2019]] ).

Transport has overall a minor importance for food system GHG emissions, with a share of 5% to 6% (Poore and Nemecek 2018; [[#Crippa--2021b|Crippa et al. 2021b]] ). The largest contributor to food system transport GHG emissions was road transport (92%), followed by marine shipping (4%), rail (3%), and aviation (1%). Only looking at energy needs, air or road transport consumes one order of magnitude higher energy (road: 70–80 MJ t –1 km –1 ; aviation: 100–200 MJ t –1 km –1 ) than marine shipping (10–20 MJ t –1 km –1 ) or rail (8–10 MJ t –1 km –1 ) ( [[#FAO--2011|FAO 2011]] ). For specific food products with high water content, relatively low agricultural emissions and high average transport distances, the share of transport in total GHG emissions can be over 40% (e.g., bananas, with total global average GHG emissions of 0.7 kgCO 2 -eq kg –1 ) (Poore and Nemecek 2018), but transport is a minor source of GHG emissions for most food products (Poore and Nemecek 2018).

'''Industry:''' Direct industrial emissions associated with food systems are generated by the refrigerants industry (580 MtCO 2 -eq yr –1 as F-gases) and the fertiliser industry for ammonia production (280 MtCO 2 -eq yr –1 as CO 2 ) and nitric acid (60 MtCO 2 -eq yr –1 as N 2 O). The industry sector data account for CO 2 stored in urea (–50 MtCO 2 -eq yr –1 ). Packaging contributed about 6% of total food system emissions (0.98 GtCO 2 -eq yr –1 , 91% as CO 2 , with CH 4 emissions of 2.8 Mt CH 4 yr –1 ). Major emissions sources are pulp and paper (60 MtCO 2 -eq yr –1 ) and aluminium (30 MtCO 2 -eq yr –1 ), with ferrous metals, glass, and plastics making a smaller contribution. High shares of emissions from packaging are found for beverages and some fruit and vegetables (Poore and Nemecek 2018).

'''Waste:''' Management of waste generated in the food system (including food waste, wastewater, packaging waste, etc.) leads to biogenic GHG emissions, and contributed 1.7 GtCO 2 -eq yr –1 to food systems’ GHG emissions in 2018. Of these emissions, 55% were from domestic and commercial wastewater (30 MtCH 4 yr –1 and 310 ktN 2 O yr –1 ), 36% from solid waste management (20 MtCH 4 yr –1 and 310 ktN 2 O yr –1 ), and 8% from industrial wastewater (4 MtCH 4 yr –1 and 80 ktN 2 O yr –1 ). Emissions from waste incineration and other waste management systems contributed 1%.

<div id="12.4.2.2" class="h3-container"></div>

<span id="ghg-intensities-of-food-commodities"></span>
==== 12.4.2.2 GHG Intensities of Food Commodities ====

<div id="h3-6-siblings" class="h3-siblings"></div>

There is high variability in the GHG emissions of different food products and production systems (Figure 12.6). GHG emissions intensities – measured using attributional lifecycle assessment, considering the full supply chain, expressed as CO 2 -eq per kg of product or per kg of protein – are generally highest for ruminant meat, cheese, and certain crustacean species (e.g., farmed shrimp and prawns, trawled lobster) ( [[#Nijdam--2012|Nijdam et al. 2012]] ; [[#Clark--2017|Clark and Tilman 2017]] ; [[#Clune--2017|Clune et al. 2017]] ; [[#Hilborn--2018|Hilborn et al. 2018]] ; Poore and Nemecek 2018) ( ''robust evidence, high agreement'' ) ''.'' Generally, beef from dairy systems has a lower footprint (8–23 kgCO 2 -eq per 100 g protein than beef from beef herds (17–94 kgCO 2 -eq per 100 g protein (Figure 12.6, re-calculated from Poore and Nemecek (2018) using AR6 GWPs based on a 100year horizon) ( ''medium evidence'' , ''high agreement'' ). The wide variation in emissions from beef reflects differences in production systems, which range from intensive feedlots with stock raised largely on grains through to rangeland and transhumance production systems. Dairy systems are generally more intensive production systems, with higher digestibility feed than beef systems. Further, emissions from dairy systems are shared between milk and meat, which brings GHG footprints of beef from dairy herds closer to those of meat from monogastric animals, with emissions intensities of pork (4.4–13 kgCO 2 -eq per 100 g protein) and poultry meat (2.3–11 kgCO 2 -eq per 100 g protein) (Poore and Nemecek 2018).

<div id="_idContainer107" class="_idGenObjectStyleOverride-1"></div>

[[File:67e8f8e10f0276f12a17395f54ec004f IPCC_AR6_WGIII_Figure_12_6.png]]

'''Figure 12.6 | Ranges of GHG intensities [kgCO''' 2 '''-eq per 100 g protein,''' '''10–90''' th '''percentile] in protein-rich foods, quantified via a meta-analysis of attributional lifecycle assessment studies using economic allocation.''' Aggregation of CO 2 , CH 4 , and N 2 O emissions in Poore and Nemecek (2018) updated to use AR6 100-year GWP. Data for capture fish, crustaceans, and cephalopods from [[#Parker--2018|Parker et al. (2018)]] , with post-farm data from Poore and Nemecek (2018), where the ranges represent differences across species groups. CH 4 emissions include emissions from manure management, enteric fermentation, and flooded rice only. a Grains are not generally classed as protein-rich, but they provide about 41% of global protein intake. Here grains are a weighted average of wheat, maize, oats, and rice by global protein intake. b Conversion of annual to perennial crops can lead to carbon sequestration in woody biomass and soil, shown as negative emissions intensity. Source: data from Poore and Nemecek (2018); [[#Parker--2018|Parker et al. (2018)]] .

Emissions intensities for farmed fish ranged from 2.4–11 kgCO 2 -eq per 100 g protein (Poore and Nemecek 2018). For Norwegian seafood, large differences have been found ranging from 1.1 kgCO 2 -eq kg –1 edible product for herring to more than 8 kgCO 2 -eq kg –1 edible product for salmon shipped by road and ferry from Oslo to Paris ( [[#Winther--2020|Winther et al. 2020]] ). For capture fish, large differences in emissions have been found, ranging from 0.2–7.9 kgCO 2 -eq kg –1 landed fish ( [[#Parker--2018|Parker et al. 2018]] ), although an environmental comparison of capture fish to farmed foods should include other indicators such as overfishing. Plant-based foods generally have lower GHG emissions (–2.2 to +4.5 kgCO 2 -eq per 100 g protein) than farmed animal-based foods ( [[#Nijdam--2012|Nijdam et al. 2012]] ; [[#Clark--2017|Clark and Tilman 2017]] ; [[#Clune--2017|Clune et al. 2017]] ; [[#Hilborn--2018|Hilborn et al. 2018]] ; Poore and Nemecek 2018) ( ''robust evidence, high agreement'' ). Several plant-based foods are associated with emissions from land use change, for example, palm oil, soy and coffee (Poore and Nemecek 2018), although emissions intensities are context specific ( [[#Meijaard--2020|Meijaard et al. 2020]] ) and for plant-based proteins, GHG footprints per serving remain lower than those of animal source proteins ( [[#Kim--2019|Kim et al. 2019]] ) ''.''

In traditional production systems, especially in developing countries, livestock serve multiple functions, providing draught power, fertiliser, investment and social status, besides constituting an important source of nutrients ( [[#Weiler--2014|Weiler et al. 2014]] ). In landscapes dominated by forests or cropland, semi-natural pastures grazed by ruminants provide heterogeneity that supports biodiversity ( [[#Röös--2016|Röös et al. 2016]] ). Grazing on marginal land and the use of crop residues and food waste can provide human-edible food with lower demands for cropland ( [[#Röös--2016|Röös et al. 2016]] ; [[#Van%20Zanten--2018|Van Zanten et al. 2018]] ; Van Hal et al. 2019). Animal protein requires more land than vegetable protein, so switching consumption from animal to vegetable proteins could reduce the pressure on land resources and potentially enable additional mitigation through expansion of natural ecosystems, storing carbon while supporting biodiversity, or reforestation to sequester carbon and enhance wood supply capacity for the production of bio-based products substituting fossil fuels, plastics, cement, etc. ( [[#Schmidinger--2012|Schmidinger and Stehfest 2012]] ; [[#Searchinger--2018b|Searchinger et al. 2018b]] ; [[#Hayek--2021|Hayek et al. 2021]] ). At the same time, alternatives to animal-based meat and other livestock products are being developed (Figure 12.6). Their increasing visibility in supermarkets and catering services, as well as falling production prices, could make meat substitutes competitive in one to two decades ( [[#Gerhardt--2019|Gerhardt et al. 2019]] ). However, uncertainty around their uptake creates uncertainty around their effect on future GHG emissions.

<div id="12.4.2.3" class="h3-container"></div>

<span id="territorial-national-per-capita-ghg-emissions-from-food-systems"></span>
==== 12.4.2.3 Territorial National Per Capita GHG Emissions from Food Systems ====

<div id="h3-7-siblings" class="h3-siblings"></div>

Food systems are connected to other societal systems, such as the energy system, financial system, and transport system ( [[#Leip--2021|Leip et al. 2021]] ). Also, food systems are dynamic and continuously changing and adapting to existing and anticipated future conditions. Food production systems are very diverse and vary by farm size, intensity level, farm specialisation, technological level, production methods (e.g., organic, conventional, etc.), with differing environmental and social consequences ( [[#Václavík--2013|Václavík et al. 2013]] ; [[#Fanzo--2017|Fanzo 2017]] ; [[#Herrero--2017|Herrero et al. 2017]] ; [[#Herrero--2021|Herrero et al. 2021]] ).

Various frameworks have been proposed to assess sustainability of food systems, including metrics and indicators on environmental, health, economic and equity issues, pointing to the importance of recognising the multi-dimensionality of food system outcomes ( [[#Gustafson--2016|Gustafson et al. 2016]] ; [[#Chaudhary--2018|Chaudhary et al. 2018]] ; [[#Hallström--2018|Hallström et al. 2018]] ; [[#Zurek--2018|Zurek et al. 2018]] ; [[#Eme--2019|Eme et al. 2019]] ; [[#Béné--2020|Béné et al. 2020]] ; [[#Hebinck--2021|Hebinck et al. 2021]] ). Data platforms are being developed, but so far comprehensive data for evidence-based food system policy are lacking ( [[#Fanzo--2020|Fanzo et al. 2020]] ).

To visualise several food systems dimensions in a GHG context, Figure 12.7 shows GHG emissions per capita and year for regional country aggregates ( [[#Crippa--2021a|Crippa et al. 2021a]] ; [[#Crippa--2021b|Crippa et al. 2021b]] ), indicated by the size of the bubbles. The GHG emissions presented here are based on territorial accounting similar to the UNFCCC GHG inventories: emissions are assigned to the country where they occur, not where food is consumed ( [[#Crippa--2021a|Crippa et al. 2021a]] ; [[#Crippa--2021b|Crippa et al. 2021b]] ) ( [[#12.4.2.1|Section 12.4.2.1]] ). The colours of the bubbles indicate the relative contribution of the following risk factors to deaths, according to the classification used in the Global Burden of Disease Study: child and maternal malnutrition (red, deficiencies of iron, zinc or Vitamin A, or low birth weight or child growth failure), dietary risks (yellow, for example diets low in vegetables, legumes, whole grains or diets high in red and processed meat and sugar-sweetened beverages) or high body mass index (blue). The combined contribution of these three risk factors to total deaths varies strongly and is between 28% and 88% of total deaths. Figure 12.7 shows that dietary risk factors are prevalent throughout all regions. Though not a complete measure of the health impact of food, these were selected as a proxy for nutritional adequacy and balance of diets, avoidance of food insecurity, over- or mal-nutrition and associated non-communicable diseases ( [[#GBD%202017%20Diet%20Collaborators--2018|GBD 2017 Diet Collaborators 2018]] ; GBD 2017 Diet Collaborators et al. 2019).

<div id="_idContainer109" class="_idGenObjectStyleOverride-1"></div>

[[File:f5aff741bb98c5989996a74bd340024c IPCC_AR6_WGIII_Figure_12_7.png]]

'''Figure 12.7 | Regional differences in health outcomes, territorial per capita GHG emissions from national food systems, and share of food system GHG emissions from energy use.''' GHG emissions are calculated according to the IPCC Tier 1 approach and are assigned to the country where they occur, not necessarily where the food is consumed. Health outcome is expressed as relative contribution of each of the following risk factors to their combined risk for deaths: child and maternal malnutrition (red), dietary risks (yellow) or high body mass index (blue). Sources: wholesale cost of food per capita: [[#Springmann--2021|Springmann et al. 2021]] ); territorial food system GHG emissions: EDGAR v.6, [[#Crippa--2021a|Crippa et al. (2021a)]] , recalculated according to [[#Crippa--2021b|Crippa et al. (2021b)]] using AR6 GWPs; deaths attributed to dietary factors: [[#IHME--2018|IHME (2018)]] ; GBD 2017 Diet Collaborators et al. (2019).

The share of GHG emissions from energy use is taken as a proxy for the structure of food supply in a region ( [[#12.4.1|Section 12.4.1]] ), and the cost for food as a proxy for the structure of the demand side and the access to (healthy) food ( [[#Chen--2016|Chen et al. 2016]] ; [[#Finaret--2019|Finaret and Masters 2019]] ; [[#Hirvonen--2019|Hirvonen et al. 2019]] ; [[#HLPE--2020|HLPE 2020]] ; [[#Springmann--2021|Springmann et al. 2021]] ), though acknowledging the limitations of such a simplification.

While total food system emissions in 2018 range between 0.9 and 8.5 tCO 2 -eq per capita per year between regions, the share of energy emissions relative to energy and land-based (agriculture and food system land-use change) emissions ranges between 3% and 78%. Regional expenditures for food range from USD3.0–8.8 per capita per day (Figure 12.7), though there is high variability within countries and the costs of nutrient-adequate diets often exceeds those of diets delivering adequate energy ( [[#Hirvonen--2019|Hirvonen et al. 2019]] ; Bai et al. 2020; [[#FAO--2020|FAO et al. 2020]] ). Thus, low-income households in industrialised countries can also be affected by food insecurity ( [[#Penne--2020|Penne and Goedemé 2020]] ).

<div id="12.4.3" class="h2-container"></div>

<span id="mitigation-opportunities"></span>