What Changes are needed in Agricultural Climate Target-Setting?
How we got HERE
Over the past two and half decades corporate climate reporting has grown to an exponential scale. Today 20,000 companies (about 70% of global market capitalization) report to CDP. Over 10,000 of these companies (about 40% of global market capitalization) have valid targets with The Science Based Targets Initiative (SBTi).
Much of the focus of the first decade and a half of this system was on reporting and reducing industrial emissions, but in 2017, Greenhouse Gas Protocol (GHGP) released their first agricultural guidance. Prior to that time there was a vacuum in trying to understand how to report a corporate climate footprint with agricultural systems that intermingle human-caused and natural GHG impacts. Even with this support, GHG accounting was still often inconsistent due to the complexity of the land sector. In 2020, SBTi began work on the Forest Land and Agriculture (FLAG) guidance, and by 2025, over 260 companies set FLAG targets (the recent FLAG questionnaire mentions this is now over 400). Many food and beverage companies, some of which are the largest in the world, have set targets aligned with the standard.
Even with this momentum more standardization was needed to incentivize how to establish a land sector inventory, and begin to account for projects. This year the GHGP Land Sector and Removals Standard and accompanying guidance (LSRS/G) was released to supersede the 2017 guidance and provide clear requirements for how to complete an agricultural GHG inventory and account for some of these actions in their physical inventory.
So we have made a lot of progress in the past decade, but there are more questions related to taking action: how should we “count” the GHG impact of projects in the inventory when we have limited traceability and complex supply chains? What models should we use, and where? How do we navigate extreme data limitations when comparing our data-rich programs to limited regional emission factors? What are the best practices to address permanence, crop rotation, co-product allocation, quantifying removal uncertainty, etc?
While the forthcoming Actions and Market Instrument section of the anticipated GHGP Corporate Standard may help provide clarity to some of these questions, many will remain unanswered until future revisions of the LSRS. Meanwhile agriculture has four years to decrease GHG impact of the sector by 50% (from a 2020 base year) to achieve a 1.5 degree GHG reduction pathway.
What We’ve Learned
Though scaling agricultural GHG projects on the ground, we are at the point now where the industry has learned valuables lessons:
Changes take time, and are difficult to measure each year - In the first few years of a project emissions (using GHG inventory accounting methods) can increase, but that may be because a different crop is grown, manure is applied, weather, yield decreases, or multiple combinations of these factors. This can “wash out” the immediate effect of a project action. It takes time for the GHG reduction of a given practice change to emerge.
Comparing geographies is challenging - Soil conditions and weather contribute to the differences we see in carbon intensity as we move between geographies. We cannot expect areas with little precipitation, a short growing season, and poorer soil to have the same carbon intensity of geographies that have consistent rain-fed irrigation and rich soils once practices are adopted. There is a “potential” to change down to the field scale.
Shared, complex value chains, co-product allocation, and crop rotation make GHG accounting complex - There are often many intermediaries between a reporting company and the farm. Companies often don’t know the farmers they are sourcing from. Further, even if they were known, a company does not source from the same farmers each year with crop rotation and changing consumer demand. To make this more complex as an industry we want to incentivize more crop rotation in many geographies. Assuming a company can trace back to the farms they source from, they may just source a derivative of a crop (like soy meal), making justification of an intervention on that farm complex because they can’t claim the full GHG benefit.
Reaching goals often requires systems change - If a company sources from 10 million acres each year across a 100 million acre sourcing geography, 10 million acres of project-driven change is just 10% of the system, which may result in just a small impact on the average carbon intensity of the corps they source. System-wide collaboration is needed for the whole 100 million acres to achieve more dramatic reductions and achieve long-term targets, yet corporate GHG accounting is focused on individual actions.
Sector-specific standardization is needed - Once a project is underway there is no universal methodology for estimating the GHG impact of a crop, making it hard to compare project outcomes. Different data may be collected and different assumptions and quantification approaches may be used. Even if a universal quantification system existed that was standards-aligned, environmental data needs from literature, weather data, soil data, and challenges with public or collected agronomic data may limit comparability.
Understanding differences in carbon intensity can be complex - When change occurs due to natural variation knowing what claim to make can be challenging, while a practice may have contributed to a reduction, it often does not explain the entire change. Approaching conservativeness in this scenario can be complex when reporting progress towards a target. It is also challenging comparing a project to a regional geography given the variety of carbon intensity impacts due to soil and weather (as discussed above).
Project accounting is also complex and not a universal solution - While in many cases the proposed solution to some of these challenges is to use project or carbon-credit style accounting, that solution also has significant challenges. Project accounting requires a project scale, significant historical data (to define a counterfactual baseline), technical infrastructure (to build and model that baseline), and/or modeling/sampling system that is challenging to create for every geography, crop, and practice. Additionality can also be awkward to apply in agricultural systems, where each year for a farmer is often considered a “new decision”. Even then, these are system-wide estimates of impact that still need real-world validation from inventory-based reductions we can measure over the long-term. All this being said, project accounting outcomes can help, and it would be ideal in almost every case to use project and inventory accounting in concert, but this often has significant cost implications for a project and limits the scale and geography where companies can take action (at least for some crops).
It is often not as simple as reducing GHG emissions and increasing productivity - A GHG focused system often does not maximize social benefit, for example, irrigation in a water-stressed area can reduce carbon intensity but deplete an aquifer; more application of pesticides (or different pesticides) may decrease carbon intensity by increasing yield, but increase adverse biodiversity impacts; large-scale industrialization may increase yield and the ability to collect data, but exclude small family farms and decrease rural economic livelihood. In reality the land sector is a very complex system with many “levers”, and all ecosystem services and social factors need to be considered to maximize social benefit prior to pulling one of them.
What We Need
Agronomic Data
Even the best agronomic data sources have gaps - In the United States we have the National Agricultural Statistics Survey, which is an incredible resource. Still, soil health practices like cover cropping and no-till are just surveyed every five years at the county level in arrears. We have no public data between these surveys and no data at a sub-county scale. There is no national dataset of field boundaries (and maybe there shouldn’t be due to data privacy concerns). There is a county-level yield dataset, but funding for it was pulled for a short period of time in recent years (it is now re-instated). We have no public dataset of yield at a sub-county scale. Yield is critical, as it is the denominator of carbon intensity, so any risk of losing this data is concerning.
Soils data is often limited - In the US we have a nation-wide soil survey, which is a helpful resource, but it is also up to 100 years old in some geographies, and the current spatial resolution may result in inaccuracies/uncertainty when trying to understand soil properties at a field-scale.
Project and regional-scale comparison are difficult - As evidenced above, even when a county invests significant resources into developing activity data, data gaps in the region make comparisons of project results to the surrounding geography challenging. This is the exception, not the rule, as most other geographies have more limited agronomic data, which makes understanding the “typical” practices and carbon intensity more challenging.
Time for Technological Development, Research, and GHG Accounting to Mature
Technology is improving but more work/investment is needed - There are multiple technological solutions to support enrollment and the collection of farmer data, quantification, and the reporting of GHG outcomes. These solutions have improved on a rapid basis. That being said, it can be difficult to identify solutions that can characterize every crop, practice, and geography on an annual basis and meet all LSRS requirements. A solution may be capable of this for removals, but then not cover all relevant emissions sources (or vice-versa). As mentioned above, even when a product level analysis can be completed, it can be hard to contextualize results from an inventory accounting standpoint due to data limitations or lack of standardization with other projects and approaches.
Sector specific standardization to verify outcomes - A standardized Scope 3 program validation and verification system continues to be a large gap for both projects and general inventory reporting. Today there are few companies that can validate or verify the GHG programs in agriculture for GHG inventory accounting, and they often have to develop their own protocols and standardized-processes. While some companies and organizations are leading on this, it will take time for these systems to scale and mature.
More scientific research is needed - In parallel research continues to help us understand the full GHG benefits of many conservation practices. These benefits are not universal and can differ depending on crop, geography, and practice. We often know there is a general benefit to conservation actions, there are still gaps to understand the full story. It is very important as an industry we are aligned on what we know and don’t know about the outcomes of a given conservation action to manage reporting risks. In addition to this, many of these changes just take time to research (we can’t hurry the natural system along), and we will need time to understand with certainty the GHG benefit of an action.
A Collective Vision / Goal
Maximizing GHG outcomes alone is often not in our best interest - We need to apply a systems approach to understand all the pros and cons of actions across our environmental and economic systems to make decisions. Just achieving climate targets may not be in our best interest. It may be less work long-term to first evaluate the total benefit of taking a given set of actions. For example, perhaps a moderate GHG benefit is possible with many co-benefits, and this is advantageous to a larger GHG benefit with no co-benefits.
Focusing action just in locations where we can measure GHG impact with certainty may not be in our best interest - We have the most certainty in GHG accounting in areas where we have infrastructure (i.e. internet and resources to collect data), surveyed data from governments, and literature to support modeling and reporting outcomes. There are many geographies where we need to prioritize action, but do not have the same certainty when reporting GHG impact. This inverted structure of certainty and needed prioritization is the opposite of what is needed to achieve climate resilience - we need a system that prioritizes areas where the most help is needed, even if our ability to report is uncertain. For example, often an accurate GHG result may not be possible for a small farm growing many crops (modeling more than one crop is very difficult), but these farmers are often the most eager to engage and make positive changes.
We need a collective vision - focusing on achieving a GHG reduction target is often company, rather than landscape focused. This creates the risk of actors not working towards a common-definition of what a “sustainable” agriculture system looks like. Some of this was evaluated during the current definition of SBTi FLAG (such as requiring zero-deforestation commitments), but more definition of our vision for well-managed agricultural systems is needed. What vision are we all working towards?
Where we go from here
The case for activity based targets
Many companies are risk-averse and hesitate to take action because they don’t want to make GHG claims unless they know they are conservative. On the other hand often, we have already identified the general actions we should be taking as a sector (regardless of if these GHG accounting challenges have been resolved). To address this difficulty while we continue to improve our GHG accounting system we should explore activity-based interim targets, where we identify what conservation activities in each geography of our planet are needed (and where we should prioritize to maximize social impact) and take them. Many companies today already have a form of this, they set regenerative agriculture targets and track progress where they incentivize practice change. Given how uncertain it can sometimes be to identify GHG outcomes (in particular over short periods of time), we should establish a target-setting system that is focused on these goals.
To set these goals we need to first determine the activities (what does “regenerative” look like in a given area)? This will differ by geography, but this work (and our ability to verify and track practice adoption over time) is needed regardless, and can keep the momentum of actions moving until our system to model, report, and verify GHG outcomes everywhere, for all commodities and practices is complete.
Using these more-categorical approaches to measuring progress can increase conservativeness in claiming, make verification easier, and reduce hesitation on taking action by reducing the complexity of this system over the next 10-15 years (i.e. we don’t know the what the exact GHG impact will be but we know in general it will be helpful). What makes this approach more ideal is that it is not a replacement of the current GHG-focused system, but an interim solution as we continue improve our ability to develop our system (data, technology, research) to quantify GHG impacts with certainty.
It is critical to validate that GHG reductions are occurring through these actions, but perhaps this is easier done at scale, and may be more effective through collaborative partnerships with academic institutions and to track change over time.
Change has been constant in trying to reduce the climate impact of agriculture over the past two decades, and we have made massive improvements on defining standardized approaches, but solving the complexity of GHG accounting and reporting is still interfering with our ability to take action to reach 2030 targets. Activity-based target setting can keep the momentum of investment moving forward while we continue to build-out a complete GHG accounting system. It encourages collaborative and shared decision making. Over time and can report on the GHG outcomes of these approaches, which will tell us if more or less investment is needed to achieve landscapes that are managed in alignment with a 1.5 or 2 degree C planet.
