When we think about carbon and climate change, we often look up to the atmosphere where heat-trapping greenhouse gases (GHGs) accumulate, or to the green forest canopies that absorb carbon. But the largest carbon store on land lies beneath our feet. Soils hold more carbon than all the world’s trees and the atmosphere combined. They can release carbon when disturbed, such as through deforestation or intensive agriculture, or they can store more when farmers adopt practices that reduce disturbances.New research from WRI and partners shows that from 2016 to 2024, disturbed soils released approximately 3.9 billion metric tons of GHGs1 every year — nearly equivalent to the combined fossil fuel emissions of India and Saudi Arabia in 2024. Two-thirds of these emissions came from draining and burning peat and other carbon-rich organic soils, with those emissions concentrated in just 10 countries. The remaining one-third of emissions came from less carbon-rich soil spread over a much larger area. But soils aren’t a source of emissions everywhere. In some regions, soil carbon stocks have increased since 2016.Here, we unpack what’s going on with the world’s soil carbon and why keeping it in the ground — or even increasing how much is stored there — is critical for curbing climate change and sustaining food systems that billions of people depend on.What Is Soil Carbon?A handful of dirt might look simple, yet it’s anything but. Soil is alive, teeming with microorganisms in a complex mix of minerals, water, air and organic matter. It can be just a few centimeters deep or extend several meters below the surface, depending on the landscape. Soil’s texture, mineral composition and formation all shape how well it holds water, supports plant growth and cycles nutrients. They also affect how much carbon it stores and how quickly it gains or loses carbon over time. An important distinction for soils is whether they are classified as mineral or organic. Mineral soils — the soils found under most farmland, grasslands and forests — cover most of Earth’s land surface and typically contain relatively low concentrations of organic carbon, the carbon derived from decomposing or living organisms.Carbon-rich organic soils, by contrast, form in waterlogged or very wet, low-oxygen conditions where plant material decomposes slowly, causing very large amounts of carbon to build up over time, often as peat. Peatlands — the best-studied subset of organic soils — cover just 3% of the world’s land surface yet store roughly 600 billion metric tons of carbon, or about one-third of all soil organic carbon, in soil up to 10 meters deep.How Do We Define Organic Soils in This Analysis?There is no universal definition of ‘organic soil’ or ‘peatland.’ Different definitions use different minimum peat thicknesses, organic matter thresholds, soil taxonomies and underlying datasets, which results in different estimates of organic soil extent. For a globally consistent analysis, we use the broader Intergovernmental Panel on Climate Change (IPCC)-aligned category of organic soils, which encompasses peatlands. This allows us to follow standard international methods (IPCC Tier 1 framework) for estimating emissions. Our organic soil data should not be compared directly with peat extent or emissions estimates.We delineated organic soil using multiple datasets. First, we used almost 500,000 harmonized soil observations from around the globe to create a soil classification map. We combined a second dataset predicting peat depth probability with the soil classification map to produce a continuous map where each pixel represents a 30-meter-by-30-meter area of land (roughly the size of a baseball infield diamond), showing the probability that area is organic soil. To define organic soil extent from the probability dataset, we applied different probability thresholds depending on each climate zone.Emissions from organic and mineral soils are calculated differently under IPCC guidelines. Organic soil emissions are estimated by multiplying the area disturbed by an emission factor, while carbon losses and gains in mineral soils are derived from changes in soil organic carbon stocks over time. Organic soil emissions can include carbion dioxide (CO2), methane and nitrous oxide (standardized into CO2-equivalents) and are collectively reported here as metric tons of GHG. For mineral soils, methane and nitrous oxide emissions from cropland management are estimated separately and not reported here. We report changes in mineral soil organic carbon for the topsoil, or the top 30 centimeters of soil. What Causes Soil to Capture and Release Carbon? Soils are constantly exchanging carbon with the atmosphere, with plants and microorganisms acting as the bridge. As plants grow, they pull carbon dioxide (CO2) from the air and store it in their leaves, stems and roots. Over time, some carbon moves into the soil through roots and decaying plant material. Soil microbes break it down, releasing most of the carbon back into the atmosphere as CO2, while some stays in the soil, where it can remain for thousands of years. Under stable environmental conditions and land use, soils lose and gain similar amounts of carbon and there is no net change. But that balance can be disrupted. Land use change, disturbance or shifts in climate conditions reduce plant carbon entering the soil; at the same time, the release of carbon into the atmosphere and carbon transfer into water bodies through erosion and leaching continues. Deforestation, intensive agriculture, overgrazing and draining carbon-rich peatlands to make waterlogged land suitable for plantations and other uses all tip the balance toward emissions. Natural factors like drought, wildfire and rising temperatures exacerbated by climate change can also accelerate soil carbon losses.How much carbon is emitted, and how quickly, depends on the type of soil. In mineral soils, emissions tend to be gradual, driven by reduced plant carbon and ongoing decomposition and erosion. In organic soils, emissions can be more dramatic and abrupt. When soils are drained, oxygen enters and accelerates soil microbial activity, releasing CO2 year after year and causing the soil to sink over time as it breaks down. Dry organic soils are also more prone to fires, releasing carbon accumulated over centuries in a matter of days. The reverse is also possible: soils can gain carbon when more plant material is returned to the soil and disturbance is reduced. For example, improved land management practices in areas with mineral soils — such as cover cropping, leaving crop residue on fields, permaculture (when agricultural practices mimic natural ecosystems) and planting or preserving trees on farms (agroforestry) — can help rebuild agricultural soil carbon over time. However, gains are often slow, and it can take decades before there are measurable improvements.How Much Carbon Has Soil Lost Globally, and What’s Driving It? Soil emitted approximately 3.9 billion metric tons of GHGs per year between 2016 and 2024. Two-thirds of this came from organic soils (comparable to Russia and Iran’s CO2 emissions from fossil fuels in 2024), while the remaining 1.3 billion metric tons of CO2 per year came from mineral soils (comparable to Japan and Italy’s CO2 emissions from fossil fuels in 2024). Within organic soils, approximately 80% of estimated emissions were associated with drainage, with the remainder linked to fire. Drainage dominates partly because it creates an ongoing source of emissions: they recur year after year for as long as the soil remains drained. Fire adds episodic emissions when burning occurs, and drainage can further increase fire risk by drying out the peat. However, drainage is not the only way that humans dry organic soils. Climate change has triggered drying in some temperate and boreal peatlands due to reduced rainfall, leaving them more susceptible to fires and increasing their emissions when they do burn. For example, the vast majority of emissions from a fire in Scotland during the very dry summer of 2025 came from peatlands or bogs, including areas that were historically wet and would not burn. Where Are Soil Carbon Emissions Coming From? Emissions from organic soils are heavily concentrated: 93% come from just 10 countries. This is because organic soils are relatively uncommon and form only where waterlogged ground has allowed organic matter to accumulate over centuries. Further, large emissions occur only where these carbon-rich soils have also been drained, cultivated, extracted or burned, concentrating emissions in relatively few countries. Organic soil emissions are highest in Indonesia due to its extensive tropical peatlands, many of which have been drained over recent decades for logging and conversion to oil palm, pulpwood and other agriculture. Canal networks lower water tables across the surrounding peat, accelerating decomposition and increasing fire risk. Unless hydrology is restored, drained peat can remain a source of emissions for decades. Meanwhile, 10 countries are responsible for roughly three-quarters of mineral soil emissions. This reflects the larger area of mineral soils relative to organic soils and the broader scale of the land management processes that cause mineral soil carbon loss. For example, in China, conversion of grasslands and forests to cropland has caused soil carbon loss in some regions, while Brazil has lost soil carbon due to agricultural conversion and land-clearing fires. –> In addition to varying by region, mineral soil emissions varied by land cover type. While grasslands have the highest soil carbon emissions of any land type, causes of these emissions vary greatly and can include fires, overgrazing, drought and more. Additionally, some land identified as grasslands may include cropland because it can be difficult for satellites to distinguish annual crops — like soy and cotton — from some types of grassland vegetation. Combined, grasslands and land with annual crops lost 840 million metric tons of CO2 per year from mineral soils globally between 2015 and 2020. On the other hand, soil emissions from croplands are almost all human-caused. When separated from grasslands, these lands alone lost 190 million metric tons of CO2 per year from mineral soils globally between 2015 and 2020. Soil carbon losses can be severe in some regions. For example, the U.S. state of Iowa has lost half of its soil carbon in croplands over the last 200 years, with ongoing losses over large areas. This loss is largely attributed to continuous tilling of fields (similar to plowing), which exposes soil carbon to the air and accelerates decomposition. Over the wider U.S. Corn Belt in the Midwest, nearly 1.5 billion metric tons of carbon have been lost from hillslopes, reducing crop yields by 6% and causing nearly $3 billion of economic losses per year. What Are the Future Risks to Soil Carbon? Many of the current emissions from soil are a legacy of past land-use decisions and are likely to continue. But an even larger source of potential emissions looms: thawing permafrost.Permafrost is ground that has remained frozen year-round for thousands of years, locking away enormous amounts of carbon. It contains roughly half of the world’s soil organic carbon and underlies about one-quarter of the Northern Hemisphere land area. As climate change causes temperatures to rise and fires to become larger and more frequent, this frozen ground can begin to thaw and melt. Soil carbon that has been stored for millennia can then be exposed to oxygen and microbial activity, causing it to decompose and release CO2 into the atmosphere. This creates a dangerous feedback loop. More warming leads to more permafrost thaw, which releases more greenhouse gases, driving still more warming and accelerating climate change.The scale of this potential carbon release would be massive. By the end of the century, thawing permafrost could release around 7 billion to 15 billion metric tons of GHGs per year — exceeding the current annual GHG emissions of the U.S. — although the exact amount remains highly uncertain. And the consequences of thawing extend well beyond climate. As frozen ground thaws, the land can buckle and subside, damaging roads, buildings and pipelines, accelerating shoreline erosion and forcing entire communities to relocate in some regions. What Are the Implications of Losing Soil Carbon Beyond Climate?Beyond the climate impacts, soil carbon is also central to healthy, productive land. Carbon-rich soils retain nutrients, improve soil structure and hold water that plants need to grow. When carbon is lost, soils can quickly become depleted, compacted and more prone to erosion, with valuable topsoil washed or blown away. These impacts show up quickly for farmers. In the U.S. alone, the agriculture sector is estimated to lose $44 billion due to erosion, while farmers in South Asia lose an estimated $10 billion annually. Soil erosion decreases food production by over 30 million metric tons globally per year, exacerbating food insecurity in some regions. Crop yields have declined due to land degradation in areas where 1.7 billion people live, pushing greater reliance on chemical fertilizers and irrigation to compensate, increasing food costs and driving more agricultural expansion. At the same time, depleted and degraded soils are more vulnerable to droughts, floods and extreme weather, making farming systems more fragile and putting farmers’ livelihoods at greater risk. A farmer surveys a field of withered crops. Healthy soils are essential for maintaining agricultural productivity and supporting farmers’ livelihoods. Photo by Marccophoto/iStock Therefore, losing soil carbon is a triple whammy: It fuels climate change, weakens the resilience of the global food system to its impacts and threatens farmers’ livelihoods in both the short and long term.How Can We Protect and Rebuild Soil Carbon Stores?There are two main ways to protect soil carbon: keep it from being lost in the first place and rebuild soil carbon where it’s been depleted. Both matter, as avoiding losses prevents additional carbon from entering the atmosphere, while rebuilding carbon stocks happens over decades, if ever. More broadly, reducing fossil fuel emissions that drive climate change is critical: As the climate warms, soils are more likely to dry, burn or thaw, increasing the risk of large, irrecoverable carbon losses.Globally, the largest opportunity for soils is to halt the drainage of carbon-rich organic soils like peatlands. Peatlands cover just 3% of land and produce only about 1% of global calories, but their drainage releases roughly one-third of agricultural emissions. Keeping these soils wet excludes the oxygen that drives carbon loss and reduces the risk of fires that can smolder for multiple years.Once these organic soils are degraded, it is possible to restore them, but recovery is slow. Rewetting drained peatlands and adding more organic matter, like integrating trees and cover crops into agricultural systems, can gradually boost soil carbon. But rewetting doesn’t instantly restore peatlands, and methane emissions may rise in the short term. Even so, it stops the much longer, ongoing CO2 losses and puts organic soils back on a path to recovery. For mineral soils, avoiding the conversion of forests and other natural ecosystems to more intensive uses like agriculture and settlements helps avoid carbon loss. Carbon gains are slower in mineral soils, requiring consistent management and long-term incentives. Improved land management practices — such as agroforestry, no-till farming and crop rotation — may sometimes help rebuild agricultural soil carbon over time, but reversing losses is slow and difficult. Soil carbon has been increasing in some landscapes due to land management practices. For example, China’s Northeast Plain is a major grain production region, but the highly fertile soil is prone to erosion. Soil carbon density decreased from 1980 to 2006 and has generally increased since then, potentially due to farming methods that disturb the soil less, like conservation tillage. Soil carbon has also increased in the eastern U.S. where reforestation has occurred after agricultural abandonment. From Monitoring to Protecting and Rebuilding Soil CarbonBecause soil carbon is so hard to monitor, it is often underrepresented in GHG inventories and other carbon monitoring systems. Consistently monitoring soil carbon over large areas can help mainstream it and raise awareness about the need for protection. Advances in machine learning and in standardizing and aggregating field data are helping make that possible.At the same time, we cannot let a lack of highly precise soil carbon monitoring prevent us from taking action. Other indicators like soil structure, water retention and vegetation cover often go hand in hand with increasing soil carbon, offering practical signals that soil is recovering. These other indicators can help guide action now, while monitoring improves and more precise soil carbon measurements continue to develop over time. Moreover, because we know that GHG emissions from soils largely originate from organic soils and are concentrated in a few countries, actions such as redirecting agricultural expansion and supporting alternative livelihoods in communities that would otherwise depend on draining peat should be prioritized. Protecting and rebuilding soil carbon isn’t only about climate change mitigation. Healthy soils underpin food security, support livelihoods and improve the land’s ability to handle droughts, floods and other extremes. By keeping carbon in the ground and rebuilding it where it’s been lost, we can curb climate change while sustaining the production of the food, feed, fiber and fuel we all depend upon.1 Greenhouse gas emissions are calculated as metric tons of carbon dioxide equivalent (CO2e). This includes emissions from CO2, methane and nitrous oxide.