Tag: emissions projections

Posted on

What is the size of the mitigation potential in EU agriculture by 2030?

In my previous post, I discussed the challenges of reducing non-CO2 greenhouse gas (GHG) emissions from agriculture and identified some of the strategies that are available or under development to allow farmers to reduce these emissions. But by how much would these strategies reduce projected emissions? What is the potential magnitude of the emissions reduction we should expect from agriculture in the coming decade? As in the previous post, I deliberately exclude a discussion of the potential to offset these emissions through land management and land use change although, as we will see, some insights into the potential to reduce emissions in the LULUCF sector will be covered in this post.

What I do in this post is to summarise the Commission’s quantification of the mitigation potential in agriculture up to 2030 in the context of the ‘Fit for 55’ package to be launched in June 2021. This is expected to propose significant changes in a raft of EU energy and climate laws to achieve the more ambitious at least net 55% emissions reduction target proposed for 2030. What contribution can we expect from reduced agricultural emissions towards the revised ‘at least a net 55% reduction’ 2030 target?

The post is an update to the excellent review by Allen and Maréchal ‘Agricultural GHG emissions: Determining the potential contribution to the Effort Sharing Regulation’ published by IEEP in 2017. Additional results are now available from the in-depth analysis of the ‘Clean Planet for All’ Communication in 2018, the impact assessment of the 2030 Climate Target Plan in 2020, as well as updated results from the Joint Research Centre ‘Economic assessment of GHG mitigation policy options for EU agriculture’ (EcAMPA) project. These are the results reported in this post.

Emissions trends under ‘business as usual’

First, a reminder of what the Commission expects to happen to agricultural emissions if no additional effort is made to reduce emissions, the baseline or ‘business as usual’ reference scenario. Based on the annual EEA inventories, agricultural emissions in the EU27 fell by 21% between 1990 and 2018, from 497MtCO2e to 394 MtCO2e (converting the non-CO2 gases to CO2e using AR4 weights).

The most plausible scenario to 2030 has been developed under the EcAMPA 4 project, the latest iteration of the EcAMPA initiative by researchers at the Joint Research Centre. This was published as a special chapter in the DG AGRI Market Outlook 2020-2030 publication last December. The EcAMPA project is based on the CAPRI model with mitigation technology options sourced from the GAINS model. Agricultural emissions are projected to fall to 388 MtCO2e by 2030 or by a further 1.6% compared to 2018. This reference scenario would bring the total reduction in agricultural emissions between 1990 and 2030 to 22%. This projection assumes a continuation of current CAP policies and no additional measures for GHG mitigation.

An alternative scenario using the GAINS model alone was published by a different group in the Commission in the impact assessment of its 2030 Climate Target Plan ‘Stepping up Europe’s 2030 Climate Ambition’ published in 2020. This analysis compares projected emissions in 2030 to emissions in 2005 and 2015. It projects a somewhat steeper decline in emissions by 2030 compared to EcAMPA 4 (a decline of 7% compared to 2015 levels). The 2020 impact assessment is conducted in terms of AR5 weights for non-CO2 gases, while the EcAMPA analysis uses AR4 weights. In the following analysis, I also make use of its figures for mitigation potential.

Mitigation potential from further measures targeting non-CO2 gases

There are various measures of mitigation potential, including technical, economic and market potentials. The technical mitigation potential measures the maximum mitigation possible with full implementation of all available mitigation measures and ignoring barriers to adoption. Economic potential includes only those measures that are cost-effective at a given carbon price. The market potential considers social, cultural, farm-level and political constraints to adoption that restrict the use of mitigation measures, often linked to concerns about adverse effects on farm income.

The Commission’s 2018 in-depth analysis of the ‘Clean Planet for All’ Communication took 2050 as the target date rather than 2030 which is the main focus of this post. However, for completeness I include it here. In the baseline, the impact assessment projected agricultural emissions to remain broadly stable until 2050 at 2010 levels in the absence of further mitigation incentives or changes in amount and type of agriculture goods produced. At a carbon price of €200/tonne CO2e, it projected an economic mitigation potential of around 130 MtCO2e in 2050. With baseline agricultural emissions remaining at around 400 MtCO2e in 2050, this represents a mitigation potential of 33% compared to the baseline.

We observed in the previous post that the Commission’s modelling for the impact assessment of the 2030 Climate Target Plan calculated that the potential for mitigating agricultural non-CO2 emissions ranged from 12.3 MtCO2e to 30.6 MtCO2e at CO2e prices varying between €0 and €55 per tonne CO2e. These reductions can be compared to projected baseline 2030 emissions in that Commission report of 375 MtCO2e (measured using AR5 weights to convert non-CO2 gases to CO2 equivalents) (figures taken from Tables 42 and 44 in the impact assessment part 2/2). They correspond to reductions of between 3% and 8% of baseline emissions in 2030.

Further reductions are expected arising from implementation of targets stemming from the Biodiversity and Farm to Fork Strategies in Member States’ CAP Strategic Plans. In an as yet undisclosed report (it is apparently still undergoing review, but the results are reported on page 90 of the impact assessment 2/2), the Joint Research Centre modelled the impact of the Commission’s proposals and projected a 17.4% reduction of non-CO2 GHG emissions in the agricultural sector by 2030, going up to 19.0% with the additional budget made available under the “Next Generation EU”. One assumes that these figures have been derived from the CAPRI model in the EcAMPA project but as the study has not yet been published, it is not clear how this potential was measured – what mitigation options were considered, whether it assumes full implementation of the quantitative targets included in these Strategies regardless of cost, whether each Member State is assumed to meet the targets separately, or whether the Strategies were implemented in some other way.

Another set of results reported in the DG AGRI Market Outlook publication is based on a previous iteration of the EcAMPA project, the EcAMPA 3 study by the Joint Research Centre (Pérez Domínguez et al, 2020). It reports that adoption of mitigation practices and technologies incentivised by pricing non-CO2 emissions has the potential to lead to a reduction of between 15 and 35 MtCO2e in agricultural emissions where the CO2e price varies between €20 and €100 per tonne CO2e. These figures are similar to those estimated in the 2030 Climate Target Plan impact assessment, though allowance must be made for differences in the metrics used.

The EcAMPA 3 iteration incorporated a carbon cycle model into the underlying CAPRI model used for simulations. This allows the model to report changes in soil carbon stocks resulting from changes in land management and land use to be calculated as well as reductions in non-CO2 gases. DG AGRI reports that the EcAMPA 3 iteration estimated that, at the same carbon prices (between €20 and €100/t CO2e), there would be a further reduction in LULUCF emissions of between 37 and 45 MtCO2, making a total mitigation potential of up to 80MtCO2e.

It is important to underline that this LULUCF mitigation arises solely as a by-product of targeting non-CO2 gases only in the model. The CO2 savings in the LULUCF sector are a secondary effect resulting from imposing a carbon price on non-CO2 emissions only. They should not be taken to represent the potential LULUCF abatement if, for example, a carbon farming scheme were introduced that in addition put a price on CO2.

These reductions refer to the direct impacts of implementing these practices. Because these practices are costly, they will induce further changes in production levels and production mix as farmers reassess their production decisions in the light of the new cost structures. The DG AGRI Market Outlook observes that, when changes in production intensity and production mix following the imposition of a carbon price on non-CO2 emissions are included, these emissions savings are amplified, by a further 40% at the lower carbon price (€20/tonne) and by a whopping 180% at a carbon price of €100/tonne. At the higher carbon price, the mitigation potential in agriculture including LULUCF reductions could be as high as 224 MtCO2e (i.e., 80 Mt CO2e * 2.8).

To understand where these figures come from, it is helpful to take a closer look at the EcAMPA 3 study.

A closer look at the EcAMPA 3 study

What are the mitigation practices that would give rise to these abatement potentials? The following table shows the technologies included in the EcAMPA 3 study and their mitigation potential under various assumptions. All technologies focus on non-CO2 gases but in some cases there can also be an indirect effect on CO2 emissions. Three of the measures shown in the table – increased share of legumes in temporary grasslands, winter cover crops, and the fallowing of organic and peat soils (histosols) – are included because they can help to reduce N2O emissions, but if implemented their contribution to reducing CO2 emissions would be much more important.

Source:  Pérez Domínguez et al, 2020. Each technology is compared to the baseline scenario 2030. Figures refer to EU28 prior to Brexit. Figures shown in red are where LULUCF emissions increase.

The mitigation potential shown in this table is the (theoretical) maximum technical mitigation potential for each mitigation option assuming that it is applied to the maximum extent possible. The figures are calculated by turning on each mitigation technology one at a time, while freezing the uptake of the other technologies to their baseline levels. Results for two scenarios are shown for each technology (except for the breeding technologies where this does not make sense).

The first scenario is the ‘tech only’ option. This estimates the direct impact of the applied technology, assuming no response by farmers in terms of production levels or production mix. We can think of this as a scenario where the introduction of the technology is fully subsidised so there is no change in farmers’ costs. This direct impact includes reductions in both non-CO2 emissions as well as CO2 emissions from land. Both impacts are added together and shown as a combined total in the table.

The second scenario is the ‘overall effect’. This scenario assumes there is a cost associated with adopting the technologies, so their forced adoption leads to adjustments in the optimal land use allocation and livestock production. These adjustments in the production intensity and production mix are due to the profit maximization framework of CAPRI. This overall effect is separated into the reduction of agricultural emissions and the change in LULUCF net emissions/removals in the table.

For each technological mitigation option the net effect on the aggregated LULUCF and agriculture emissions is a reduction in total emissions. In general, the positive mitigation effect (‘tech only’) on the ‘targeted’ emissions type is augmented once the land use allocation is adjusted to the application of the mitigation technology (though not in the case of fallowing histosols or anaerobic digestion). The ‘forced’ adoption of mitigation options reduces profitability, and this income loss is minimised by shifting away from the affected activities.

Relative costs of mitigation technologies in EcAMPA 3

The previous table does not tell us anything about the relative costs of pursuing mitigation through the different technological options. Costs in the EcAMPA project are of two kinds. The first type are the accounting costs (amortised investment costs plus annual implementation costs) of the mitigation option. The second kind refer to ‘adoption costs’. Even where a more profitable technology arrives, not all farmers adopt it immediately. CAPRI considers this is due to costs that are known to farmers but not included in accounting costs that increase more than proportionally as use of the technology expands.

Examples might be labour or machinery bottlenecks, risk premiums or rotation constraints. For this reason, the unit cost of adoption by the last farmer is assumed to be higher than for early adopters. These non-linear adoption costs have a further modelling advantage that they smooth adjustments to changes in technology and avoid sudden and large switches in production that would occur under a simple linear programming model.

The next chart shows the average unit costs of each mitigation option (averaged over all adopters) on a stand-alone basis, that is, when each option is implemented on its own to the maximum technical extent possible. The mitigation achieved for each option (shown by the horizontal width of each option) does not take into account induced changes in production, so it corresponds to the column heading ‘tech only’ in the previous table.

Source:  Pérez Domínguez et al, 2020.

The attractive measures are those with high mitigation potential and low abatement cost. They include nitrification inhibitors, fallowing of histosols, anaerobic digestion, precision farming, variable rate technology and higher legume share on temporary grassland. On the other hand, measures such as vaccination against methanogenic bacteria in the rumen, winter cover crops, and the two feed additives nitrate and linseed can deliver high mitigation but at high cost. For the two feed additives, EcAMPA estimates the average cost would be over €700/t CO2e abated.

An alternative way to present the mitigation abatement cost curve is through a combined approach. All technological mitigation options are assumed to be available at the same time and can be adopted cumulatively by farmers. Adoption rates and mitigation potential will depend on the relative costs of each option (which will vary depending on the existing level of uptake) and the carbon price applied in different scenarios. This carbon price is only applied to non-CO2 emissions. The following chart shows the projected mitigation potential and the practices that contribute to this potential at different carbon prices between €20 and €100 per tonne CO2e. This mitigation potential includes both the direct abatement due to the use of the mitigation technologies and the associated changes in production intensity and production mix.

Source:  Pérez Domínguez et al, 2020.

The solid rectangles at the bottom of each column show the direct contributions of the technological options to the total mitigation achieved under the combined measures approach at aggregated EU-28 level. The abatement potential for the technologies assessed in the EcAMPA 3 project amounts at the highest carbon price to 85 MtCO2e including emissions reductions in the LULUCF sector. This corresponds to the 80 MtCOe reported in the DG AGRI chapter, the difference most likely arising because the results reported in Pérez Domínguez et al, 2020 refer to the EU28, while the DG AGRI results refer to the EU27 following Brexit.

By far the highest mitigation is achieved by fallowing histosols at all carbon price levels. In fact, it realises about 75% of its maximum potential already at a carbon price of €20/t and further increase is limited. This is followed by anaerobic digestion, higher legume share on temporary grassland, winter cover crops and nitrification inhibitors as well as the feed additives and vaccination against methanogenic bacteria in the rumen. Some of these measures were classified as high-cost in the stand-alone approach. Their appearance in the combined measures approach indicates that they can be cost-effective at least for some farmers in some regions.

The chart makes clear, however, that most of the abatement potential arises from changes in production levels and production mix rather than from direct application of the chosen mitigation technologies (these are the two stippled rectangles at the top of each column). At the lower carbon price 61% of the abatement potential shown in the previous chart comes from production changes. This share increases as the carbon price increases, and accounts for 75% of the total abatement when the carbon price reaches €100 per tonne. The effects of the breeding measures related to milk yields and ruminant feed efficiency are included as part of the production level and mix changes. At the highest carbon price, the total mitigation potential reaches almost 350 MtCO2e. Somewhat surprisingly, this figure is reported in the DG AGRI Market Outlook chapter as 224 MtCO2e which is a considerable drop even allowing for the impact of the departure of the UK on these figures.

Another striking finding is that, on average, about 65% of the total mitigation in the scenarios relates to avoidance of CO2 emissions and carbon sequestration (reported in the LULUCF sector). This LULUCF contribution arises both from the direct effects of the technology as well as from subsequent changes in production levels and production mix. However, it is still only a secondary effect as the carbon price targets only non-CO2 emissions from EU agriculture.

The DG AGRI chapter also reports changes in soil organic carbon (SOC) arising from the baseline projections of agricultural output to 2030 in EcAMPA 4. SOC is increased as a result of increasing organic inputs to the soil and/or decreasing their loss or slowing down the mineralisation process. Organic inputs consist mainly of crop residues (including root inputs), manure from grazing animals and manure spread over the field. The study estimates this increase to be around 10 MtCO2 annually to 2030.

The impact of the three mitigation options with significant impacts on CO2 emissions included in the first table should be added to this. In the EcAMPA 3 publication, summing the LULUCF reduction in removals in the ‘overall effect’ column in the first table amounts to 64 MtCO2. The figure quoted in the DG AGRI report (for a smaller EU27) is 75 MtCO2. These estimates assume each of these measures are implemented on a stand-alone basis, assume no interactions between them, and include the indirect effects on production.

The impact of diet changes

Diet changes are expected to play an important role in reducing EU agricultural emissions (see previous post). However, dietary changes have not been explicitly modelled in the studies under review. The Commission 2030 Climate Target Plan impact assessment refers to a sensitivity analysis undertaken as part of the in-depth analysis accompanying its 2018 ‘Clean Planet for All’ Communication to understand the possible implications of differing trends in consumer preferences by the EU population on greenhouse gas emissions in the next decades.

Five diet scenarios were analysed in addition to the baseline which is based on the same assumptions as the EU Reference Scenario 2016. Diet 5 was seen as consistent with reaching in 2070 levels of meat consumption in line with recommended diets in several studies. Each scenario also included a reduction by half in the generation of food waste in all EU Member States. The following chart shows the total calories contributed by animal source foods in the different diets (on the left) and the contribution of the different foods to the overall total in 2050 (on the right).

Source:  Commission, 2018

The analysis assumes that the decrease in EU animal products consumption is entirely reflected in EU production levels and no increase in exports of animal products to the rest of the world takes place. Under that rather heroic assumption, the 2020 impact assessment concluded that the potential emissions reduction from dietary change by 2030 could be of a similar order of magnitude to the technical reduction potentials in the agricultural sector.

Source:  Commission, 2020 based on the GLOBIOM and GAINS models.

Conclusions

In the spirit of Allen and Maréchal’s review of estimates of the mitigation potential in EU agriculture which covered studies up to 2017, this post extends their review to take account of more recent studies by the Commission. The following table provides an overall summary of the findings. The table mainly uses estimates from the JRC’s EcAMPA 3 project, supplemented by information from the Commission’s impact assessment for the revised 2030 emission reduction target.

Source: EcAMPA 3 estimates as reported in DG AGRI, Market Outlook 2020-2030, Chapter 7 Environmental Aspects; 2020 impact assessment estimates are from Commission Staff Working Document accompanying the Communication ‘Stepping up Europe’s climate ambitions’, SWD (2020) 176, Part 2/2, Table 44. EcAMPA 3 estimates use AR4 weights, the 2020 impact assessment AR5 weights to aggregate non-CO2 gases. The reduction percentages to account for the additional agricultural mitigation due to production impacts are derived from the ratios of production impacts to technology impacts at different carbon prices in Table 21 of Pérez Domínguez et al, 2020. The potential mitigation from dietary change assumes all the projected reduction in consumption of animal source foods would be reflected in a reduction in EU domestic production. Additional reductions in carbon emissions in the LULUCF sector are not considered in this table.

It is a complex table but the following main points can be highlighted. It concentrates on the abatement potential for non-CO2 emissions from agriculture, although we saw that the secondary emissions reductions in the LULUCF sector will likely be more important. The emissions reductions from the technological options included in the EcAMPA 3 project are first reported, both for low (€20/tonne) and high (€100/tonne) CO2e prices. These are compared to the corresponding abatement potentials reported in the Commission’s impact assessment for the 2030 Climate Target Plan. Although both studies use different carbon price scenarios, and use different weights to aggregate non-CO2 gases to CO2 equivalents, there is broad agreement in their findings.

The next part of the table adds the abatement potential from production changes that would occur when levying a carbon price on non-CO2 emissions, again for the low and high carbon price scenarios. It also shows the upper end estimate for potential reductions due to dietary changes up to 2030. The total mitigation potential is estimated by adding these three elements together, assuming no interactions between them. This certainly overestimates the maximum potential mitigation as some of the production changes induced by the dietary changes may overlap with the changes induced by implementing the technological options.

As a benchmark, the estimated reduction potential from implementing the Farm to Fork and Biodiversity Strategies is also shown. We stress again that these are results from an unpublished study so the measures included in that simulation are not known. For example, whether it includes some demand-side changes from the food waste reduction target in the F2F Strategy, what carbon price is assumed, or whether it includes the production impacts of the technological options selected, is not known.

Despite these caveats, the conclusion from the table is that there is potential for mitigation in the agricultural sector, but this depends on how willing we are to see production fall. Without any production impacts, the adoption of technological options alone would reduce agricultural emissions by between 4-9% below the 2030 baseline depending on the carbon price. Allowing for production impacts including from the impact of dietary changes, the modelling studies suggest that a reduction in agricultural emissions even at a low carbon price of between 15-17% below the baseline trend would be a realistic target by 2030. At higher carbon prices (implying greater policy efforts and greater production impacts) this could increase up to 28%.

Still, reducing agricultural emissions is more challenging than in other sectors. The low-end estimate for agricultural emissions reduction in 2030 even including production impacts compared to 1990 is 34% when the EU economy as a whole is aiming for a net reduction in emissions of 55% (implying a gross reduction of around 53%). However, if a greater impact on production is deemed acceptable, the percentage decrease in agricultural emissions could be even higher. Further secondary emissions reductions take place in the land sector but these would be reported separately under the LULUCF heading in the inventories.

These estimates depend on heroic assumptions about accounting costs, their rate of change over time as technology advances, and adoption rates. Previous literature has shown that the range for these accounting costs is very wide. There may also be wide differences in the abatement expected from a particular practice, not least because agricultural emissions are due to biological processes that are influenced by soil type, temperature, precipitation, and management practices. Given these uncertainties, any estimate of the mitigation potential in agriculture should be reviewed with a critical eye.

One example can be given from the studies reviewed in this post. Both the Commission impact assessment and the JRC EcAMPA studies rely on the GAINS database for the costs and mitigation potential of different mitigation technologies. While the EcAMPA 3 study sees great potential in the fallowing of histosols (75% of the technical potential would already be exploited at a carbon price of €20/tonne CO2e), this technology would not figure  in the Commission’s mitigation cost curve until the carbon price is well over €100/tonne CO2e (see Figure 74 in the Commission’s 2018 in-depth analysis). Part of the explanation is that the EcAMPA 3 study also counted the (considerable) CO2 mitigation in the land sector, even when the technology is incentivised by a carbon price on non-CO2 gases.

The mitigation potentials shown in the previous table reflect the estimated cost-effective potential at different carbon prices. Carbon prices are not currently an instrument used to abate agricultural emissions and are used to stimulate a response in the different models. They are an indication of the relative stringency of the policy measures that might be used to incentivise mitigation reductions of this magnitude.

One might also surmise that the instruments used might influence the rate of adoption of the different measures. Both the EcAMPA project and the GAINS model try to model adoption rates so some account is already taken of resistance to implementing the mitigation options. My instinct is that these adoption costs underplay the difficulty of organising mitigation actions on the ground given the millions of individual actors who are expected to play a role.

The big unknown from the published studies to date is what the production impacts included in these figures mean for production levels and farm incomes in the EU. The EcAMPA 3 study was intended primarily to show ‘proof of concept’ and to demonstrate the implementation of the carbon cycle model in CAPRI. It did not give details of the production impacts or impacts on farm income of setting carbon prices at these levels. Nor did it indicate the extent to which these emissions might be shifted off-shore if domestic EU production is replaced by increased production elsewhere.

This information may be included in the yet-to-be-disclosed study of the impact of implementing the Farm to Fork and Biodiversity Strategies referred to in the Commission’s impact assessment of the 2030 Climate Target Plan or in the as yet unpublished EcAMPA 4 study if this is different. Until now, it has been taboo to discuss policy interventions that have a significant negative impact on agricultural production in the EU. This taboo was broken with the publication of the F2F Strategy where several of the interventions highlighted will result in lower EU production. Knowing the production and farm income impacts will help to better evaluate the political feasibility of reaching the 2030 mitigation potential projected in these studies.

This post was written by Alan Matthews.

Update 7 April 2021: The method used to calculate the additional reduction due to production impacts in the summary table was changed. The conclusions were redrafted to emphasise more the contribution of production impacts.

Photo credit:  Pxhere, used under a Creative Commons licence

Posted on

Climate measures in agriculture

The need and opportunities to accelerate the reduction in agricultural greenhouse gas (GHG) emissions have been underlined in a number of recent reports (see, for example, the IPCC Special Report on Climate Change and Land (2018) or the IEEP report Net-Zero Agriculture in 2050: How to Get There (2019)). Following a period from 1990 to 2012 with a steady decrease in EU agricultural emissions amounting to 22% in total, these emissions have begun to increase since then, growing by 4% over the 2012-2017 period.

In this post, I examine the projected trend in agricultural emissions to 2030, drawing on the most recent European Environment Agency (EEA) report on Trends and Projections in Europe 2019 as well as the inventory of policies and measures that Member States have taken or plan to take to reduce these emissions in future. These policies and measures are discussed in the Eionet report Overview of reported national policies and measures on climate change mitigation in Europe in 2019 and collected in the EEA database on climate change policies and measures in Europe (both accessible from the EEA webpage on Policies and Measures).

The projections indicate that Member States do not expect to make further significant reductions in agricultural emissions by 2030. Even with additional measures planned but not yet implemented in 2019, agricultural emissions are expected to fall by less than 5% between 2017 and 2030. Overall emissions are projected to fall by 18% but the EEA assesses that this will not be sufficient to meet the existing 2030 target of a 40% reduction in emissions compared to 1990.

If, in addition, the climate ambition of the European Green Deal is to be realised, further efforts to reduce emissions in the non-agricultural sectors but also in agriculture will be required. However, to properly incentivise, motivate and track progress in the farming sector, I argue that a focus on agricultural emissions alone is misleading and should be supplemented by also considering changes in land use emissions and removals that are under the control of farmers. This will require changes in the way that the EEA presents its inventories as well as rethinking the relationship between emissions in the effort sharing sectors and emissions and removals in the land sector in EU climate policy.

Trends and projections of agricultural emissions

The chart below shows the historical trend in agricultural emissions as reported in the EEA annual inventories as well as projections to 2035 as estimated by the EEA based on Member State projections. The figures are for the EU28 meaning the EU27 Member States plus the United Kingdom. They are based on converting all greenhouse gases (GHGs) to CO2 equivalents (CO2e) using equivalence weights based on global warming potential (GWP) values reported in the IPCC’s Fourth Assessment Report (AR4).

Agricultural emissions in the inventory consist almost entirely of the non-CO2 gases of methane CH4 and nitrous oxide N2O (with very minor amounts of CO2 emissions associated with liming, urea and the use of other carbon-containing fertilisers). These emissions are associated with livestock production (both enteric fermentation and manure management) and the application of fertilisers to agricultural soils (which stimulates the production of nitrous oxide as a by-product of microbial transformations of nitrogen in the soil).

The figures do not take account of carbon sequestration or emissions from agricultural land as these are reported in the IPCC sector Land Use, Land Use Change and Forestry (LULUCF). Changes of CO2 in agricultural soils due to cultivation particularly of organic soils or the incorporation of crop residues in soils are not attributed to agriculture in the EEA inventories This is also the case where carbon is sequestered in hedges, trees or through rewetting of drained organic soils.

Trends in agricultural emissions are driven primarily by changes in activity data (mainly changes in ruminant livestock numbers (cattle and sheep) and in the volume of organic and chemical nitrogen fertiliser applied. Other factors that play a role over time include changes in livestock feeding regimes, in the management of manure, in rice cultivation as well as changes in emission factors due to technical and productivity improvements in agricultural production.

The graph above shows that there was a particularly sharp fall in agricultural emissions between 1990 and 1993 (of around 12%). It then took twenty years for a further fall of 12% in emissions (from the 1993 base, 10% from the 1990 base) before emissions started increasing again. The sharp reduction between 1990 and 1993 most likely reflects the disruptions caused to agricultural production in the newer EU Member States following the fall of the Berlin Wall in 1989 and the complex transition to a market economy in these countries. It may also be associated with the introduction of the Nitrates Directive in 1991 as cattle numbers fell by 11% and inorganic N use by 14% over that short period. 

The fall in agricultural emissions mirrored the fall in EU total emissions over the period (in 2017, total emissions were 22% below 1990 levels). This is shown in the dotted line in the chart above which plots the share of agricultural emissions in the total. This shows a general stability at just under 9% of total emissions over the period 1990-2013, though this had risen to over 10% by 2017.

The chart also shows projections of agricultural emissions under two scenarios – WEM With Existing Measures, and WAM With Additional Measures. The WEM scenario is based on policies and measures existing and adopted when the projections were submitted by Member States in March 2019 under the 2013 mechanism for monitoring and reporting emissions Regulation (EU) No 525/2013 (MMR). The WAM scenario includes in addition policies that were at the planning stage when the projections were submitted.

In the EEA spreadsheet that reports the projections data the various series start at 2015 so there are three years that overlap with the inventory data. Historical emissions in the overlapping years have grown more rapidly than foreseen in the data used in the Member State projections, suggesting that the projections are likely to underestimate the trends in emissions in the two scenarios. These scenarios do not take account of any changes in economic activity expected as a result of the measures taken to prevent the spread of the coronavirus in the first half of 2020.

The projections show that Member States expect a relatively modest reduction in agricultural emissions up to 2030. With additional measures, a further reduction of around 20 Mt CO2e is projected compared to 2017 emissions of 429 Mt CO2e, or a reduction of slightly less than 5%. Economy-wide emissions are expected to fall by 18% over this period, resulting in a rise in the share of agricultural emissions to closer to 12% by 2030.

Are planned measures sufficient to reach existing 2030 targets?

The EU as a whole is currently committed to reducing its emissions by 40% compared to 1990 by 2030. EU climate policy has three legs each with separate targets: the Emissions Trading Scheme (ETS) for the power sector, large industrial plants and aviation inside the European Economic Area; effort-sharing in the non-ETS sector which includes transport, buildings, agriculture and waste; and a ‘no debit’ rule in the land use, land use change and forestry (LULUCF) sector meaning that emissions cannot exceed removals. If this sector creates net emissions, these need to be compensated by using allowances from the effort sharing sectors.

As the cap-and-trade system under the Emissions Trading Scheme guarantees that 2030 targets will be met (even though the most recent projections suggest that existing and additional measures are still not sufficient to meet the target), the main uncertainty in reaching the overall EU target is progress in the non-ETS sector. Member States have individual national targets under the 2018 Effort Sharing Regulation. Together, these are intended to achieve a 30% reduction in non-ETS emissions by 2030 compared to the 2005 base year.

There are two parallel reporting procedures that allow us to track whether Member State initiatives are likely to be sufficient to meet this target. In addition to the projections submitted biannually since 2015 under the MMR previously mentioned and which are collated and published by the EEA, Member States are also developing their National Energy and Climate Plans (NECPs). These are mandated under the Governance of the Energy Union and Climate Action Regulation (EU) 2018/1999. They represent the framework within which Member States have to plan, in an integrated manner, their climate and energy objectives, targets, policies and measures.

NECP projections

Each NECP contains a list of planned measures and stated ambitions for national GHG emissions reductions. Countries must develop their NECPs on a rolling basis, with an update half-way through the implementation period. NECPs for the first period 2021 to 2030 should ensure that the Union’s 2030 targets for GHG emissions reductions, renewable energy, energy efficiency and electricity interconnection are met.

The plans are developed on an iterative basis between the Commission and the Member States. Member States first submit a draft NECP. The Commission reviews the draft plans and can issue recommendations for increased ambition to Member States if it assesses that they are not doing enough. For the share of energy from renewable sources and for energy efficiency, where there are EU-wide targets but no specific national targets, a set of objective criteria are set out in Annexes to the Governance Regulation on which the Commission can make its assessment. For GHG emissions reduction targets, the relevant benchmarks are the effort-sharing targets for each Member State.

Member States are expected to take these recommendations into account when submitting their final plans. The Commission can also make recommendations if at any point in time a Member State is falling off track in delivering its targets asking them to take additional measures to bring them back in line.

Member States were required to submit their first draft NECP by 31 December 2018. The Commission returned these drafts with its recommendations in June 2019. The Commission assessed that the overall GHG reduction for the Union as a result of these plans would be in line with the -40% emissions reduction target for 2030 compared to 1990.

For the effort sharing sectors, it concluded that the planned measures in the draft NECPs showed that the EU could achieve a 28% reduction in emissions in non-ETS sectors. It thus sought additional measures to fill the remaining 2 percentage point gap in the final NECPs. Although the final NECPs were due to be delivered by 31 December 2019, the Commission’s NECP web page shows that, as of 18 March 2020, many major players such as Germany, France and Spain had yet to do so.

In making its assessment, the Commission relied on Member States’ own projections of the emissions reductions associated with additional measures (where this information was provided) or else the emissions reductions associated with implementing existing measures. The Commission assessment does not assess whether the measures outlined in the NECPs are capable of delivering the reductions that Member States have pencilled in to their projections.

In principle, such Quality Check/Quality Assurance should be carried out at an earlier stage. It is clear that, despite the detailed guidance available from the EEA, most Member States are struggling to provide the relevant information. This becomes clearer when the information on agricultural mitigation strategies is examined later.

EEA MMR projections

Based on the MMR projections submitted in March 2019, the EEA concludes that accounting for planned additional measures the EU27 and UK together expect to reach emission reductions totalling 36% by 2030 (compared with 1990 levels). The EEA explains that the differences with the NECP projections are due to differences in the additional measures assumed as well as differences in the way missing data are interpolated.

The EEA estimates that for the EU27 and UK together effort sharing emissions could decrease to a level 27% below that of 2005 by 2030 with additional measures. Similar to the Commission’s finding based on the draft NECPs, it concludes that a more focused effort will be necessary to reach the existing 30% reduction target.

In a further briefing on these trends earlier this month which is based on updated emissions trends to 2018, the EEA noted that non-ETS emissions had fallen by 11% between 2005 and 2018 and indeed were below the corresponding target level for the period 2013-2020. However, the pace of reduction must dramatically increase in order to reach the 2030 target, as shown in the following chart.

Source:  EEA. National action across all sectors needed to reach greenhouse gas Effort Sharing targets, https://www.eea.europa.eu/themes/climate/trends-and-projections-in-europe/national-action-across-all-sectors/national-action-across-all-sectors

“Today – more than halfway through the 2005-2030 period – the total reduction achieved so far in the Effort Sharing sectors represents only one third of the reduction needed by 2030 to achieve the target of a 30% reduction compared with 2005 in these sectors. From 2018 onwards, the annual rate of emission reductions at EU level needs to nearly double to achieve the cuts foreseen under the Effort Sharing Regulation”, according to the EEA.

During the 2005-2018 period agriculture contributed just 1% of the emissions reduction, a mere 2 Mt CO2e of the 309 Mt CO2e reduction. Most of the projected reductions in the coming period up to 2030 are also expected in the transport, buildings and industry sectors. As the following chart illustrates, agriculture remains the sector where countries foresee only minimal change in emissions over the next decade. In the next section, the measures taken and planned by Member States are analysed further.

Source:  EEA National action across all sectors needed to reach greenhouse gas Effort Sharing targets, 2020. Notes:  WEM ‘with existing measures’ projections: WAM ‘with additional measures’ projections.

Agricultural mitigation policies and measures

As part of the biannual MMR reporting, Member States are asked to report on the existing and additional policies and measures they are taking within each sector using a standard template. The EEA has created a database of the reported policies and measures based on the 2019 submissions. This allows an overview of what Member States are doing or planning to do to reduce agricultural emissions although the current standard of reporting leaves a lot to be desired as discussed later.

Gaps in reported measures

The EEA’s own status update states that the number of reported agricultural policies and measures showed a relatively larger increase between 2017 and 2019 compared to other sectors. According to its analysis, Member States in 2019 reported 212 policies and measures having effects on agricultural GHG emissions. Most (72%) were implemented in response to the EU’s Common Agricultural Policy. The most reported objectives are reduction of fertiliser/manure use on cropland (38% of all agricultural policies and measures) and improved animal waste management systems (30% of all agricultural policies and measures). However, in my extraction of the reported agriculture measures in this spreadsheet I could only identify 148 measures. The spreadsheet also includes the brief description of each measure provided by the Member State as well as the specific agricultural objective or objectives it addresses.

Of these 212 measures identified by the EEA in 2019, 55 are planned policies or measures by 12 Member States targeting additional GHG reductions from agriculture. These measures most commonly aim at improving cropland management (19 measures), reducing fertiliser use and manure application on cropland (15 measures), and improving livestock management (10 measures). Policies and measures reported are often related to the implementation of European policies such as the Common Agricultural Policy, the Nitrates Directive and the Renewable Energy Directive.

However, this catalogue of measures should come with a strong health warning, even apart from the apparent discrepancy in the total number of agriculture policies and measures just noted. This is partly because countries report policies and measures at different levels of aggregation. For example, Austria reports just one measure “This measure summarises the implementation of the programme for rural development 2014-2020 and the implementation of the Common agricultural policy (CAP). Herein measures such as improved feeding of pigs and poultry, covering of manure storage, low-loss application of manure and biogas slurry, promotion of organic farming, reduced usage of mineral fertiliser and promotion of grazing are summarised”. Other countries would report each of these as separate individual measures.

More disconcerting is that there are clear gaps in what Member States report. For example, all Member States have put in place support for organic farming. Some countries report this as a GHG mitigation measure, many don’t. Ireland lists just one agricultural mitigation measure – the use of nitrification and urease inhibitors in conjunction with nitrogen fertilizers under its agri-environment-climate scheme GLAS. But its National Mitigation Plan published in 2017 and which was the relevant planning document in March 2019 when the MMR report was submitted lists many more initiatives in its agricultural chapter. I suspect there are similar gaps for other countries.

Absence of quantification of measures

The way in which policies and measures are reported would be less significant if they were associated with clear quantitative targets. Ideally, one would like to see a specified reduction target for a sector consistent with a Member State’s overall targets, and then a series of individual policy measures with associated emissions reductions which, when summed together, would add up to the reduction target. The MMR reporting template encourages Member States to provide this information, but very few do. According to the EEA, only 18% of single measures affecting the agricultural sector have at least one quantitative estimate on expected emission savings. And the information that is provided does not seem very reliable, despite the QC/QA process it undertakes.

For example, take that Irish planned measure (listed as planned in early 2019 but also listed as implemented in 2018) to encourage the use of nitrification and urease inhibitors. Ireland reports that this measure will lead to a reduction in CO2e emissions of 187,240 tonnes in 2020, 187,340 tonnes in 2025 and 187,470 tonnes in 2030. The adoption curve of this innovation by farmers implicit in these figures does not seem credible. But Ireland at least has attempted some quantification and remains very much the exception.

The technical difficulties in estimating the impact of specific mitigation measures should not be underestimated. However, without information on the expected emissions reductions that can be achieved with different measures, and an estimate of the cost of each measure, it will not be possible for Member States to prioritise and choose between different measures nor to undertake meaningful policy evaluation.

Many Member States will have available Marginal Abatement Cost Curves (MACCs) showing various possible interventions and their abatement potential at different economic prices which could be used as the basis for developing these policy impacts. There is a very helpful review of the lessons learned in the development of agricultural MACCs in various EU countries in this 2018 paper by Eory et al in the Journal of Cleaner Production (open access version here). We are clearly a long way from this desirable state in agricultural mitigation policy.

Additional measures in the Farm to Fork Strategy

The Commission’s proposal for a Farm to Fork Strategy as part of the European Green Deal (now delayed at least to the end of April by the measures taken to control the spread of the coronavirus) will have some implications for future agricultural emissions mitigation potential but how important it will be remains to be seen. Some clues are available from the latest (13 March 2020) draft of the Strategy published by politico.eu under preparation in the Commission but it is important to remember that this is still a document under development and further changes can be expected.

Specifically, as far as food production is concerned, the Strategy is expected to increase the level of ambition to reduce significantly the use and risk of chemical pesticides, as well as the use of fertilisers and antibiotics. It is also expected to push for an expansion in the area under organic agriculture. To promote efforts to encourage carbon sequestration, it is likely to propose an “EU carbon farming manual” to quantify emission reductions and carbon removals in farms and forestry system as the basis for payments (e.g. under CAP support) and for labelling.

Additional initiatives will be proposed on the food consumption side, including measures to reduce the environmental impact of the food processing and retail sectors, reducing food waste and labelling to nudge consumers into making more sustainable food choices. Environmental NGOs are pressing for the inclusion of further initiatives to reduce meat consumption and production.

A number of these measures would indeed help to reduce agricultural emissions. Reduced nitrogen fertiliser use would help to limit N2O emissions and, depending on how it was implemented, along with an increased area under organic farming, could have an indirect effect on ruminant livestock numbers. Improving the robustness of carbon accounting on farms could open the way for additional market-based incentive policies whether financed by the CAP or by the private sector.

Successful implementation will require these targets and measures to be fully integrated into national CAP Strategic Plans. It is reported that the Strategy will encourage Member States to set ambitious targets and to put in place reporting arrangements on the delivery of the objectives of the Strategy. For this to happen, the Strategy will need to strengthen the governance arrangements for the preparation and approval of national CAP Strategic Plans.

Conclusions

The minimal reduction in EU agricultural emissions that has taken place in recent years (if measured with 2005 as the starting date) and the limited reduction foreseen in Member State projections raise many questions. It is generally acknowledged that agricultural emissions cannot be reduced to zero, but what would be an appropriate target? What is the scope for reducing emissions by adopting different agricultural practices or techniques or can emissions only be reduced by reducing activity levels? If activity levels are reduced in Europe, what is the impact on global emissions?

Agricultural emissions are predominantly non-CO2 gases. Does the current metric of aggregating different gases to CO2 equivalents using the Global Warming Potential (GWP100) accurately reflect the impact of agricultural emissions on global temperature, given the different characteristics of short-lived gases such as methane and long-lived gases such as CO2 and N2O in the atmosphere? The answer to this question will influence the priority given to reducing emissions from ruminant livestock.

The biannual review of the policies and measures reported by Member States to reduce agricultural emissions under the MMR remains deeply problematic, even if the 2019 review was the third time the exercise was undertaken. The data reported are incomplete, both with regards to the coverage of measures and their quantification. Where data are reported, they are not always accurate or reliable. As emissions reduction targets become more stringent, it is in the interests of Member States themselves to improve the quality of this reporting. The information sought is the necessary basis for designing the most cost-effective pathways for emissions reductions when establishing the climate action elements in CAP Strategic Plans.

Despite the limited reduction projected for agricultural emissions up to 2030, a number of national farm organisations (for example, in Denmark and the UK) as well as various food firms (notably concentrated in the dairy sector, see chart below) have announced their intention to have net-zero emissions by or before 2050. Given that agricultural emissions cannot be reduced to zero, these commitments assume that agricultural emissions will be offset by carbon sequestration in sinks or displacement of fossil fuel emissions in other sectors.

Source:  Image designed by Alan Matthews

Official EU policy is only partly supportive of these initiatives. The idea of offsetting carbon dioxide emissions by land-based removals remains controversial although it is fully congruent with the Paris Agreement. This sets out the objective of “a balance between anthropogenic emissions by sources and removals by sinks of greenhouse gases in the second half of this century” in order to achieve the long-term temperature goal of holding the increase in the global average temperature to well below 2°C above pre-industrial levels and pursuing efforts to limit the temperature increase to 1.5°C above pre-industrial levels.

Suspicion that carbon sequestered in natural sinks is highly uncertain, difficult to measure, may not be permanent and can be used to justify less ambitious targets to reduce fossil fuel emissions influenced the design of the 2030 Climate and Energy Framework. Fungibility between net sink removals in the LULUCF sector and emissions in the effort-sharing sector (including agriculture) is severely limited for these reasons (the linkages between the two policy regimes are explained in my blog post here). The various arguments are reflected in the responses to the Commission’s 2015 public consultation on addressing greenhouse gas emissions from agriculture and LULUCF in the context of the 2030 EU climate and energy framework and in the subsequent impact assessment.

While a refusal to allow fossil fuel emissions to be offset by LULUCF credits can be justified, this is not the case for agricultural emissions which, as we have seen, cannot be reduced to zero. Furthermore, for a farmer making decisions on how best to manage his or her land, a sharp distinction between agricultural emissions and land-based removals does not make sense. These activities are so closely linked together that the appropriate signal to farmers should be a combined target.

This is recognised in the way the Scottish government presents its emissions inventory, which reports net emissions from ‘agriculture and related land use’. In addition to emissions from livestock and agricultural soils, this includes net emissions from cropland, grassland along with net emissions from land converted to cropland and grassland. It also includes energy emissions from stationary combustion sources and off-road machinery.

In the Scottish approach, forestry removals are reported separately in a forestry sector that is kept separate from agriculture and related land use. It includes changes in net emissions resulting from afforestation, deforestation and harvested wood products. The basis for Scotland’s approach is discussed in this paper ‘The true extent of agriculture’s contribution to national greenhouse gas emissions’ by Bell et al, 2014.

My key conclusion is that it does not make sense to evaluate the efforts made by the farming sector to tackle climate change by looking solely at the trend in agricultural emissions without taking into account at the same time its contribution to carbon removals (as in Scotland and Ireland, when reporting agricultural emissions, on-farm energy use should also be included). While LULUCF emissions and removals are fully integrated into the EU’s climate policy from next year, the signals and incentives given to farmers are not coherent.

The EEA must continue to report EU emissions inventories and trends and projections to the UNFCCC in the format prescribed. However, for the purpose of tracking progress within the EU it should prepare a supplemental set of accounts along the lines of the Scottish approach that integrate agriculture and related land use. In countries where woodlands and new tree planting takes place on agricultural land, consideration might be given to including this under the ‘agriculture and related land use’ heading rather than forestry. This would both facilitate more credible target-setting for the agriculture sector as well as describe more accurately the efforts being made by the farming sector to address climate change. Focussing solely on agricultural emissions is simply misleading and will demotivate farmers who are also required to help in stabilising the climate.

At the same time, in the context of the greater climate ambition set out in the European Green Deal and the Farm to Fork Strategy and more stringent 2030 targets, there is a need to re-open the 2018 agreement on the treatment of LULUCF emissions and removals. There may be a case to discount LULUCF removals to take into account their lack of permanence and the uncertainty of measurement. However, once that decision is made, there should be no artificial limit on the amount of removals from ‘related land use’ that can be transferred to the effort-sharing sectors where they should be reported combined with agricultural emissions as recommended above.

This post was written by Alan Matthews

Picture credit: PxHere CC Public Domain

Update 22 October 2020: The disruption caused by the transition to market economies in the newer EU Member States has been added as a possible cause of the sharp fall in emissions between 1990 and 1993.