The race for carbon in agriculture

Share this article!

Carbon is the only word on everyone’s lips. Finger-pointed as the evil of the moment – due to climate change – we almost forget that we are made of it. We should stop emitting it, transforming it or storing it – we don’t want it to show its face. And the subject is a hot one in agriculture. It is one of the sectors that emits the most greenhouse gases, but also the one that has the greatest potential for storing carbon in its simplest form, the soil. It is a sad paradox that the atmosphere contains too much carbon while the soil lacks it.

Carbon measurement tools, models and monitoring methods are being developed. Labels and certification frameworks are emerging. A carbon market is being established. We will focus here on agriculture, of course, but we will allow ourselves a few digressions on the forestry sector. As a source of motivation or inspiration for some and concern for others, the subject of carbon is not unanimously supported. The fact remains that the subject is fascinating. Between climate advocates, agriculture advocates, and opportunists, carbon has a way of making heads spin.

As usual, for readers of the blog, this article is based on telephone interviews with players in the sector (whose names you will find at the end of the article), whom I would like to thank for the time they gave me. Several articles, reports and seminars have enabled me to complete the feedback from the interviews.

PS: I always give a digital prism to my articles since I work in the field of digital tools applied to agriculture. You will also find it here, but to a lesser extent than in other articles I have written in the past.

PS2 : Be aware that the article relates more to France and Europe than in the rest of the world.

Soutenez Agriculture et numérique – Blog Aspexit sur Tipeee

Legal references – Laws and regulations

There is no longer any doubt that climate change is at work. The terrible frost episodes observed in France at the beginning of April and the destruction of almost the entire village of Lytton in British Columbia by fire at the end of June will certainly have convinced the last sceptics. The IPCC (Intergovernmental Panel on Climate Change) and other international bodies keep hammering it home with ever more alarming reports. The number of objectives, protocols, agreements and strategies put in place over the last few years to guide international and national trajectories towards a reduction in greenhouse gas emissions has been countless. Who really knows where we stand? There have been so many of them that it is difficult to find one’s way around. A few points of reference can’t hurt. This part is not very sexy, but it will have the merit of restating climate policies and introducing certain acronyms that we will need later on. We’ll discuss the mechanisms, emission allowances, and carbon markets associated with all this later in this article. I promise you, we will be talking about agriculture very soon.

At the international level, it was the United Nations Framework Convention on Climate Change (UNFCCC) that set the ball rolling in 1992 (it was adopted in 1994). Signed at the Rio Earth Summit by nearly 190 countries, the convention’s main objective is to stabilise greenhouse gas concentrations “at a level that would prevent dangerous anthropogenic (human-induced) interference with the climate system”. Although the proposed ambition was very commendable, it must be admitted that the framework was still rather vague. Nevertheless, the convention had the great merit of highlighting the responsibility of the industrialised countries for climate change, and of pushing these countries to take the lead by doing their utmost to reduce emissions on their territory. It was not until a few years later, in 1997, that the Kyoto Protocol finally clarified the legally binding emission limitation and reduction obligations of industrialised countries. Over the first Kyoto Protocol period (2008-2012), the industrialised countries that signed the protocol – the so-called “Annex I” countries – committed themselves to reducing their greenhouse gas emissions by an average of 5%. The other signatories, but non-industrialised countries, known as ‘non-Annex I’ countries, have not committed themselves to any emission reductions so that they can continue to develop. The attentive reader will have noted the time taken between the signing of the Kyoto Protocol and its first period of application. A second period of application of the protocol (2013-2020), the Kyoto Protocol II, will set as a common objective for the signatory countries a 20% reduction in their emissions by 2020 compared to the base year 1990. This extension of the Kyoto I Protocol will be narrowly concluded at the Doha climate summit in 2012, as previous COPs and climate summits, notably the Copenhagen summit in 2009, had failed to reach a consensus. It should be noted that some of the countries that signed the first Kyoto Protocol had in the meantime withdrawn from the second… The Kyoto Protocol ended in 2020! It was replaced by the Paris Agreement, signed in 2015 at the COP21, which came into force on 1 January 2021. Considered historic, the Paris Agreement is the result of a common understanding among its signatories on the importance of limiting global warming to well below 2 degrees Celsius, preferably 1.5, above pre-industrial levels. The Paris Agreement remains an agreement, however, without any constraints. The increase in greenhouse gas emissions between 2015 and the start of the Covid-19 pandemic should not completely reassure us… Let’s conclude with a good point for the agricultural sector: carbon storage in the soil was recognised as a means of combating climate deregulation during COP23 under the United Nations Framework Convention on Climate Change.

At the European level, in 2020 we saw the arrival of the European Green Deal, whose initiatives aim to make Europe the first carbon-neutral continent by 2050, no less. Initial greenhouse gas emission reduction targets have been set within the European Union for 2030, with emission levels at least 55% lower than in the 1990s. This Green Deal follows on from the Energy and Climate Package (also known as the ‘energy-climate package’), which included a target to reduce greenhouse gas emissions by 20% by 2020 compared to 1990 levels. The Green Deal has therefore increased these ambitions for 2030. In agriculture, the European Green Deal has been translated into the “Farm to Fork” strategy, with a strong commitment to reducing the use of fertilisers, pesticides and antimicrobials. On the carbon side, the Farm to Fork strategy was an opportunity to consider carbon market systems to finance carbon offsetting in the agricultural sector. We will come back to this later in the article! The Green Deal and the Farm to Fork strategy should also serve as a framework for the guidelines of the new CAP, which should come into being during 2021-2022. Green credit schemes (eco-schemes, eco-conditionality…) are expected. Each member state will have to define its National Strategic Plan for the CAP (NSP CAP). For France’s NSP CAP, two levels of credits (standard and higher) should be accessible through three access routes: practices [improvement of existing practices such as plant cover or crop diversity], environmental certification [use of existing certification frameworks such as HVE or organic], and agro-ecological infrastructures also known as AEI [implementation of elements favourable to biodiversity]. The level of the amount of these green credits is not yet fully decided.

At the national level, Cocorico, France has put in place its law on the energy transition on green growth (LTECV) in 2015. This law aims to reduce greenhouse gas emissions by 40% compared to 1990 and to reduce fossil fuels by 30% compared to 2012. The LTECV gave a general orientation which was then accompanied more recently by a roadmap to fight climate change, the National Low Carbon Strategy (SNBC). In this framework, greenhouse gas emissions are to be halved, including in the agricultural sector – and carbon storage is also widely considered. The national low-carbon strategy complements and is linked to the national plan for adaptation to climate change (PNACC) programmed following the Grenelle Environment Forum. And it is also within the framework of the SNBC that the “Label Bas Carbone” was launched (we will come back to this later). In May 2021, France ratified its “Climate and Resilience” law, following the proposals of the Citizens’ Climate Convention (3C). These proposals included several articles on agriculture and food. The Climate and Resilience Law is currently being criticized quite a bit, especially by the citizens of the Citizens’ Climate Convention, who gave it a nice score of 3.3/10. Not exactly reassuring… On the agricultural side too, former agriculture minister Stéphane le Foll launched the international 4/1000 initiative in France in 2015, or to put it another way, the annual increase of the carbon stock in the world’s soils by 0.4%. Intrinsically linked to the SNBC, and by the same token to agriculture, one could also add the National Biodiversity Strategy (SNB) – biodiversity whose decline has been widely pointed out by the IPBES (Intergovernmental Platform on Biodiversity and Ecosystem Services), and the “Net Zero Artificialisation” (ZAN) strategy for soils by 2030. Following the Covid-19 pandemic, the French government deployed its recovery plan (called France Relance), which includes several measures in favour of the energy and climate transition. These include the Carbon Diagnostic Vouchers, which we will discuss later in the article.

To go a little further into this mishmash, and if you haven’t had enough, I suggest you take a look at the following graph, taken from an ADEME report. It shows the link between several of the strategies and objectives presented above.

Figure 1. International and national policy to combat climate change.

Carbon stocks and flows

The subject of carbon in agriculture is rather paradoxical. We are constantly told that there is too much carbon in the atmosphere, yet we are trying to increase the organic carbon content of soils. And for good reason, decades of chemical agriculture and intensive modern techniques will have led to a loss of 50-70% of carbon stocks in our largest storage compartment, the soil. The IPCC reports estimate that there are around 1,500 Gigatonnes of carbon in the soil, which is about three times more than what is found in the atmosphere, but with a large margin of uncertainty. Add to this the fact that carbon stocks are often measured down to a depth of 1 metre, whereas even more could be found by digging a little deeper. With climate change on the rise and carbon levels in the atmosphere rising dramatically, the subject of carbon storage has become particularly topical, in the sense that it could help solve the climate equation. However, it should not be seen as a quick fix and it should be remembered that storage will never replace emission reductions at all possible scales, including for agriculture.

Figure 2. Map of French soil organic carbon stocks (in tonnes of carbon per hectare) in the first 30 cm of soil (GIS Sol data). Not all countries have carried out this carbon stock survey

The main carbon storage process in the soil is photosynthesis. It is this absolutely incredible process that enables the transformation of atmospheric carbon into organic carbon in the plant; carbon that will then be stored in the soil when the plant biomass, loaded with carbon, decomposes (cover residues, roots, litter…). This carbon, contained in the organic matter of the soil, will be found in the soil in several forms, more or less easily degradable or mineralizable. In particular, there will be labile forms of organic matter, which can be mineralised fairly quickly by the soil micro-organisms to ensure soil fertility and the circulation of carbon throughout the soil food web, and intermediate or stable forms which will remain in the soil over long periods of time.

Figure 3. Maximum climate change mitigation potential of soils in 2030 in the pathways of the forest, agriculture and grassland, and wetland biomes, with safeguards. To the left of the vertical bar are the mitigation potentials through storage. On the right are the mitigation potentials through emission reductions. The dark parts of the bars represent soil organic carbon. The white parts represent vegetative biomass and the dotted part represents CH4 and N2O avoided through better nutrient and animal management. Source: Bossio et al (2020).

It is important to understand that carbon in the soil follows a dynamic, it is a continuous inflow and outflow. Carbon is stored in the soil by the addition of organic matter (endogenous or exogenous to the farm) and is removed by mineralisation of the organic matter. The challenge is to ensure that carbon stocks (especially intermediate and stable forms) increase as much as possible and remain in the soil as long as possible. We are talking here about the effect of reversibility or non-permanence of carbon in the sense that everything that has been stored can be destocked if so-called “destocking” practices are implemented on the farms. Hence the primary motivation to preserve as much carbon as possible where it is abundant, and in particular to avoid deforestation, the turning over of forests, or the drainage of organic soils and wetlands.

It should be borne in mind that carbon levels in the soil will always eventually reach a state of equilibrium (except in certain special cases), i.e. the levels of carbon entering and leaving will be the same. As storage progresses, adding more carbon to the soil will become more and more complicated, but the challenge will be to store as much as possible. However, it is not enough for a soil to be rich in carbon to be alive and highly functional. It is the autonomy of fertility and the capacity of the carbon to be well integrated into the soil network that should be sought. As someone else would say, if you just want to add carbon to the soil, you can bury used tyres in it…

And these stocking or destocking practices are beginning to be widely known and documented. In agriculture, although some of them are still the subject of heated debate – particularly the effect of no-till – there is a consensus on many of them: introduction of cover crops, agroforestry, management of temporary grasslands, etc. (Figures 4 and 5). We can then differentiate between practices that are more in the nature of optimising existing practices and those that are more transformative, where the farming system is really undergoing a profound transformation (introduction of cover crops, for example). These practices have a more or less marked interest in soil carbon storage, and are above all more or less costly to implement (Figures 4 and 5).

Figure 4. Additional SOC storage potential for 12 natural climate mitigation pathways. The dark grey part of the bars indicate low-cost mitigation levels (<10 US dollars per MgCO2e per year). The light grey parts of the bars represent cost-effective mitigation levels under the assumption of a global ambition to keep warming below 2°C (<100 USD per MgCO2e per year). The white parts represent the maximum additional storage on top of these two previous bars.

Figure 5. Contribution of practices to the additional storage obtained for a carbon price. On the left for 0 €/ tCO2e, in the middle for 55€/tCO2e, on the right the maximum additional storage. Source: INRAE (2019).

What is the soil storage potential? There are actually several ways of looking at it. It could be seen as a biophysical potential (what could be found under a forest or a permanent grassland) or a technical potential (what is technically possible), which are the references that are most used. But we could also define this potential from an economic point of view (the practices have a cost to be implemented) or an achievable point of view (ensuring that the practices are adopted, and setting up an environment that is favourable to the development of low-carbon projects: training, support, networks of actors, tools available, financial incentives, etc.). We will discuss this further below.

The storage of organic carbon is an issue in the fight against climate change, but it should not be considered only from this perspective, quite the contrary. It is the number one factor in all the functions that the soil can perform in the first few horizons: maintaining fertility, reducing soil erosion, structuring and carrying the soil, etc. The carbon issue is therefore a co-benefit that must be taken advantage of, but for which the soil cannot be summed up.

Different types of emissions and Scope 1 – Scope 2 – Scope 3

So far, we have talked a lot about carbon storage in the soil. To come back quickly to the part on emissions in agriculture, they are mostly known and well identified. In the French sector, agriculture is in third place with almost 20% of French emissions. The main culprits are carbon dioxide (CO2), methane (CH4) and nitrous oxide (N2O). And it is CO2 that is emitted in the greatest quantity. The problem, however, is that methane and nitrous oxide have far greater warming effects than carbon dioxide. This is a factor of 25 for methane and a factor of 300 for nitrous oxide. One unit of CH4 can therefore be converted into 25 units of CO2 equivalent, and one unit of N2O can be converted into 300 units of CO2 equivalent, so that all greenhouse gas emissions can be compared on the same basis (we will therefore speak of tonnes of carbon equivalent or tCO2eq for short). Even if they are emitted in smaller quantities, it is difficult to ignore methane and nitrous oxide… The main emissions are due to the enteric fermentation of ruminants, the manufacture and use of nitrogenous mineral fertilisers, the use of diesel to power agricultural machinery and changes in land use. These emissions are well documented in the literature.

One way to characterise the emissions of a farm or business is to think in terms of “scope”. There are three different scopes – Scope 1, Scope 2, Scope 3 – which Ademe (French Agency for the Environment and Energy Management) defines as follows:

  • Scope 1: Direct emissions from sources owned or controlled by the organisation: combustion from fixed and mobile sources, non-combustion industrial processes, ruminant emissions, biogas from landfills, refrigerant leaks, nitrogen fertilisation, biomass, etc.
  • Scope 2: Indirect emissions associated with the production of electricity, heat or steam imported for the organisation’s activities.
  • Scope 3: Other emissions indirectly produced by the organisation’s activities that are not accounted for under 2 but are linked to the entire value chain, such as the purchase of raw materials, services or other products, employee travel, upstream and downstream transport of goods, management of waste generated by the organisation’s activities, etc.

When we talk about emissions, we are often only interested in Scope 1 and 2, either because they are simpler to calculate or because they avoid asking too many questions about what happens upstream or downstream of the operation or company. However, Scope 3 emissions are often far from negligible! And they vary greatly depending on the sector under study (Figure 6). For agricultural production (number 1 in the figure), it is not surprising that Scope 1 remains the priority, since the majority of emissions take place on the farm. But for the food industry (number 2 in the figure), this is no longer the case…

Figure 6. The different Scope levels.

Not taking into account Scope 3 can lead to completely biased results. Many companies or territories point to reductions in greenhouse gas emissions in recent years, but this is only because they are not considering their Scope 3, which would then completely reverse the trend. Forgetting about scope 3 often gives the artificial impression that we are on the right trajectory for reducing greenhouse gas emissions. However, we can see that a large number of the things we use on our territory are not produced there, which means that there are emissions that need to be accounted for (we are talking about imported emissions, such as the import of feed for livestock, which leads to deforestation in the countries producing the feed; this deforestation is therefore indirectly imported). One could also define the concept of avoided emissions, which is best understood as emissions that could be prevented by implementing a project (e.g. avoiding de-stocking practices or preventing a forest from being deforested). Combating climate change therefore requires attention to reductions in all these direct and indirect emissions.

Monitoring carbon emissions and storage

Measuring carbon in agriculture

To improve a system, you have to measure it; for me, this remains a fairly relevant leitmotif. In the case of carbon, what we would like to measure in order to do things properly are carbon emissions on the one hand and carbon storage in the soil on the other. Emissions in agriculture are not so much a problem. Let’s be clear, it’s not that there are no emissions, on the contrary – carbon dioxide, methane, nitrous oxide, among others – but it’s that, overall, we know them. There is an enormous amount of literature on the subject, databases of emission factors (how much this or that product or practice emits). And these databases are still being improved. In other words, we know more or less what we are dealing with. The big part of the current problem is monitoring carbon stocks in the soil. We know how to measure carbon in a soil sample at a given time. Quite classically, we analyse the organic matter content of the sample in the laboratory. The carbon is measured and a result is generally given in terms of organic matter (the proportion of carbon in the organic matter is fairly stable, so we get by). You can obviously go further by analysing different forms of organic carbon and different forms of organic matter, but that’s not the point (or at least not for the moment). But assessing carbon stocks on a larger territory (a plot, a farm, etc.), and especially monitoring the evolution of these carbon stocks over time, is a different matter. The physical measurement of this carbon has a cost, or rather several.

Firstly, there is a cost to sampling – in terms of time and manpower. A sufficient number of samples must be taken from a plot to be exhaustive and representative of the variability of organic content at a given time, which must then be sent (and paid for) to the laboratory for analysis. And for some, this spatial variability, sometimes even within a plot, is so great that the quantity of samples to be taken would be almost impossible to achieve (several hundred samples per hectare). However, not everyone thinks so, with some stakeholders questioning the quality of current sampling. For the latter, it is necessary to carry out soil analyses at constant bulk mass or constant bulk density, but especially not at constant depth, so as not to bias the results – and soil sampling would then be more than sufficient.

In order to monitor organic matter levels over time, regular sampling campaigns would be necessary. This could be done on an annual basis, along with other conventional soil tests, or at each key stage of a project to introduce carbon-storing practices – particularly at the beginning and end of the project. One of the main problems with this temporal monitoring is that changes in soil organic matter content would be too small to be detected properly. In other words, over short timescales (of the order of only a few years) the changes in organic content in soils would be smaller than the uncertainty of the laboratory measurement of carbon. In the introduction to this section, I made it clear that there was no technical issue with laboratory carbon measurement – we know how to measure content – but I did not say that the measurement obtained was extremely accurate. In fact, you would have to wait at least 6-7 years to detect significant variations in OM in the soil and make sure that you are not in the uncertainty range for laboratory measurements.

How, then, to avoid these two main problems, namely the quantity of samples required to correctly represent a plot or a farm, and the low temporal variability of carbon in soils, which would prevent temporal monitoring. A first, fairly classic response in science is modelling. Starting from a known initial stock of carbon in the soil, we model, from experimental data, how the farmer’s practices on site would influence the storage or removal of carbon in the soil. The model thus gives an average trajectory of carbon, and this trajectory can be given at the plot, crop system or farm level depending on how the modelling is carried out. We are therefore more in a logic of means (we look at what is done, we look at the means used) and not really in a logic of results (we observe what we obtain at the end), even if nothing prevents us from checking again at the end that the modelling has not been nonsense. One can very well imagine rather hybrid approaches in which data collection is carried out on an ongoing basis, allowing the modelling to be regularly adjusted to what is observed in the field (this is known as data assimilation). Let us also add that modelling is not necessarily static (the modelling parameters are fixed at the outset) but can also be dynamic (one or more modelling parameters are measured over time). Nevertheless, modelling cannot do without an initial carbon stock measurement so that the reference state is at least known with the best possible accuracy.

As the measurement of carbon stocks in absolute terms is already somewhat uncertain, there is a tendency to work on variations in stocks – a relative evolution of carbon, shall we say – rather than on changes in stocks in absolute terms. This is what one of the latest INRAE reports on the subject recommends (Yogo et al. 2021). In the framework of the AMG model tested (we will come back to all these models shortly), we can see in particular that monitoring stock variations rather than absolute stocks makes it possible to reduce modelling errors quite considerably. It is therefore preferable to say that in 5 years, the soil has stored 3 tonnes of carbon, rather than trying to estimate the output stock with precision. It is therefore a variation, a delta, or a differential that is perhaps more relevant to measure.

To compensate for the exhaustiveness of field sampling for monitoring carbon in soils, several players have positioned themselves on data acquisition systems. The main one is the satellite, whose images are taken regularly and cover relatively large areas. I will not dwell here on explaining all the characteristics of an image. The curious or neophytes can go and dig around the spatial, temporal, and spectral resolutions of the main satellite constellations used in agriculture. Keep in mind that in the context of carbon, the satellite image could be used for two things. Either to directly measure a carbon stock in soils by combining the different spectral information of the image, or to follow over time a parameter of interest integrated in a soil carbon estimation model (biomass in particular). At present, the satellite seems to be of more interest for its temporal dynamics and its monitoring of input parameters for agronomic models than for an absolute carbon measurement in due form, for several reasons. The image already gives a superficial view of the soil, whereas organic carbon is generally studied at soil horizons of 0-30 cm or even much deeper. The satellite signal does not penetrate many of the soil layers. The switch from optical to radar signals is not likely to help matters much. And the satellite is also faced with relatively small changes in soil carbon content over time. If variations are difficult to detect in the laboratory, they are even less likely to be detected by satellite. Some work in France is nevertheless interested in satellite tools for measuring surface organic carbon levels (Vaudour et al., 2021). In the near future, however, one might wonder whether this work will still be relevant for agricultural areas if all farmers have their soils permanently covered – the satellite will then no longer be able to observe bare soil colour to measure the level of organic matter on the surface. Besides, once you put carbon into the soil, the soil becomes darker, and it has a lower albedo. Once you store carbon in the soil, you would have to cover the soil to prevent the benefits of storage being lost through these albedo effects. Satellite work is based on the analysis of spectral information in the image, known as spectrometry. Similar approaches to spectrometry, but very close to the ground, are also being deployed. On-board spectrometers – for example the Verris technology used in France by players such as PreciField – or portable spectrometers are used to measure organic matter levels in soils. One of the main limitations or points of attention of these systems is the calibration of the sensors and the dependence of the spectral signal on soil conditions (water content, structure, etc.).

If large-scale sampling capacity is not technically or economically feasible, is it then better to monitor relatively few samples but always at the same location and over a long period of time, or to model the evolution of carbon content in soils using dynamic soil or vegetation parameters that can be monitored by spectrometry? Is soil monitoring on a very precise GPS point in time really possible insofar as the passage of machinery (even without tillage) and climatic agents can move some soil? We could also add that soil sampling is very concrete, and that it is something that speaks to farmers. It allows us to lay a certain number of foundations. For the deployment of a carbon reduction and storage dynamic in agriculture, there is a need for simple field indicators to be put in place on the farms. The organic matter content has the great advantage of making people think and move forward. As far as modelling is concerned, one thing is certain: on-site data are fundamental, particularly for initialising models. And modelling will be all the more accurate if it is possible to intelligently combine field measurements (initial carbon stock), farmers’ data (practices, interventions, etc.) and satellite data (dynamic monitoring of parameters of interest).

Market Models, Tools and Methods

Some terminology

Let us begin by clarifying a few terms. We will try to stick to them as much as possible in the rest of the document, but this will at least be an opportunity to provide a framework so as not to get too lost.

  • The term “data” will already include anything that is used in any way to quantify or describe carbon levels or contents, and several of these have already been mentioned in the previous section. Example: samples, laboratory measurements, and to which we can add farmers’ data, soil databases…
  • Then come the models which, as described, are based on a set of experimental data, and whose objective is to estimate the content or evolution of carbon content over time. It is important to check the conditions of applicability of the models (according to soil types for example). Example: AMG, SAFY-CO2, RothC, STICS, CHN…
  • These models can be integrated into tools, which are in fact nothing more or less than a sort of slightly more human interface for the model(s) integrated into them. We can therefore, for example, parameterise a tool, have it tested by a user, and so on. Keep in mind that a tool does not necessarily use a model. It can simply use data from measurement systems, or carbon emission factors. Some tools and/or indirect models focus on carbon emissions while others will focus on soil carbon storage. Some focus only on carbon, while others will consider a wider range of greenhouse gases, including methane and nitrous oxide. Some tools may be certified (Ecocert, ISO, 2BSvs…), others not. Some tools may have several data entry modes (a very simple but coarse data entry interface, and an advanced data entry interface for more detailed results). Example: Simeos-AMG, MyEasyCarbon, Carbon Track, Cool Farm Tool, CAP’2ER, Systerre…
  • The methods provide a framework for projects that aim to reduce emissions or implement carbon storage practices. Some methods can be certified, others not. In the context of carbon, a clear distinction can be made between environmental certification (which will tend to target a farm – it will be said to be in the process of progress) and environmental labelling (which will tend to target a product). Methods can be considered as standards if everyone considers them to be a reference. Methods may or may not recommend measurement tools. Methods may or may not comply with national or international recommendations or reports (GHG Protocol, IPCC). Example: Label Bas Carbone, Carbon Agri, Carbocage, ABC’Terre, Gold Standard, Verra VCS, Regen Network…

To compare the different tools, one way is to use the hierarchy proposed by the IPCC under the name of “Tier”. Tier 1 corresponds to tools that use emission factors from very generic models, which are not specific to a country or region. For example, such an agricultural practice stores or emits so many tonnes of CO2. Tier 2 goes a step further, with emissions factors that are often a little more specific to the country, region or condition in which the practices are implemented. Tier 3 tools are the most advanced and make use of dynamic carbon models (soil and/or vegetation parameters are followed over time to refine the models). The idea in general is not to say that this or that tool is better than the others – it can be used for that too – but rather not to compare cabbages and carrots. If we can, for example, avoid emission factors, we can find interest in going for Tier 3 tools. Within the framework of a territorial dynamic, it may be relevant to use what we know how to do with dynamic satellite monitoring.

To compensate for the quality or uncertainty of the available data that may or may not be used to feed models and/or tools, several methods have set up a discount system. These discounts are often recommendations, as in the case of the Label Bas Carbone arable farming method, which should be officially validated in mid-2021 (Figure 7). These discounts are in addition to mandatory discounts for the risks of non-permanence of carbon in soils (we will come back to this later). Each method decides, guides or proposes how to consider the discounts.

Figure 7. Example of the calculation of the discount linked to the uncertainty of the input data of the soil carbon storage models for the Label Bas Carbone arable crop method. Source: AgroTIC Chair and Elsa-Pact Chair seminar.

Without being exhaustive, let’s go into the details of some French and international methods and tools. To complete these details, you will find comparison tables and tool and method sheets in a recent INRAE report (Yogo, 2021).

The Label Bas Carbone method: LBC

The Label Bas Carbone (LBC) method is a government-sponsored strategy under the auspices of the French Ministry of Ecological Transition and Solidarity (MTES), in support of the National Low Carbon Strategy (SNBC). It is a government initiative to encourage French farmers to reduce their emissions and implement carbon storage practices, and to encourage French and international companies to take an interest in projects carried out on French territory. The aim is to provide guarantees on the ground and to seek remuneration for these emission reductions. It is an incentive to implement low-carbon practices.

Depending on the sector and context, the LBC method has either been deployed, is being deployed or is being developed. For example, there are variations on livestock farming (the method is called Carbon Agri), field crops, vines, orchards, hedgerows and methanisation plants. It is not the ministry or the agro-industries that impose new methods but the players in the sectors that propose these methods. For each of these sectors, the consortium, agricultural professions, federations or technical institutes mobilised are different.

The LBC method is a voluntary carbon certification framework for private projects. Each structure or group of farmers (the LBC method does not really target individual farmers) wishing to have their emission reduction or storage project certified can set up their project and apply to the Ministry of Ecological Transition. The ministry is not really looking to target individual projects but rather to aggregate a critical mass of farmers to collectively seek funding. Only projects on French territory can be labelled by the LBC. Following a project labelled by the LBC method, it is the emission reductions that will be certified and not the farm itself. A farm that embarks on an LBC certified project will not be certified as a low carbon farm. What will be certified is the fact that the farmer’s practices have reduced emissions and/or stored carbon in the soil. Once recognised and certified, the emission reductions can be marketed, the so-called “carbon credits”, to which we will return later. The carbon reduction units are the property of the project owner (farmer or forester)

The LBC method is an obligation of means, there is no obligation of results. An initial diagnosis is carried out on the farms by collecting field measurements and data from farmers. These data are then used to define a reference scenario. This scenario will finally be projected into the future after modelling with the farm’s current practices, and will be compared to one or more scenarios that are also theoretical but use more virtuous practices on the farm. It is the difference between these two scenarios, both projected, that will allow the assessment of the emission reductions and/or carbon stocks on the farm. Be clear that the scenario with virtuous practices is not compared to the reference state at the time of data acquisition, but to the reference state projected in time, considering that the practices on the farm do not change. For the soil carbon storage part, the absence of an obligation to achieve results does not require a return to the field to establish a new state achieved. The modelling is sufficient in itself. The LBC method considers that if the agronomic levers are applied, the project is necessarily heading in the right direction.

On the emissions side, the reductions are evaluated at the end of the project by studying the farm’s invoices (consumption of nitrogenous mineral fertiliser, fuel, etc.).

The LBC method is suitable for all types of farming profiles, whether it is someone who has not implemented any virtuous practices, or on the contrary, someone who is already very committed to reduction or storage practices and who would like to have what has already been put in place recognised. The difference will depend on the reference scenario chosen. Let’s take the example of carbon storage. In the first case, where it can be assumed that the margin of progress is significant (we will speak rather of additional carbon storage), the farmer can compare himself with or without practices. This farmer, perhaps going from 50 to 52 tonnes of carbon in his soil, will have stored 2 tonnes. In the second case, where the margin of progress is already more limited (we will speak rather of maintaining stocks), the farmer will be better off choosing a generic reference – such as an average on farms with similar soil and climate conditions to his own. A farmer who already has a stock and who will maintain it over time will therefore be able to have it recognised. The objective is to have him maintain his practices to avoid destocking carbon. This farmer, for example, who already has a soil stock of 52 tonnes of carbon, will avoid falling back to 50 tonnes of carbon if he stops his storage practices; he will then have avoided destocking 2 tonnes.

The LBC method looks at the farm and/or cropping system scale, but it does not consider each crop independently.

At this stage, the LBC method is not linked to the CAP.

The CAP’2ER tool

The CAP’2ER tool is the Carbon Agri method tool, which in fact corresponds to the LBC method for livestock. Set up by the CNIEL and IDELE, the Carbon Agri method has been submitted to the Ministry and certified in September 2019. The CAP’2ER tool did not come out of nowhere. For nearly 10 years, IDELE and its partners have been working on the issue of agricultural impacts on the greenhouse effect, air quality and the preservation of energy resources through the Ges’tim project, which will also be updated soon.

The first carbon project deployed in 2019 – Carbon Diary – made it possible to carry out a carbon diagnosis on 4,000 farms with the CAP’2ER tool. Today, by mid-2021, there are 15,000 farms involved for 22,000 diagnoses (some farms have already carried out two diagnoses)

In the same vein as the LBC, the Carbon Agri method is a voluntary approach, the aim of which is to involve as many producers as possible. The CAP’2ER tool is a Tier 1 tool, based mainly on emission factors and empirical data from the literature. Although it also looks at carbon storage in soils through grassland use, its main contributions are on greenhouse gas emission reductions, carbon dioxide, methane and nitrous oxide.

We will discuss carbon credits later, but in the context of the emission reductions generated by the Carbon Agri method and to manage the carbon reduction mechanisms, the “Jeunes Agriculteurs” (Young Farmers) have created the France Carbon Agri association, of which IDELE is the technical operator. Unlike livestock farming, where the project agent – France Carbon Agri – is very well identified, the field crop sector will see many different agents (Coops, CETA, GIEE, CER, consultancies, etc.) to develop collective projects for farmers. The operation will be the same but this ecosystem will be much more fragmented and decentralised than it is in the livestock sector.

A European version of CAP’2ER is under development.

The SAFY-CO2 model

Developed by the CESBIO and the CNRS, the SAFY-CO2 model (or SAFYE-CO2 when a water component is integrated into the model) is an agro-meteorological model whose objective is to estimate carbon storage by the plant, by making a direct link between this storage and the biomass observed by satellite. The use of Sentinel 2 satellite data makes it possible to show the carbon balance of the year (CO2 fluxes and carbon exports at harvest) at a plot scale. Note also that the model does not stop at biomass but is also interested in yield estimation. As you may have understood, the SAFY-CO2 model is a Tier 3 approach.

For the moment, SAFY-CO2 is only parameterised for four crops (wheat, sunflower, maize, rape) with a rather generic parameterisation, which can nevertheless evolve according to the type of cover present. For winter crops, two parameterisations are available, one for short cover crops (mandatory cover crops in relation to the nitrate directive) and another for long cover crops, those destroyed at the end of March or beginning of April. For summer crops, the task is a little more complicated insofar as the distinction between intermediate cover, regrowth, and/or weeds is not obvious. The composition of the cover crops is not taken into account for the moment, due to lack of information, but may be in the coming years depending on the data provided by the farmers.

In the current version of the SAFY-CO2 model, the estimates of organic matter evolution are only valid on soils depleted in organic matter, which is the case for the vast majority of agricultural plots. A thesis is underway for soil contexts richer in organic matter with the aim of having relevant parameters to enter into the model for these situations.

Several projects and initiatives are underway:

A heavier treatment chain, AgriCarbon-EO, is currently being deployed. This chain will make it possible to consider both Tier 1 and Tier 2 approaches, but also Tier 3 by integrating the SAFY-CO2 model. And the processing chain should be able to recover data sources from land management software, in particular from the Mes Parcelles software. A coupling between the AMG soil dynamics model (presented in the next section) and the SAFY-CO2 model could improve the estimation of organic matter decomposition in the soil. SAFY-CO2 currently only estimates it with a simple approach based on temperature levels. AMG, on the other hand, goes much further by integrating the chemical properties of the soil. The combination of the two models will make it possible to combine soil and biomass models to better estimate changes in soil carbon stocks. It should also be noted that, for the moment, AgriCarbon-EO does not take into account the nitrogen applied and its impact on greenhouse gases. Unlike the SAFY-CO2 model, which was initially oriented towards research, the developers of the AgriCarbon-EO processing chain have worked on propagating the uncertainties of the input data to the outputs of the treatments. No decotations are currently required by the CAP or any other mechanism depending on the accuracy of the model, but in a broader context, it is conceivable that this could become the case. The objective of this uncertainty propagation work is to quantify the uncertainty on each of the simulated variables (CO2 flux, yield, biomass, etc.) in order to produce uncertainty maps. The AgriCarbon-EO processing chain has been designed to be consistent with the MRV (monitoring, reporting, verification) strategy of the CAP, and is compatible with a low-carbon label method.

The AgriCarbon-EO chain is still a tool that is closer to the research domain. It is not yet operational, and the question of scaling it up, or in other words who will be in charge of managing the satellite data and using the AgriCarbon-EO chain, will arise. Could the ministry do this? Or perhaps the European Copernicus programme? Someone will have to step in to provide the service.

In the wake of COP21, the H2020 Circasa project has defined a methodology to address carbon footprints in a spatialised manner. The project ended a few months ago but a follow-up should be set up. Another H2020 project, the NIVA project, is still ongoing in 2021. NIVA aims to develop a nested approach with three levels of complexity – still following the Tier 1, Tier 2, and Tier 3 taxonomies – on carbon, nitrate leaching and biodiversity indicators. The Tier 1 level is based only on satellite data and the graphical parcel register (RPG). For the carbon balance maps at the plot level, a link is made between the duration of soil cover and the quantity of CO2 absorbed (the link was established on about twenty sites with flux exchange measurements by Eddy Covariance). The relationship is generic and can be applied to many crops (except rice). The objective is to produce maps, at the plot or pixel level, of areas that fix or not CO2, and to monitor plots close to equilibrium. The Tier 2 level still uses this empirical relationship between the duration of soil cover and the quantity of CO2 absorbed, but also integrates farmer data to achieve a more complete carbon balance by calculating the quantities of carbon exported at harvest (in grain, straw, fodder, etc.) or added in the form of amendments. Tier 3 is based on the SAFY-CO2 model described above.

The Quantica project focuses on the specific contribution of intercrops. The objective is to specifically measure the biomass of the cover crops with the aim of refining the carbon balance at the plot level: taking into account both what happens in crops and intercrops over one or more years. We are specifically interested in intercropping and the biomass of the cover crop to store carbon in the soil. The detection of intercropping works quite well to discriminate bare soil from covered soil. Radar data can even be used as a complement to compensate for the presence of clouds. Estimating biomass by radar is far from simple, as these signals are sensitive to soil moisture and water content (the company OneSoil already offers biomass maps interpolated with radar data). The most important thing is to have satellite images at the end of the growth of the canopy, as this is what will be buried in the soil. However, the question of the cover crop is not obvious, especially as these are often mixtures.

The AMG model

Developed by AgroTransfert and its partners, the AMG model is a dynamic soil model – it is therefore a Tier 3 model – which focuses on soil carbon storage and organic matter management. The model considers a set of simple input data to describe the farm (yields, intermediate cover, tillage, irrigation, etc.) and to estimate the carbon that will enter the soil through crop residues and be transformed into humus. By adding the type of soil and climatic data, it is possible to obtain a humified soil carbon balance. The output of the AMG model is the evolution over time of the carbon stocks and its content in the surface layers.

The AMG model could also be coupled with remote sensing data, as discussed in the previous section about SAFY-CO2. Classically in AMG or in other dynamic soil models, biomass is not measured but estimated, often from yield and allometric functions. This source of uncertainty could be reduced by measuring a biomass level from satellite imagery.

The AMG model is integrated into the Simeos-AMG tool, which is basically AMG with an interface. The software is made to help see what could be expected from new practices compared to a reference situation and to make a decision on the long term. The software is also used to make simulations and to project oneself (by asking for example if there would be interest in exporting straws, and if so at what rate…)

The AMG model was developed for vines in collaboration with IFV and INRAE. The mineralisation functions used are the same, only the model inputs change. Discussions are underway to implement the AMG model on grasslands.

The AMG model is also an integral part of the ABC’Terre method, the objective of which is to carry out a greenhouse gas balance on the scale of the territory, by integrating, among other things, the effects of carbon in soils thanks to the AMG model. The inputs to the AMG model are determined using relatively large data sources, in particular the graphical parcel register (RPG) to reconstitute rotations by soil type and farm, but also the soil analysis database (BDAT) set up by the Gis Sol. Understand that ABC’Terre uses the outputs of the AMG model, but is not exclusive to it. It also includes emission factors from the Agribalyse database and IPCC emission factors.

Cool Farm Tool

Originally developed by the University of Aberdeen, Unilever and the Sustainable Food Lab, Cool Farm Tool (CFT) is a tool for calculating the carbon footprint of farms, considering both emissions and soil carbon storage. The CFT considers Tier 1 and 2 for livestock (dairy and beef) and crops, as well as a “simple” Tier 3” model when it comes to N2O emissions and soil carbon sequestration. The tool is partly aligned with a number of standards and protocols (IPCC, GHG protocol…) but does not necessarily seek to be fully compliant. Even if some approaches are still rather simplistic – for example on plant cover – the tool continues to be improved regularly. Keep in mind that Cool Farm Tool is mainly referred to for its carbon aspects, but the tool also looks at water resources, nitrogen efficiency, and biodiversity.

The tool is relatively simple to use, with many default values (it is also possible to define ranges rather than single values), and the majority of the input data is filled in manually. The tool links to external databases, notably to retrieve climate data and water availability. Unlike other tools or models we have described, the tool does not use satellite data to calculate soil biomass or carbon levels. However, APIs would be available to work with a geospatial player, GeoFootPrint.

The Cool Farm Tool is not yet widely used in the carbon markets – it was initially a tool for decision support. Cool Farm Tool is still a measurement tool. It is not a certification tool and does not currently qualify a carbon credit. However, the Cool Farm Alliance is working to improve the tools’ structure and methodology so it becomes compatible with the variety of carbon crediting schemes globally. For carbon markets, the Cool Farm Tool is promoted by members such as the Danish company Commoditrader or the Belgian company Soil Capital, which considered that the tool enjoyed a sufficiently solid scientific and industrial consensus (many industrialists participated in its financing) to rely on it. Soil Capital is setting up carbon reward schemes for farmers, certified to ISO 14064 (an ISO standard for quantifying, monitoring and reporting greenhouse gas emission reductions or removal improvements), and the ISO certificates generated are sold by its partner South Pole.

The Verra VCS method

Verra has developed its own programme – VCS – describing a set of procedures for calculating greenhouse gas emission reductions. Specific methodologies are then developed for each activity and sector (industrial processes, construction, waste, transport, etc.). There are several methodologies specific to agriculture, including the SALM methodology (VM0017), and more recently, a more integrated and general methodology – IALM (VM0042). The proposed methodologies are reviewed by Verra, but also by independent third parties. A public consultation is also set up. VCS-labelled projects are also audited by a structure independent of Verra. The company is not involved in carbon credit transactions at all (Verra does not sell credits directly), as contracts are set up outside the VCS programme.

At the moment, the majority of VCS standardised projects are in the forestry context. Few projects are still oriented towards agriculture. The VCS standardised methodologies are currently very oriented towards soil sampling (revisions are underway to combine soil measurements and dynamic models), and are very rigorous on the verification and monitoring requirements

Identified limitations and weaknesses of existing tools

From a purely agronomic point of view, the subject of carbon is still the subject of much debate among the main stakeholders. Without getting into parochial disputes – and this blog post is not a scientific article – let’s quickly review the main points of attention raised by the community (the points are not exhaustive). The aim here is not to provide a clear answer but rather to raise a number of questions:

  • Progress seems necessary to better simulate the evolution of the carbon stock in deep horizons and in particular to take into account new mechanisms such as the “rhizospheric priming effect”. Roots produce thousands of exudates depending on the context in which the plant is located. In soil that is stuck to the roots – known as rhizospheric soil – carbon could be mineralised 2 to 3 times faster. In current models, the quantity of carbon returned by the soil could therefore be underestimated. This priming effect could also lead to a revision of the coefficients of mineralisation in soils.
  • Some people criticise the long-term experimental data used in the carbon estimation models for lacking completeness, with too few soil types and textures considered. Results from experimental stations are sometimes criticised for being in contradiction with results in the field from pioneer farmers (not the same agri-equipment [seeders…], not the same working methods…). More specifically, the French carbon estimation models would not be calibrated for high biomass plant cover crops (whose quantity of carbon stored would then be largely underestimated). Some of the tools currently proposed would nevertheless make it possible to overcome this problem by artificially multiplying the frequency of crop returns. Other criticisms relate to the fact that direct seeding under cover would not be considered (direct seeding alone is). The plant cover used would also not be deep enough to store organic matter in depth (lack of soil compaction management). Organic matter profiles in grasslands would not be deep in the sense that they are cut all the time (or grazed)
  • The effect of no-till on soil carbon storage/removal is still a hotly debated topic. The latest INRAE report and recent meta-analyses suggest that no-till has no effect on the total amount of carbon stored, as the carbon is distributed over the entire vertical soil profile. Opponents criticise the experimental data for having been studied by considering each agricultural practice (stopping ploughing, introducing cover crops, etc.) independently, whereas these practices would have significant combined effects. The latter add that the effect of no-tillage would be all the more important as the soils are already very rich in organic matter.
  • Certain agricultural practices, notably fertilisation and irrigation, would be in opposition to the storage of carbon in the soil. Fertilisation and irrigation would increase the mineralisation of carbon. There would therefore be a balance to be struck between fertilisation and yield levels so as not to encourage the removal of carbon. This dilemma could particularly affect farms undergoing organic conversion which, if they do not use strong intercropping, could be led to destock carbon by adding nitrogenous fertilisers. Nitrogen fertiliser seems to be necessary at the beginning to reach the self-fertilising state mentioned above.
  • There would also appear to be a trade-off between soil carbon storage and nitrous oxide emissions.
  • One of the main levers for reducing greenhouse gases – reducing the number of animals on the livestock farm – would not be taken into account in some of the methods for scoping greenhouse gas emission reductions.

Some additional information

  • INRAE and Planet-A are developing the SOCCROP indicator to monitor the evolution of carbon stock in agricultural soils
  • The organisation PADV (for a living agriculture) is working on a farm regeneration index. This index is not specific to carbon.
  • Some actors to follow on the subject of carbon: MyEasy Farm, Soil Capital, AgroTransfer, Carbone Farmers, Rize Ag, Indigo, Nori, Truterra, Climate Action Reserve, TerraCarbon…
  • Some information on carbon standards

A carbon market that is being set up

The main principles of carbon payout

Carbon markets are subject to a number of key principles that must be kept in mind to ensure that a euro spent really does go towards a project that has an impact:

  • Additionality: To demonstrate the additionality of a carbon project, one must be able to prove that this carbon project could not have taken place without the proposed aid and/or financing. In other words, it must be shown that this project goes beyond the conventional trend, beyond what would have been done naturally, or beyond what the regulations would have imposed anyway. If the farmer or other beneficiary of the project has their own funding, there is no reason why they should get carbon funding as they could have somehow set up the project on their own.
  • Double counting: It is necessary to ensure that a GHG emission reduction or carbon storage action implemented as part of a carbon project is not counted or paid for twice for the price of one. This principle of double counting will become increasingly important because there will be more and more offsets and schemes put in place. A concrete example: An agricultural cooperative is ensuring that its members develop a low-GHG sector. Who gets the carbon credits? To the members who have reduced their footprint? To the cooperative? Or perhaps to the downstream distributor who has imposed a low-GHG product? And if a member has several low-GHG sectors, how can the carbon count of the actions he or she has implemented for these sectors be shared, knowing that certain global actions on his or her farm will have served both sectors? We could already allow this member to exclude the emissions of one of these crops, but we would not be completely out of the woods. How does an agribusiness trade with another industry if one of them wants to finance a low carbon project? Should we separate emission reductions from carbon storage? Do the stored credits belong to the agribusinesses? If an agribusiness finances a farmer and the latter wants to make even more effort, should new credits be generated? In short, an accounting puzzle that is far from being completely resolved. In agriculture, this issue is extremely complicated because the sector is multi-subsidised, and there are many players. As the sector feeds several markets, all of which will (have to) share emission allocations, the task will be far from easy. Add to this the fact that if there were only one counting method, it would be fairly straightforward. But when several methods are put in place – we can take the example of the Label Bas Carbone (LBC) for livestock, field crops or methanisation – it becomes much more complicated. Note, for example, that the Carbon Agri method will soon be updated to include the LBC method for field crops. To reassure ourselves a little, there are some reduction/storage methods that do not overlap too much either, and that will allow us in some cases not to get our brains in knots (for example between the Carbocage project on hedges and the low carbon label for arable crops).
  • Permanence or non-reversibility: it is all very well to implement an action to reduce emissions or store carbon, but it must be sustainable over time! The first example is a fairly simple forestry project where I plant a tree. When this tree grows, apart from the fact that it is cut down, burnt or diseased (that’s quite a lot, you might say), the carbon storage is sustainable. In agriculture, it’s a bit more complicated than that because the storage comes from the fact that the farmer is going to put in place practices. But if five years later, either because he’s fed up, or because he sells his farm, or for any other number of good reasons, he re-emits the carbon he helped store, the permanence of the credit is no longer guaranteed… Some methods impose mandatory discounts to take into account these aspects of non-permanence or non-reversibility; we’ll come back to this a little later.

All these major principles directly or indirectly entail a need for traceability of both the action and the financing of the action. Many methods impose strict conditions for auditing and verifying projects, and this is done in a totally independent manner by third-party auditors. Information and investment must be tracked. It is also important to work at the system or farm level. Working on a crop-by-crop basis certainly seems more attractive because of its simplicity – it is relatively easy to find relevant environmental monitoring indicators – but the approach is not very robust to side effects and cross-crop impacts.

Carbon markets

The bond markets under the Kyoto Protocol

In 2005, following the Kyoto Protocol, three main carbon markets emerged:

  • An Emissions Trading Scheme (ETS). You may have seen the acronym EU-ETS, which is in fact nothing more than the European Emissions Trading Scheme. Keep in mind that many countries or groups of countries have set up emissions trading systems (Europe, China, Korea, Switzerland…) but that these different systems of allowances are hermetic (for the moment, it is not possible to trade allowances between different systems). The European Emissions Trading Scheme has been broken down into several phases over time (the 3rd phase runs from 2013 to 2020 and the 4th phase from 2021 to 2030). This system was set up to encourage the most polluting companies (energy-intensive industries, electricity producers, etc.) to reduce their greenhouse gas emissions and limit their carbon footprint. An emissions quota (what they are allowed to emit at the most), called EUA for European Allowances, is then set by sector. Companies can resell allowances if they have emitted less than they were allocated, or buy them if they have emitted more than expected. The idea is to progressively reduce the ceiling of EUAs distributed so that companies emit less and less. In the European market, the reduction target for 2020 compared to 1990 was 20%.
  • A joint implementation mechanism (JI). In this JI market, industrialised countries, i.e. the Annex I countries of the Kyoto Protocol (see the first section of this article), have the right to buy emission reductions from other industrialised countries through ERUs (emission reduction units) generated by JI projects.
  • A Clean Development Mechanism (CDM). The principle of the CDM market is the same as the JI market, with the major exception that an industrialised country buys emission reductions from emerging or developing countries (countries that are not in Annex I of the Kyoto Protocol), and not from other industrialised countries. The credits generated by CDM projects are called CERs (certified emission reductions). The aim was to give industrialised countries the opportunity to finance projects in southern countries and reduce their emissions there. The idea was that the constraints and costs of reducing greenhouse gas emissions in the South would be lower than if these emission reductions had been implemented in an industrialised country. This clean development mechanism (CDM) is nevertheless widely criticised for the uncertainties surrounding the additionality of projects, the sectoral and geographical distribution of projects, and the governance and ethics of the projects implemented (Demaze, 2013).

Carbon credits can be traded between players (buyers and sellers) either on marketplaces, directly or via an intermediary, or over the counter. These three markets, the ETS, JI and CDM, are considered to be carbon bond (or regulated) markets in the sense that states and companies are subject to emission reduction targets and constraints. As you may have already realised, these three markets are quite different, apart from the fact that each one involves accounting for tonnes of CO2 equivalent. In an emissions trading scheme, companies and countries trade emission allowances (okay, we’re not that far along…). These emission allowances (we have talked about EUA : European Allowances) are in a way rights to pollute. But you have to realise that they are rights to pollute in the future. On the contrary, in the JI and CDM markets, countries and companies buy emission reduction credits (ERUs or CERs) which are in fact offsets for emissions that have already taken place. They are therefore not rights to pollute in the future but compensation in the past.

The EU ETS only applies to the most polluting sectors. Not all sectors are subject to it and therefore do not have the same objectives and regulations to respect. In addition to the EU-ETS, there are two other climate policies for Europe:

  • The effort sharing policy, referred to by the acronym ESD (Effort Sharing Decision) over the period 2013-2020, then ESR (Effort Sharing Regulation) for the period 2021-2030. The sectors subject to this effort sharing (transport, construction, agriculture, waste) are subject to annual emissions caps, expressed as a percentage of 2005 emissions. By 2030, emissions from these sectors must be reduced by 30%.
  • Land use, land-use change and forestry (LULUCF), which are changes in soil and forest carbon stocks. LULUCF is currently the only sector that allows for negative emissions thanks to natural carbon sinks: biomass (forests, hedges, agroforestry, etc.) and soils (agricultural soils, etc.).

The European objectives are then broken down in each country, which then break them down by sector. Each company in its sector (ETS, ESR or LULUCF) therefore has different reduction targets to meet. However, a number of flexibility mechanisms have been put in place to help countries and companies meet their targets. For example, it is possible to carry over credits from one period to another: emissions trading schemes have been defined over several different periods – it would then be possible to carry over a credit from, say, the 2013-2020 period to 2021-2030. By deferring a credit, one abstracts oneself from the obligation to repay one’s carbon debt at a given moment, and assumes that one will reduce it later. It would even be possible in some cases to finance credits from one sector with credits from another. For example, it would be possible to finance the credits of the ESR sector with those of the ETS or LULUCF sector. In principle, there might be no problem with this, except that some of these sectors are or have been heavily over-credited (and not necessarily for the right reasons, as we will discuss later), which means that it is quite easy to move credits between sectors and not engage in a low carbon transition. Each trading scheme manages how credits can be used. CERs (CDM projects) or ERUs (JI projects) can, for example, be used to meet part of the obligations that countries may have under a trading scheme (an AAU will then have to be cancelled to generate an ERU so as not to double-count emissions reductions). In addition, it would appear that the European Commission is setting emission reduction starting points theoretically higher than they actually are, making it all the easier to meet targets.

The price of carbon credits on the bond market has fluctuated quite significantly (see Figure 8), for several reasons. The first is that a number of emission allowances in the ETS were allocated for free. If you look at it that way, you might think it’s a scandal, because if you sell allowances for free, how can you expect companies to reduce their emissions on their own? In Europe, the explanation lay in the fact that credits allocated free of charge to European companies could avoid penalising European industry in relation to its competitors, and avoid carbon leakage. The second important reason was that on the one hand the initial reduction targets of the Kyoto Protocol were quite low and therefore easily achievable, and on the other hand the economic crisis of 2008-2009 naturally led to a decrease in carbon emissions in a number of countries. As a result, companies had a lot of allowances to sell and the price of allowances collapsed, simply by following the principle of supply and demand. The price of these carbon allowances fell to a few euros, and remained very low until recently (Figure 8). To raise the price of carbon, two mechanisms were put in place a few years ago: a mechanism for purchasing allowances at auction and a market stability reserve mechanism. The reserve mechanism makes it possible to remove excess allowances from the market (and artificially reduce the supply of credits, thus raising prices), and to potentially put them back on the market later. Could we also imagine a floor price for carbon to avoid these pitfalls?

Figure 8. Carbon price changes in the EU-ETS market. EUA stands for European Allowances

The governance of the ETS as presented above has been established until 2018. The fourth phase of the ETS (2021-2030) is expected to initiate a number of changes. The cap on distributed allowances will continue to fall, but at a greater rate than before (from 1.74% per year to 2.2%). The allowances allocated so far for free will continue to be allocated over this new ten-year period, mainly to those sectors most exposed to the risk of relocating their production outside the European Union. Sectors that are less exposed to the risk of relocation will see their free allowances decrease, until they are definitively eliminated in 2030, at the end of phase 4 of the ETS. In view of these fairly strong constraints on the reduction of free allocations of allowances since the Paris Agreement, the investment kinetics are weaker than this regulation. The demand for allowances becomes high and as the supply is low, the price of carbon rises – still following the classic paradigm of supply and demand. In the bond market, the tonne of carbon is currently around 60 euros. Again, keep in mind that the agriculture sector is not included in the ETS, the objective being to force companies under quota to reduce by themselves by investing on their carbon impact.

The CDM projects should have seen their days over at the beginning of 2021, i.e. at the end of the second application period of the Kyoto Protocol. I remind you that the Paris Agreement is supposed to have taken over from the Kyoto Protocol on 1 January 2021. It was initially decided that CDM projects would be transformed into SDM projects [Sustainable Development Mechanisms) (in general, when you start hearing about sustainable development, it doesn’t smell very good…). So a few questions arose: What are we going to do with the CDM projects already underway? Should they be maintained or cancelled? How to transfer credits from CDM projects to SDM projects? How do we properly account for the credits to ensure that they are not counted twice, both for the country financing the project and the country implementing the project at home? Or in other words, how to account for CERs. It is important to understand that before the Paris Agreement, under the Kyoto Protocol, some countries had committed to reducing their emissions (Annex I countries) while other so-called non-industrialised countries (non-Annex I countries) had not. CDM projects then allowed Annex I countries to finance projects in non-Annex I countries. With the Paris Agreement, as all countries have committed to emission reductions, the principle of CDM projects no longer makes sense. The last COPs (Nos. 24 and 25) did not lead to a clear consensus on what should be done with these CDM projects. COP26 is expected to resolve this issue. A provisional consensus has nevertheless been reached: CDM project applications from 1 January 2021 onwards will still be registered but their approval will be provisional. They will only be definitively accepted if, at the end of COP-26, it is decided that the CDM mechanism is maintained.

The voluntary markets under the Kyoto Protocol

Quite intuitively, voluntary markets are different from bond markets in that countries or companies commit to emissions reductions without being bound by a particular regulatory framework. These markets therefore allow those who are not subject to the emissions trading scheme to participate in carbon trading as well. However, the total credits generated in voluntary markets remain much lower than those generated in regulated markets.

A number of voluntary certification frameworks – such as the Label Bas Carbone – are being set up. I would like to stress that voluntary certification frameworks should not be confused with a taxation system, they are two different things (we will discuss this later). In the framework of voluntary certification, credits are generated by mutual agreement and a relationship is built between an investor (the one who buys carbon credits), a project leader or agent (the one who organises, facilitates and coordinates a low-carbon project) and an implementer (the one who sets up the project, often an agricultural player in our case).

Two examples will perhaps be more telling:

  • The company Soil Capital positions itself as a project developer, between farmers and buyers of carbon credits. Soil Capital approaches farmers, calculates the potential for carbon credits using the Cool Farm Tool, certifies the credits with its partner South Pole (an expert in carbon remuneration), which then sells the credits generated to private companies that want to buy carbon credits. Soil Capital buys the credits at €27 (farmers are therefore paid €27 per tonne of carbon) and sells them, via its partner South Pole, at around €40 to credit buyers.
  • The France Carbon Agri association aims to facilitate the implementation of the Carbon Agri method. The association organises calls for projects so that producer groups (producer associations, cooperatives, etc.) can support producer projects during the period of labelling. France Carbon Agri trains and supports the use of the Carbon Agri method and the Cap’2ER tool, and ensures the certification of projects via the certifier Bureau Véritas. France Carbon Agri also sells the carbon credits generated. These credits are currently sold at €38 per tonne of carbon, of which €30 goes to the farmers.

Is it then possible to create fungibility or a strong link between the bond and voluntary markets? This is obviously under consideration. The problem of double counting is still prevalent. Let’s take the example of starch producers or maltsters who are under quota because of their energy-intensive factories, but who are also supplied by the agricultural sector. Their impact therefore also comes from agriculture (and this is notably the case for most of their Scope 3). Nevertheless, one may wonder whether opening up bond credits to the voluntary market will not add significant complexity to the systems in place, and make it even more difficult to set up projects and/or access carbon credits. As we shall see shortly, there are many possible financing mechanisms within the framework of the voluntary market, and these mechanisms may be sufficient in themselves.

How are carbon markets evolving in the wake of COP26?

COP26 in Glasgow took place in October 2021. It was an opportunity to discuss Article 6 of the Paris Agreements, which mainly deals with the way in which emission reductions will be accounted for (the establishment of carbon credit markets and trading mechanisms)

In order not to clutter up this already long blog post, I have written a special one on the feedback from COP 26. I invite you to read this complementing dossier

Ancillary financing methods

The voluntary certification framework is interesting, of course, but it must be borne in mind that it is an independent auditor who will generate credits and then sell them, and this within a relatively long period of time, 5 years or more. We can therefore imagine many other mechanisms in place to avoid putting everything on the farmer.


One of the first things we can think of is the tax! There are so many different ones, why not put one on carbon? And we could, for example, use the money from this tax to finance projects with a high environmental impact. Today, there is a tax on petroleum products and agriculture is exempt. It was proposed very recently to raise this tax on fuels, and everyone knows what happened in France… Little hint, it’s yellow. The fuel tax was subsequently frozen at €44. It seems for the moment that we can forget about the solution on taxation

Aid – State intervention

The state can also intervene directly. This was recently the case with the introduction of the “Carbon Diagnosis vouchers” in 2020 in France. The approach was aimed directly at recently installed farmers with the objective of carrying out a greenhouse gas assessment on the farm and then defining an action plan to guide them towards levers for improving the farm’s carbon performance. With €10 million on the table, the ambition is to carry out around 5,000 farm carbon diagnoses – at a cost of €2,000 per farm (i.e. 7% of the total number of farms owned by recently settled farmers). The process is supervised and accompanied in the field – farmers are not left completely on their own to carry out this diagnosis. These good carbon diagnostics are not part of a sales logic but rather a transition support logic. It remains to be seen how this dynamic can be channelled (and continued over time) and how these schemes can be opened up to as many people as possible. The diagnosis is not seen as an end in itself but as the beginning of something bigger, of a profound transition. Nevertheless, one could question the state’s ambition. Ten million euros is certainly a lot of money, but the number of people reached is still quite low.

Should the decarbonisation of agriculture be encouraged or made compulsory? The question remains open. One could argue that the state should be there to create incentives and thus encourage farmers to integrate GHG reduction programmes. A good way to do this is to finance diagnoses, in the same way as it could be done for private individuals. The state could also defiscalise a certain number of things on carbon transactions (no VAT, transactions deductible from taxable income for farmers and possibly tax deductible for buyers). All of this could have an important catalytic effect.

Channel premiums

The principle of a chain premium is also quite attractive insofar as downstream players have a vested interest in financing the decarbonisation of the farmer, and the chains are obviously also seeking to reduce their carbon footprint. There are already industry premiums for a number of crops, such as rapeseed and sunflower. Some premiums are also redistributed to make biofuel (there are also volume incorporation constraints in fossil fuel), for example insofar as several European directives encourage the introduction of rapeseed, oilseeds or sunflower to make biodiesel. The way in which the premium is offered should also be considered. We could, for example, set up a premium for plant cover or a carbon premium considering that the carbon stored is proportional to the cover introduced. We would then move from a practice premium to a low-carbon premium, which could be totally different for communication actions, and also for the associated remuneration, especially if the price of carbon increases. The nuance is fine, but the devil is always in the details. Two examples:

  • The company Nataïs has replaced an industry premium on plant cover with a carbon premium. The credits generated are not yet sold on the voluntary market because they are not certified.
  • The company Saipol organises its sourcing of rapeseed and sunflower seeds with greenhouse gas reduction properties that are used in the form of biofuels. Saipol buys the seed from an intermediary (cooperatives, traders) for the most part (80-90%) or from a farmer to crush it. The company has set up its Oleoze offer. A bonus for greenhouse gases is paid to the farmer according to the results of an internal Saipol calculator based on a life cycle analysis approach. Oleoze remunerates producers up to €24 per tonne of carbon. The remuneration follows a gradient (from 0 to 24€) according to the practices implemented and the reduction calculations made. The credits are 2BSvs certified, and sold to oil customers (the end users of the credit).

It should be noted that this graduation system is also interesting for companies that do not have enough equity to finance large projects. These companies could put an amount on the table according to what they are prepared to give, and the project leader or farmer opposite could decide to implement a certain number of actions for the price proposed. We have already talked about this, but a recurring problem with chain premiums is again that of double counting – who benefits from the greenhouse gas reduction. Under the LBC, the methodology for biofuels is not yet out, but the LBC accepts for the time being the cumulation of payments at the farmer level.

Product Premiums

Contrary to the commodity chain premium, which remunerates farmers for a crop/product within a specific commodity chain, the product premium would remunerate the farmer for the simple fact of implementing this crop/product. This is not yet widely discussed. Disconnecting the remuneration from the product may also allow a more global vision of the farm and support changes in practices. Implementing actions to reduce a product’s emissions is a good thing in itself, but it depends on what is done with it afterwards. If we reduce the emissions of a product in order to make even more, does it really make sense?

Environmental payments

When we think of carbon, we think of the environment and we think that there are already payments in this area: HVE (High Environmental Value), MAEC (Agri-Environmental and Climate Measures), CAP payments, organic farming aid…. Could we then imagine playing on the complementarity of these aids to combine and/or merge them? The question that arises is whether or not these aids already take carbon into account, at the risk of contradicting the principle of additionality of the carbon market that was discussed. If, for example, MAECs are targeted at practices already covered by the LBC, it must be shown that this aid is not sufficient to trigger these practices and/or that the MAEC will make it possible to go further than the LBC taken in isolation. In Brittany, for example, there are soil MAECs where farmers implement practices that are fairly close to soil conservation agriculture and are paid to follow these specifications. A certain permeability is being established. However, is a product or farm label as valuable as a voluntary carbon certification framework? It really depends on the confidence that credit buyers will put in these labels. We have discussed this, but there are big debates going on around the new CAP about how future eco-regimes will be structured. But can carbon credits really be awarded to an HVE or organic certified farm? The framework of practices is totally different. Will we favour easier access to carbon aid for agriculture, regardless of the capacity to reduce emissions or store carbon in the soil?


Given the current price of carbon on the voluntary market in France (around €40-50 per tonne in mid-2021), we can wonder whether private companies will be prepared to pay the price. There are serious doubts and concerns that the voluntary market will not be able to absorb the carbon credits issued (with the risk that the price of carbon will collapse again if supply exceeds demand). There may therefore be a need for co-financing for the time being, both between different mechanisms (those seen above) and also different financiers. The diversification of funding sources would effectively allow a low-carbon orientation of agriculture to be stepped up. Care must be taken to avoid a situation where, in practice, farmers will have to choose between private and public funding. We must consider territorial financing (payments from the region, taking charge of carbon diagnostics, taking charge of transaction costs, etc.) – which would not only focus on carbon – a local commitment with companies or associations that want to change things, or even sectoral bonuses if the food industry better promotes low-carbon products or if they manage to communicate better with their CSR policy. Over time, some projects could be co-financed by other companies because the initial funder could not take on the project alone (e.g. increasing the number of producers). If a project is implemented on a co-product that is not sufficiently structuring for the company, the search for co-financing could be very relevant to prevent the company from disengaging.

Some additional elements

We are starting to see many players emerge in the carbon sector. This is reassuring because it confirms that there is a market and that it is starting to interest a lot of people. The challenge will be to find a balance between what is solid and credible for the markets and what is practical for farmers. And of course what really makes a difference in terms of greenhouse gases reduced or avoided…. The risk that can only be tested in the field is whether the programme put in place will be sufficiently acceptable to the rest of the stakeholders (not too much administrative and prohibitive burden…) to be recognised over time as something that works.

In order to be able to enter the voluntary carbon market, there is no choice but to have standards and to have the credits certified. Finally, anyone could issue uncertified credits, but there is relatively little chance of anyone buying them. The market is rather a good way to get long-term project finance because it strengthens project implementation and control (or at least gives the impression of it…). Everything that revolves around standards tends to reassure and give confidence: verifiability of credits, independent third-party auditors, transparency of credits, credibility of labels, etc. Carbon should be considered as a lever for financing. In the context of development aid projects, some would even say that this lever would make it possible to change the current logic of being on the lookout for financial backers and always having to respond to ensure an income, by something much more profound, namely becoming a producer of a non-tangible commodity. By becoming a producer of this new commodity, one can then sell this carbon on the long term market (if the market does not fall apart, of course), allowing one to become independent of donors and other sources of funding such as development aid.

Will companies nevertheless want to finance carbon projects upstream? Would they not prefer to wait until they can buy carbon assets once they are certified? The debate remains open… Some territorial actors could take the risk themselves by pre-financing projects or pre-purchasing credits, and sell them once the credits are certified. A territorial actor or a company could say to a farmer: “You commit to introducing such and such a lever, and we consider that you will generate X carbon credits. I will pre-finance Y and you will get the rest at the end if you have implemented everything we have committed to”.

Let’s try to take a step back

The subject of carbon in agriculture is clearly a topical one, and farmers are therefore very directly or indirectly involved. But have they really been asked their opinion? Do they feel the need to get involved in this issue? We could start by saying that farmers have a strong need for recognition of the practices they implement. Remuneration is of course important, but recognition of their work in the field, right down to the citizen, is fundamental. The agribashing of recent years must evolve towards a consideration of farmers and positive communication. And carbon could be one of the levers towards this recognition, since carbon storage practices in the soil could show everyone the involvement of farmers in the fight against climate change: “Look, with my farming practices, I have managed to store 50 tonnes of carbon in the soil. And what have you done on your side? “

On the other hand, more and more farmers are finding themselves in agronomic impasses, and carbon, through all the co-benefits it brings to the soil, is a particularly relevant way in. Nevertheless, we must keep our main focus on soil quality and not only on carbon. If we just want to put carbon in the soil, we might as well not bother putting used tyres in it – although everyone will agree that this will not do much for the state of the soil on the planet. Above all, we need to keep a global logic of soil quality and not a purely economic logic as could happen with carbon. Markets and carbon credits are a plus, certainly not negligible, but they will never do everything by themselves. They should be seen as a contribution to a farm transformation project with other objectives. These carbon mechanisms should be put alongside soil quality because it would provide incentives for soil quality, and there will undoubtedly be an indirect effect on carbon. All this will not prevent carbon payments. Just telling a farmer that he has to finance his carbon transition is a hidden way of making him go further into debt or adding capital, and in the best case of all, subsidising him, but it is not a way of giving him economic autonomy. It is necessary to encourage the creation of a turnover from carbon income, but it is also necessary to admit that after a certain time, the farmer will continue the practices he has put in place, even without carbon income (insofar as we may end up contradicting the principle of additionality of aid). The carbon income is simply the icing on the cake.

Entering through the environmental component will then make it possible to support strong agronomic components and important economic components. The low-carbon transition is nevertheless very technical and will require a lot of human support for farmers. Structures that have acquired a great deal of knowledge on the subject will have to redouble their efforts in the field. How then can we ensure that farmers really feel involved in emission reduction or carbon storage projects? And how can we get farmers to change or transform their practices in some way, and not just optimise some of their practices, which are not necessarily the most relevant for engaging low-carbon levers? The debate is quite open between mechanisms favouring an obligation of means (we judge the practices in place) or an obligation of results (we look at what has been done in the end). Those in favour of the obligation of results would tend to say that this mechanism encourages innovative practices, initiatives and creativity. Farmers invent and need to share. The farmer would become a real actor of his system and, after careful consideration, could apprehend and imagine new practices to be put in place. This would also avoid the risk of deadweight loss, guarantee the effectiveness of funding and avoid having to deal with regulations that are not always well thought out (for example, not planting a cover crop before or during a rainfall just to comply with a regulation). For the supporters of the obligation of results, there are in any case no specifications of means that allow to achieve the result because all farms are different. Supporters of the obligation of means will tend to say that monitoring recommended practices is much simpler than monitoring results, and that this mechanism allows for the widest possible dissemination of the practices deemed most relevant. A recent report by I4CE nevertheless seems to show that the obligation of results would not necessarily be more costly to investigate and control than an obligation of means (the distribution of costs is however quite different between the costs of design, administrative operation, monitoring and verification). According to the same report, it is not so much the debate between obligations of means and results that should be addressed, but rather the ambition of the proposed measures.

The rigour of carbon monitoring tools and methods are fundamental, that is certain. But we must agree on the fact that no project will see the light of day if the systems to be put in place are unbearably cumbersome and if the tools proposed are not at least ergonomic. What about the complexity or heaviness of putting together a file to have a project labelled? What about the time spent on the initial diagnosis of the project? It should not and cannot be a gas factory. How can the proposed programmes be simplified? Perhaps by proposing certain pre-organised frameworks. Could we imagine a somewhat dirigiste model where the project leader would propose, for example, 3 practices to be implemented, which the farmer would sign or not sign? Or propose 2-3 different projects (payment date, price, funders, data ownership, etc.) to farmers and ask them to choose their favourite? Or should we push for a more open model where all the low-carbon levers are presented on the farm and each agri makes a bit of a choice? Or perhaps we should turn to the system of premiums (e.g. sectoral premiums) which are more flexible to implement?

The costs and risks of a carbon project

Is a carbon project expensive to implement? This is not an easy question to answer. Perhaps the most important sticking point at present is that the economic model associated with carbon markets is not yet stabilised. The majority of the current market is concentrated around the carbon bond market. On the voluntary market, prices are very volatile, and the price ranges, depending on the country and the project, are quite wide. In France, on the voluntary market, there are certainly local projects to be financed with credit values that can be around €40-50 per tonne of carbon, but this is still perhaps a niche market. How deep is this market really? Is it really possible to deploy large projects over large areas? If we believe in the carbon transition that is taking place and if we also believe in voluntary compensation, prices will rise. And it will then be very important to organise, administer, and set up a carbon market with a utility. Farmers may still want to wait a little while for prices to rise, or for more companies to start engaging with carbon. But by waiting too long, is there not a risk that the practices in place will no longer be considered as additional and thus contradict the 3 main principles of carbon remuneration mentioned above?

In the meantime, for the farmer, the commitment to a complete decarbonised transition is quite expensive. We can think of it in terms of economic cost per tonne of carbon stored (see Table 1, and review Figures 4 and 5). All the techniques that can be put in place are risk-taking, and farmers do not want to take these risks alone, especially since some certification frameworks generate credits after 5 years of the project. And this is all the more the case for farmers who are in a rather difficult economic situation, who have to make a transition, and who cannot necessarily take this level of financial risk. Some very virtuous practices are going to be very expensive – one can think, for example, of the establishment of intermediate grasslands during which there would be no cereal biomass, or the testing of a new cover. Generally speaking, virtuous practices are known – except for some slightly innovative ones – but these practices are not always implemented, either because the investment cost is too high or because farmers feel that there is a significant risk associated with this change in practice. What is important is to organise and prioritise the levers to accelerate the increase of organic matter in the soil. Farmers should also have the choice to go at their own pace according to their level of risk.

These farmers will need to be economically supported and all the funding mechanisms presented above will need to be intelligently articulated and their implementation will need to remain flexible. The example of the recovery plan proposed by the government is interesting, but the Carbon Diagnostic Vouchers are only aimed at young farmers, and the total amount allocated to projects is still quite low given the proportion of French farmers. The fundamental question to be asked is whether, in the end, the game of being labelled by an existing mechanism and/or setting up carbon projects will make it possible to compensate for the additional costs of implementing low-carbon actions for the farmer.

Table 1. Summary of costs per practice, on average at national level, depending on whether the storage horizon considered is 0-30 cm or the entire profile. Source: INRAE, 2019

And this risk can materialise in different forms. If prices per tonne of carbon increase, will the prices be passed on to the farmer? If the project holder cannot fulfil his part of the contract due to events beyond his control (e.g. fires, crop losses involving the import of feed, loss of hedgerows, etc.), the consequences will depend on the terms of the contract he has signed. Let’s also imagine that a farmer has signed a certificate and committed himself to a carbon market. If tomorrow, more serious or restrictive rules are put in place, will the farmer be able to sell his carbon elsewhere when he has already sold it? The sources of financial uncertainty for farmers are both in the amount of revenue farmers will get – revenue that depends on the price of the tonne of carbon – but also in the fact that their actual payment is uncertain (a buyer could very well make a mistake). With a carbon market that looks so attractive, we should see new players arrive quickly – and this is already quite the case – some more opportunistic than others. Projects will come in with potentially very different contracts, and scams will never be far away. Farmers will have to be careful and aware of their payment terms. As a farmer, make sure that the players who approach you with carbon calculation or remuneration tools have them.

Not all farmers have the same chances to store organic matter. Could we go so far as to say that this is a bit unfair? The results are quite variable between farms. In 10 years, some farmers will be able to gain several points of organic matter under certain conditions, but not everyone will. And the comparison of these conditions is not completely obvious either. The 4/1000 initiative launched in France by Stéphane le Foll aimed at a global annual storage target, which is difficult to individualise for each soil – some soils will store more, others less. Selling a tonne of carbon when you have a very poor soil thanks to additional storage is definitely not the same thing as when you already have a very carbon-rich soil (we mentioned this point earlier in the context of the low carbon label). The ton is also more expensive in temperate countries because the storage capacity is less. Current conditions make it more financially logical to set up carbon projects in tropical areas (for example in forestry projects) simply because the growth is faster and the carbon will be stored more quickly. However, more and more French companies are finding it worthwhile to develop carbon in France because they want to make a commitment to their territory and give more meaning to their commitment. Perhaps this is a reaction to the Covid-19, which has made some companies feel the need to favour more local projects, who knows; we’re not going to miss out. As incredible as it may seem, French farmers are helping to feed the French… It is obviously also a way for these companies to give a little boost to their communication, by showing their involvement in their territory. Perhaps it will be possible to differentiate the carbon remuneration according to the difficulty of the levers to be implemented in certain pedoclimatic contexts…

There is a risk that the majority of the added value will not go to the farmer. When we see the distribution of value in the current food distribution chains, there is indeed cause for concern. For carbon, the costs are multiple: between diagnosis, verification, auditing, or intermediaries; who will really benefit from carbon markets? Getting the prices right and allocating the value of the tonne of carbon will be extremely important. However, some intermediaries may benefit farmers because not all project developers have the capacity and/or time to sell carbon credits on the market. Buyers or sellers of carbon credits – whether they are aggregators or intermediaries – can facilitate access to carbon credits for farmers by aggregating projects. Marketplaces offer additional verification of ongoing projects and can provide more transparency to customers, make it easier to understand projects and make the market more fluid. What is more problematic, however, is when multi-buying/selling of carbon credits is set up (there is talk of “brokers” who would be there to carry out high-frequency carbon trading for example). In this case, we can be sure that the intermediation costs will artificially inflate the prices of carbon projects and that little money will ultimately reach the farmers. At the national level, there will also be work to be done to align the work of intermediaries, in particular so that the prices offered do not vary by double between intermediaries. The choice of one or more intermediaries for the farmer (or group of farmers) should remain free.

So far we have talked about farmers, but the price of a tonne of carbon can also be a barrier for companies that finance credits. First of all, there is the initial cost of the initial diagnosis of the project, which could be paid for by the company, but also the audits carried out. As for the certification itself, the budgets can be substantial for international labels insofar as an audit must be carried out every five years. For projects that are very structuring for a company internally – within its distribution chain – it is understandable that the company is ready to assume this cost. On the other hand, on small-scale carbon projects and for smaller companies, the cost of certification may be more difficult to swallow. Some players will therefore prefer to turn to labels with a lower budget that do not necessarily generate credits (e.g. Value Chain Initiative). In addition, a company that does not want to sell its carbon credits on the market does not really need to generate carbon credits anyway.

Talking about money associated with carbon can be reassuring for farmers but the purely economic reasoning is still rather limited because the associated services are not monetised. The implementation of virtuous practices is of course of interest for emission reductions and/or carbon storage in the soil but it is far from being the soil. Soil regeneration, reduction of erosion, improvement of the bearing capacity and structure of the soil, reduction of leaching, intellectual stimulation through advanced agronomic thinking, all this is directly or indirectly linked to the implementation of these practices and cannot be summed up in a simple economic value.

How much should the price of a tonne of carbon be set at? 20, 50, 100, 200, 1000? And as we have already said, these costs should also reflect the cost of the farmer’s practices. However, be careful not to confuse the carbon price with the climate action value, which is supposed to help choose the best actions and projects to fight climate change. Under this framework, all projects with emissions costs below the climate action value would have to go ahead in order to stay within the Paris Agreement. Other projects should be left aside. This value, currently around €100 per tonne of carbon, is expected to rise to around €800 by 2050.

From compensation to carbon contribution

Listening to the media and the propaganda of certain large companies, we should be reassured that everyone is committed to carbon neutrality, especially through carbon offsetting. Reforestation projects, rehabilitation of areas, low-carbon projects, it almost brings a tear to the eye. The problem of climate deregulation has obviously not been well understood. Saying that we compensate and that we are carbon neutral when we are not a state does not really make sense. Offsetting is a non-issue and the term itself is not appropriate. And the idea of offsetting can lead to perverse effects and aberrations, such as the cessation of virtuous practices and the release of carbon into the atmosphere in order to be able to restock this carbon afterwards. The examples of reforestation projects are also very telling. These projects only allow carbon to be stored over the long term; the time it takes for the trees to grow, whereas greenhouse gas emissions are often immediate. Carbon storage does not take place at the time of purchase of the credit. To maximise the investment in forest carbon credits, some project developers may want to accelerate production cycles, with fast-growing species, favouring forest monoculture, and potentially exploiting planted forests for commercial purposes. In addition, climate change is a threat to the development of forestry projects. Fires, drought, disease, pests – all threats that will affect forest production and, if they occur, will make it impossible to offset what has been emitted upstream. The carbon balance sheet will then be completely skewed.

The much more accurate term contribution should be preferred. A company is only contributing to the low-carbon transition, it cannot claim to be neutral. One fear, however, is that companies in the market are not yet really aware of this and have not really assimilated it. How will these companies that finance carbon projects react when it is explained to them that they cannot call themselves carbon neutral? Will they continue to invest the same amount of money? Companies are still taking a very accounting approach to the transition and their objective is to put as little money as possible into each project. Nevertheless, perhaps if we tell them that they will never be neutral, some French companies will allocate their budgets to French carbon projects to take an interest in their territory. More generally, the carbon contribution lever should be reserved as a priority for sectors that are important for the common good (food, for example) and for which certain emissions are incompressible.

Surprisingly enough, a project whose emissions increase every year could obtain a low-carbon certification framework if its emissions increase less than they would have in a baseline scenario. In other words, if a carbon project is less worse than it would have been without changing anything, it is already better than nothing. And it also depends on how the carbon calculations are done. Reducing greenhouse gas emissions per litre of milk produced is in itself a good thing, but it becomes meaningless if the total amount of milk produced increases significantly and/or if a farm grows significantly. Wouldn’t this favour the most productive and intensive farms? This is the classic Jevons paradox, more commonly known as the “rebound effect”, theorised by the English economist and logician William Stanley Jevons about the rapidly increasing consumption of coal despite the increasing energy efficiency of steam engines. In the context of field crops in France, the common divisor chosen was the hectare. In some certification frameworks, for carbon storage in soils, a baseline – calculated from a reference scenario – models the evolution of carbon stock in soils if current practices are not changed. A farmer who de-stocks but with a lower de-stocking than under the baseline scenario could also be able to recover carbon credits. This argument is defended by the fact that storage approaches must be considered over a long time. Some carbon de-stocking situations are linked to land use changes that took place several decades ago and that are not always the fault of the farmer. In this context, limiting the loss of carbon would already be a very positive result.

In the same vein, the concept of avoided emissions was discussed in this article. It is likely that more work needs to be done to clearly define its contours. There is still uncertainty as to how much carbon is actually sequestered or avoided. The aim is to ensure that compensation is not paid for avoided emissions that would not actually have occurred. In the context of forestry projects, we can imagine a first simple case of planting where we calculate a number of trees, an expected diameter, and where we estimate a carbon storage knowing that we are planting on areas that would not have been planted. Second case: the defenders of a forestry project announce that if their project does not go ahead, the forest in question will be deforested. In this case, the avoided emissions are the difference between a theoretical deforestation curve that would have occurred without the project and the level of deforestation that would be expected if the project were implemented. We are therefore selling here a difference between two deforestation curves, assuming that the project leads to less deforestation than the baseline situation. The problem lies in the consideration of the baseline situation, often based on historical data. Can we really be sure that the rate of deforestation will remain the same and that the project will really have served a purpose? In other words, it is always possible to adjust the baseline scenario on the assumption that the forest is more threatened than it actually is. In this way, avoided emissions are financed for nothing. As another example, some projects in India and China propose to replace fossil energy sources (especially coal) with renewable energy. The avoided emissions could then be sold on. The problem is that in several cases, organisations have found that there is already a local law requiring this conversion or that existing projects already have the objective of replacing coal with renewable energy. If the avoided emissions are sold, then the principle of additionality is completely violated. Hence the importance of clarifying the concept of avoided emissions as precisely as possible. The last example is agriculture. We have already talked about this, but a farmer who maintains his carbon stock by keeping stocking practices can be remunerated for this, and that is a good thing. But is it normal for a company to buy these credits, or in other words to offset them, knowing that the carbon is already in the soil? Isn’t there a form of schizophrenia?

The contribution of companies to the transition via the existing carbon markets can certainly be seen as an interesting practice, but it should not and in any case cannot be a substitute for reducing its emissions. We simply do not have the time. Setting up carbon projects, whatever one may say, is time-consuming. You have to put together files, get them validated, set up the project, find buyers… Climate change is knocking on our door, and requires us to work on reducing greenhouse gas emissions as a priority. Companies must build a strategy to reduce their polluting emissions. The contribution cannot be sufficient on its own and must be accompanied by a real dynamic. Calculating emissions that are subtracted from contributions is rather limited, it is not enough to go and plant trees. However, some will say that as long as there are ways to offset their footprint, companies will continue to do so before reducing their emissions. The contribution system exists, so we might as well use it to channel the flow of money towards projects that will serve the carbon and much more than that; projects with social, ethical, human development or biodiversity impacts. The challenge will be to get companies to finance carbon projects that are as close as possible to their activities. This is known as “insetting”, or integrating a carbon project into the sector, making it a project that is close to the company’s issues. A food company that sets up a project in its own distribution chain has less and less interest in greenwashing but rather in considering its producers better (better traceability, important environmental dimension, etc.). Climate deregulation will create real supply chain risks for certain food companies and these companies will have to tackle the problem head on. And a company that contributes to the decarbonised transition by buying carbon credits certainly does not want NGOs or other organisations to question the carbon certificates that this company may have bought. These companies want to be unassailable.

The price of carbon credits can also play an important role in the balance of greenhouse gas emission reductions. The price goes up to reflect the real price of the externality, to compensate the farmer and give them a real incentive to change. For a buyer of carbon credits, the higher the price, the more the purchase looks like a fine. For companies with high emissions in certain sectors of our economy (industry, transport, etc.), this can become very expensive. The high price of the credits will then be an incentive to reduce their emissions. To take the example of France, a majority of the most powerful companies have already made commitments to reduce their carbon emissions and have certainly already made provision for the carbon debt they will have to pay. Everyone is aware that these companies will have to decarbonise. If they are at the front of the train, they will have the benefit of better communication (afterwards, of course, if Scope 3 emissions are not taken into account, it’s still not great…) and the stick won’t be too painful as long as the credits aren’t too expensive to buy. Will all companies voluntarily take this direction? Certainly not, but the notion of the polluter pays is still very present. The EU is creating a taxonomy of investments that are considered sustainable or not. If the EU defines them in a legal way, these taxonomies can act as a strong incentive or even a punishment. Many rating agencies are now proposing socially responsible investment (SRI) criteria that include climate criteria. Even if one can always criticise the European Union’s taxonomy or the rating agencies’ criteria, the carbon issue is clearly changing.

Project owners will also be able to play a role in the balance, as they will be able to choose to whom they sell the carbon credits they have helped to generate. The carbon project holders – farmers or collectives – will then be in a strong position and will be able to direct the credits to buyers they consider responsible and well-intentioned. This could be an opportunity for them to ensure that the buyers also implement actions to reduce greenhouse gas emissions and that they participate in a general improvement of the agricultural sector. Farmers must seize this position of strength and this freedom to sell carbon credits, and take advantage of this position to resist the siren calls of very large structures (organisations, cooperatives, etc.) that could lure them with a carbon payment or a chain that would cause them to lose their rightful share of the added value.

It will be necessary to ensure that the projects set up are long-term, that they are sustainable over time. Or at least that the duration of the accounting of emission reductions is longer than the validity of the project, which is not really the case at the moment. If the carbon is destocked very quickly afterwards, the purchase of a credit will really only be to ease one’s conscience…

Carbon at national and international levels

In parallel with the social dumping that can be observed in Europe – both in agriculture but not only – could we see the appearance of climate dumping on the part of certain countries which, voluntarily or involuntarily, are moving more slowly than others towards a less carbon-intensive agriculture that stores more carbon in its soil. This is, in any case, what some stakeholders, including Copa-Cogeca, are concerned about. As we have seen, certain virtuous practices are expensive and sometimes risky from several points of view for the farmer who wants to implement them on his farm. One could oppose this argument by considering that if these practices are indeed risky and time-consuming, they have the great merit of reflecting on the problem over a long period of time and of getting out of the short-termist logic in which we are all stuck. That said, Copa-Cogeca and other organisations are calling for a Carbon Border Adjustment Mechanism (CBAM). The objective of this mechanism would be to ensure that the cost of imported products on the European market reflects the extent of their carbon emissions. The argument put forward seems relevant, but it would be necessary for all countries in Europe to move in the same direction.

At present, the price of carbon credits on the voluntary market is much lower internationally than in France. From a purely economic point of view, how can we fight when the price of a carbon credit internationally ranges from a few euros to several dozen (and this is still not enough) on French territory? This is why the vast majority of carbon projects are carried out in southern countries. On the voluntary market, there is nothing to prevent a company from having several labels for different credits, and an international company can perfectly well use the low-carbon label framework for a carbon project on French territory. Carbon credits, whether they are stamped with the Low Carbon Label, Gold Standard or Verra VCS, all have the same value, i.e. the credibility that a buyer will have in a given credit. We have already talked about this a little, but for French companies, the challenge will be to ensure that these companies do not buy too much credit internationally. We will have to show them the advantage of financing local projects, whether for reasons of distribution chain, communication, value or commitment. We need to create interest and willingness for territoriality. On the other hand, is it not defining for farmers to be locally attached?

Will a single carbon price be set for everyone? The great advantage of a single carbon price would be that it would be technologically and politically neutral, which could avoid wasting public funds on subsidising a particular sector. This is doubtful, however, given the failure to develop common financial and taxation indices across Europe. Having said that, the G7 has recently set up a common tax rate for large multinationals… One can dream.

What about the diversity of carbon monitoring tools and methods

Carbon calculation tools, whether for emissions reductions or storage, methods and certification frameworks are beginning to be numerous, some of which we have reviewed in the course of this blog post. And they continue to develop. In the context of the low-carbon label, for example, several methods are under construction: viticulture, methanisation, perfume plants… What do we think of the proliferation of these methods? Opinions differ.

In view of the heterogeneity of soil and climate contexts and national and international production, one might start by thinking that it is complicated to have very strict methodologies set in stone. The productions are different and the associated problems are not at all the same. Instead, we should not focus on one method, but remain open and flexible. Having a maximum number of methodologies and tools would make it possible to deal with specific subjects. And in order to be precise in the calculations and carbon modelling carried out, it would be necessary to be able to choose the most appropriate tool for each subject. One could very well imagine a tool for emission reductions, a tool for carbon storage, and perhaps even use a different tool between questions relating to compost and others relating to plant cover. Keeping an open mind on existing methods and tools would also make it possible to adapt to various logics of obligations of means or results. If measurement systems develop – or are proposed as a replacement for existing measures – and make it possible to measure soil carbon levels in greater detail, why not take advantage of them? And in this context, a logic of results will perhaps be more appropriate than a logic of means.

How then to juggle between these different tools and methods, and above all how to compare them when the results are significantly different (some results may be 2 to 3 times better than others depending on the tools used – who is right or wrong then, it may be both)? Would we be able to retro-pedal or retro-pay if we realise that a tool is no longer suitable? And how should the associated credits be managed? Will the initial economic model have to be completely revised? In the context of the low-carbon label, we raised the subject of discounts associated with the uncertainty of input data. At national and international levels, standards have different ways of applying discounts, and this freedom is quite criticised. This is all the more the case when the logic of obligations of means or obligations of results is quite different. The aim of a label is to develop measurement methods. But the way in which the measurement is carried out and the accuracy of the measurement can have a strong influence on the price of carbon. Each method specifies how it will evolve and track things. This brings us back to the Tier 1, Tier 2, and Tier 3 taxonomies (see above) – some based on fairly simple approaches to emission factors, others looking at more complex mechanisms with dynamic models. The simplicity of the methods is not a bad thing in itself, but the validity of the results of a simple method cannot be put on the same level as more robust methods. In order to compare methods, it will be necessary to be able to compare assumptions (which must be very clear) and concrete results. For example, under the Label Bas Carbone, an automatic discount of 10% is set to take into account the risk of non-permanence of carbon – because there may be a fire, or a climate risk. But this discount can be different depending on the country. For example, it is 20% in Australia and varies between 7.5 and 15% in Canada. As the voluntary carbon market in agriculture is over-the-counter, the price of carbon credits can be more easily negotiated if the buyer has confidence in the credits he is buying. One of the big advantages of the performance obligation framework is that you have to go and re-measure the soil to see if any carbon is stored. This is perhaps a more expensive approach, but it could encourage greater precision. If no obligation of results is required, will project developers really try to verify that their emissions and/or storage have really evolved? Nothing is less certain…

In France, for example, one wonders whether Soil Capital will comply with the Label Bas Carbone framework and review its approach to the territory. If not, then there will be different methods and the project owner will be able to choose his preferred methodology. The main thing is perhaps that farmers find a remuneration that suits them, that the methodology can reassure the financiers. We would still like to measure changes in emissions and stocks that are close to reality…

Nevertheless, it is difficult not to see behind this diversity an increasingly blurred message. A buyer wishing to buy credits will certainly find it difficult to find his way. And this buyer will perhaps tend to choose a method that suits him or her or a method that is more advantageous for him or her, and not necessarily a very scientific approach or one with a sufficient level of precision on carbon stock quantifications. It could also be argued that by trying to develop more and more specific methods, we could lose the overall vision of the operation. Building too many sectoral methods could have consequences on farms or even transfer of impact – we would still end up with major problems on the principle of double counting carbon. Sectoral methods may be necessary, but there is a need to maintain consistency at farm level, and perhaps use aggregative methodologies for this. The methods must be placed in a more global context.

A lack of transparency and openness?

In the French context, anyone wishing to develop their own carbon certification methodology can contact the government and ask to have their method labelled. Officially, the carbon file therefore seems relatively open. Unofficially, however, there may be some reason to question this openness. In view of the time spent by the current scientific committees to propose and validate the framework of the Label Bas Carbone, one might wonder whether the ministry would really accept a new methodology. It could be argued that developing a method on an existing perimeter is cumbersome and redundant, and that it would be preferable, if we really want to produce a new method, that it should rather focus on a new perimeter that has not yet been validated. However, it could be argued that the current scientific committees do not represent all the farmers’ groups either, and that the fact that the Ministry has framed a method would tend to lock in the existing process.

Officially still, it is normally not possible to start developing a tool until the associated method has been validated by the Ministry. However, this is not really what is happening at the moment. Agrosolutions participated in the drafting of the Label Bas Carbone method for field crops, and has already developed its Carbon Track tool integrating the equations and approaches described in the Label Bas Carbone method. The tool will therefore be one of the pioneers on the market, and some other players may be a little late. Insider trading? Political stakes? But is it really a problem in the end? Not necessarily if we want tools to be available quickly. And Carbon Track will not be the only tool available on the market. Nevertheless, some transparency is needed and smaller players wishing to develop their own tool will certainly face stiff competition. For those who are not Invivo and FNSEA big fans, they will need to have strong backs. It should be noted that this has already happened in the past, with Idele having developed both the Carbon Agri method and the CAP’2ER tool.

Still on the subject of transparency, it will become important to ensure the traceability of contributions to emission reductions and carbon credits. And it is surely the state, via the Ministry of Ecological Transition and Solidarity (MNTS), that should take charge of this. Some proposals are in favour of a centralised and public register to frame the terms of the contract. Monitoring and quality assurance of credits seems fundamental, and all this should also be required by the farmers themselves. The credits must have been generated in serious settings and following recognised protocols.

The estimation of stock levels and changes in soil carbon stocks is complex. This complexity naturally makes the estimates uncertain. Strangely, these uncertainties are almost never displayed in the results of models or in the tools made available on the market. Yet it should be normal for a project developer, a credit buyer, and especially a farmer, to evaluate the margins of uncertainty that he would have in implementing his emission reduction or carbon storage practices.

What about digital technology?

Could IT tools, and perhaps digital technology in a broader sense, accompany the carbon transition that is taking place? Some actors would tend to say yes. To carry out a carbon diagnosis and model the evolution of carbon on a farm with current means, it would be necessary to collect several thousand data – both on the farm itself, the cultivation practices used, the interventions, and the production results. This is not insignificant. However, one might ask how much of this data is needed to assess the state of stocks and futures on the farm, or in other words, to find a compromise between the quantity of data collected and the quality of the modelling. No one wants to spend their life entering their data into a modelling tool. This is to some extent what the Label Bas Carbone framework already proposes with the discounts granted, depending on the uncertainty and quality of the data input to the modelling (see above). These tools must be ergonomic and easy to use if they are to be deployed on a large scale. The carbon pay experience should be made as easy and enjoyable as possible. Let’s keep in mind that the data collected could be used for other things anyway (other sources of remuneration, CAP declarations…). But these general observations are after all not specific to the carbon topic, but to all IT projects in the broadest sense.

All this being said, the greatest complexity lies in the recovery of data. Data entry will certainly be more straightforward for a farmer who has had computerised traceability in place for some time, as many players are developing APIs (Application Programming Interfaces) so that calculation and modelling tools can interface with land management software. This is the case, for example, with the Simeos-AMG tool, which is in the process of being converted to an API, but which will also keep its current interface. In animal husbandry, we can also think of dairy farms, monitored by the milk control, and for which a certain amount of data is standardised and easily retrievable. These interoperability issues are at the heart of many debates in the field of agriculture, as we have already discussed in a previous blog post. The connection between heterogeneous data sources – whether they come from the farmer, field measurements, management software or satellite data – is the sine qua non condition for existing digital tools to deliver accurate indicators. It should also be borne in mind that, at present, relatively few farmers are equipped with land management software. A 2017 study by the Observatory of Digital Uses in Agriculture gives a figure of 25%. And I would even add that the data entered in this software is often of a regulatory nature and has little agronomic value (sowing density, yield, etc.). There are two important observations to be made. Firstly, there is very little data in this software that could really be used for carbon modelling tools. Secondly, and this is not completely independent, it could be difficult for a farmer who commits to a carbon project to carry out carbon simulations on practices prior to his carbon project, as the data would not necessarily be available. Despite ongoing efforts on the subject, it is clear that the interface between existing tools is far from being resolved. For the farmer who keeps paper records, data entry will certainly be a little more laborious – but not impossible insofar as his records are organised and prepared for diagnosis or modelling. For arable and beef farms, some have not seen advisors for a long time. A large proportion of producers do not have technical support and data collection may therefore take much longer. Some tools, such as Carbon Track from AgroSolutions, will offer quick input levels where, in less than an hour, a farmer can already fill in the main characteristics of his farm to set up an initial carbon status.

Carbon is undoubtedly the subject of the moment; calculation tools and methods are developing quite widely in France and internationally. From a logistical point of view, this proliferation of tools and methods makes it easy to juggle between calculation tools by transferring data from one to the other according to the carbon projects in progress. In the future, we can expect the methods to become increasingly complex. The discovery or updating of agronomic processes, the updating of biomass estimation models – these are the developments that can be expected. But digital tools, on the contrary, could be there to simplify everything. The farmer will be able to understand the functional and agronomic complexity of the method, but its technical complexity will be deviated from the computer tools. It should be noted, however, that the tools are not yet capable of taking into account certain specificities of cultivation techniques or certain innovative techniques (double cropping, multi-annual permanent cover, highly diversified cover). They certainly reflect models that perhaps do not consider them, but the complexity of agronomy is not yet fully integrated. This rise in complexity could be seen from rather diametrically opposed points of view. On the one hand, this complexity is not a problem in itself because the level of complexity represents the heterogeneity of France. Each territory, each farmer, each plot of land, each soil is different. A general methodology can never be completely adapted to a local context. On the other hand, this complexity could scare off some collectives or associations that would like to develop their own IT tool to remain independent from the main market players. The deployment of APIs should rather be seen as reassuring since it will make it possible to avoid intermediaries and to connect directly to the raw calculation tools. Finally, it could be added that as the complexity increases, it will become increasingly complicated to verify the quality of the monitoring indicators obtained. Some indicators may depend on the production objectives of the farm and some input data may be subject to interpretation. Add to this potential problems or errors in the transfer or updating of data, and you could get lost in the maze of calculations and modelling, and find it difficult to get your head above water.

Digital tools are also expected to serve as a major cost driver for carbon projects. The new CAP should come with strong demands on MRV (Monitoring, Reporting, Verification) systems to ensure that the credits granted – for carbon but not only – go to the right place. One example is Cesbio’s SEN4CAP chain for monitoring agricultural practices and verifying CAP declarations. This processing chain already offers maps of leaf and vegetation indices (LAI, NDVI, FCover, etc.) and indicators for monitoring nitrate leaching, biodiversity and the albedo of agricultural land. This demand for monitoring will unquestionably require very high project management costs. Costs that digital tools could help to reduce insofar as we are dealing with the management of agricultural practices, with an obligation of means. The data on practices must be traced in order to make the link between practices and carbon density. Digital tools would facilitate the acquisition and transfer of data, and would make the current processes more fluid. Spending several days setting up a carbon diagnosis or modelling, between data collection, formatting or even the simulation of different scenarios), can weigh down a carbon project economically. Automation and simplification therefore go hand in hand, while ensuring that simplifications do not undermine the robustness of the methods in place.

I won’t go into detail on the subject, but digital tools will always raise issues of data ownership and use, especially on an issue as hot as carbon. Some charters and labels have been set up – for example Data Agri – but some farmers will never be completely reassured. The obligations of means, the repeated monitoring of farms (particularly with the new CAP measures) or the gaps and failures of agronomic practices on the declarations (management of crop residues and organic fertilisers, presence or absence of cover crops, etc.) leave one pensive. A relationship of trust must necessarily be established.

Finally, let’s raise the paradox linked to digital tools that I had already highlighted in a previous blog post. Does developing digital models and tools to limit the carbon footprint of agriculture make sense in view of the carbon impact of these tools themselves? Between data collection, storage, analysis and the development of IT tools, is the balance on the side of agriculture or digital technology? Is it worth it? The answer is not obvious, but we need to think about it.

Some final remarks

In view of the carbon storage capacity of soils, could it be imagined that the carbon dossier would encourage speculation on land? Pressure on land could significantly increase the price of agricultural land and make access to property even more complex than it is at present, especially for new farmers. A point of attention seems appropriate; some associations are calling for an impact study on the subject.

If we consider that photosynthesis is one of the most powerful mechanisms for carbon storage, the question arises as to how the biomass will be used. How can we choose between returning the biomass to the soil as much as possible and using it in a value-adding process such as methanisation? Are the two approaches compatible? Does it depend on the spatial scale considered? One might add that the question of the distribution of the added value of carbon will arise again. Who, the biogas plant manager or the biogas distributor, will have the carbon credits? How will they be distributed? And it should be borne in mind that a farmer could feed a biogas plant just as well with virtuous practices as with practices that emit a lot of carbon or that destroy it.

From an agronomic point of view, carbon is still mainly considered in relation to organic matter in the broad sense. There are several forms of this organic matter (stable, labile, etc.). The technologies for analysing the forms of organic matter are beginning to arrive, but they are not yet in complete agreement with each other. Will the refinement of the forms of organic matter considered be of interest for monitoring the carbon issue in agriculture?

The current issues around carbon will have taken the debate away from a rather sensitive subject in agriculture, glyphosate. Insofar as advocates of conservation agriculture (CA) defend its use at very low levels and we know the potential of CA for storing carbon in the soil, could we go so far as to say that we should treat to store? Let’s remain vigilant, glyphosate remains a problem and we must continue to work on it. But this problem is far from being the worst in view of its use in the context of the conservation agriculture pioneers.

The origin of carbon on farms is still not considered very much. It can be produced on the farm (biomass of plant cover, livestock on the farm, etc.), and is then referred to as endogenous carbon. Or it may be imported onto the farm (livestock farming outside the farm, mineral nitrogen fertilisers, etc.), in which case it is called exogenous carbon. This notion of carbon origin would benefit from being taken into account in emissions and storage calculations because it is intrinsically linked to the issue of resilience and autonomy of the farm.

The generation of carbon credits is quite expensive and we have already talked a bit about it (project design, verification, auditing…). When the price of carbon dropped to its lowest point, it was therefore no longer necessarily attractive for some companies or countries to account for the emission reductions of projects they had set up, otherwise they would have lost money. However, there is a concern that if the carbon price goes up a lot, these companies and countries will bring in auditors to certify the emission reductions that have taken place in the past. In other words, the number of credits available today is much lower than the number of credits that could be available if credit prices were to rise. Given that these emissions have potentially already been avoided and that we are struggling to reduce our carbon footprint, should we not prevent these past credits from being generated? If, for example, credits from old CDM projects (pre-2020) are transferred to the 2021-2030 period, we could end up with a lot of surplus credits from 2021 onwards that will not contribute to any emission reductions after 2020….

In conclusion

One of the great fears of farmers is that they will become the turkeys in this carbon market. And when we see the situation in which some of them find themselves now, we can understand why. Farmers need to take hold of this carbon issue, understand it, integrate it and master it to ensure that they can get the most out of it. There is no doubt that carbon is an important issue. It is already an extremely powerful lever from an agronomic point of view because it brings co-benefits on the whole quality of the soil. It is also a strong lever for communication and recognition, moving from the classic punitive ecology to a positive ecology. Finally, it is a significant means of remuneration, even if we must be clear on the fact that carbon credits will only finance a part of the low-carbon transition, and that as we move forward, the share of the transition financed by carbon credits will be less and less.

To use a phrase that is not mine, not every tree is a business case. Understand that many initiatives will have to be financed independently of a well-defined project. For a company, this also means fewer indicators, fewer KPIs (Key Performance Indexes), and more common good. For a project to be sustainable, it must of course have a quantifiable impact, but for the project to work in the long term, there needs to be something more risky for the company but which brings a lot to the community, a kind of common good which will benefit the project as a whole.

At the moment, the debates around carbon are quite heated. Parochial disputes are still very visible between pioneer farmers, research centres and technical institutes. These communities are still quite isolated and still work relatively little together. There is a time to complain and fight, and there is a time to move and make things happen. The climate does not wait.

Let us conclude by reminding ourselves that although carbon is a structuring issue for the agricultural sector, we must not forget that the main mission of agriculture remains above all to feed the population. But this food can only be produced within the limits that the planet offers us. Our environmental footprint, over-consumption and climate change are eroding all the services agriculture needs to produce. As a greenhouse gas emitting sector, agriculture of course also has a role to play in combating climate change – and agriculture has an even greater interest in doing so as it is directly affected by the consequences of climate change. But agriculture will not be able to work on this alone. The responsibility lies with absolutely everyone: politicians, businesses, and consumers. It is time to wake up…

Soutenez Agriculture et numérique – Blog Aspexit sur Tipeee

Bibliography complementary to the interviews

Ademe (2016). PCAET. Comprendre, Construire et Mettre en œuvre. Plan Climat Air Energie territorial.

Bockstaller et al. (2021). Apports de la télédétection au calcul d’indicateurs agri-environnementaux au service de la PAC, des agriculteurs et porteurs d’enjeu. Innovations Agronomiques, 83, 43-59

Boivin, P. (2021). ACS et teneur en matière organique du sol. Quelques enseignements tirés de la région lémanique. TCS n°111

Bossio et al. (2020). The role of soil carbon in natural climate solutions. Nature Sustainability

Burton, R. J. F., & Schwarz, G. (2013). Result-oriented agri-environmental schemes in Europe and their potential for promoting behavioural change. Land Use Policy, 30(1), 628-641

Carbon Market Watch (2020). Introduction aux marchés du carbone. Un guide des mécanismes mondiaux de compensation.

CCFD-Terre Solidaire, Réseau Action Climat (2020). Positionnement sur le label bas carbone et la méthode pour le secteur agricole

Demaze (2013). Au nom de la lutte contre le changement climatique : le mécanisme pour un développement propre et ses travers. Controverses environnementales : expertise et expertise de l’expertise. Vertigo, vol.13

I4CE (2020). L’obligation de résultats environnementaux verdira-t-elle la PAC ? Comparison des coûts et de l’efficacité de six instruments de transition vers une agriculture durable

INRAE (2019). Stocker du carbone dasn les sols français. Quel potentiel au regard de l’objectif 4 pour 1000 et à quel coût ? Résumé de l’étude Juillet 2019

Terre-Net (2020). Livre Blanc – Stocker le carbone dans les sols

Yogo et al. (2021). Analyse des méthodologies d’évaluation et de suivi du bilan carbone des sols et recommandations pour l’écriture d’une méthode grande culture dans le cadre du label Bas-Carbone :

Vaudour et al. (2021). Temporal mosaicking approaches of Sentinel-2 images for extending topsoil organic carbon content mapping in croplands. International Journal of Applied Earth Observation and Geoinformation, 96, p. 102277

Seminar to watch – review

Agreenium : La captation du carbone dans les sols état des lieux et perspectives :

APAD : De l’Or Au Cœur des Sols ?

Chaire AgroTIC et Chaire Elsa-Pact : Evaluation de l’empreinte carbone en agriculture : quel apport des outils numériques ?

INRAE : Carbon labelling : lessons learned from the French label

Planet A : Carbone 2021 :


Anne-Sophie ALIBERTPUR Projet
Michaela ASCHBACHERCool Farm Tool
Eric CESCHIA & Thierry CHAPUISInrae/Cesbio & CNES
Gabriella CEVALLOSDeloitte – Anciennement chez I4CE
Jean-Baptiste DOLLEFrance Carbon Agri
Annie DUPARQUEAgroTransfert
Emelie HALLESaipol
Sylvain HYPOLITEAgrod’Oc
Stefan JIRKAVerra
Edouard LANCKRIETAgroSolutions
Etienne LAPIERRETerrasolis
Gaspard LESCUREKinomé
Chuck DE LIEDEKERKESoil Capital
Camille POUTRINGreenflex
Jean Pierre RENNAUDPlanet A
Thibaut SAVOYECarbone Farmers
Baptiste SOENENArvalis
Thierry TETUUniversité Picardie Jules Verne
Guillaume VIGNERONSmag
Gécica YOGOInrae

Share this article!

3 thoughts on “The race for carbon in agriculture

Leave a Reply

Your email address will not be published. Required fields are marked *