Sustainability & Environment (Agri-Food Systems) · Climate Science & Agriculture · also Soil Carbon Management, Carbon Farming, Terrestrial Carbon Sequestration
How much carbon soils and vegetation can store, how it is lost, which practices build it, and how to measure, report and verify changes credibly, including saturation, permanence and reversal risk.
In the exams: ICAR AICE subject 71 Unit 3 (carbon sequestration); carbon-market and MRV roles; PG and PhD research
Soils hold more organic carbon than the atmosphere and all vegetation combined. Common estimates put the organic carbon in the top metre of the world's soils at roughly 1,500 billion tonnes of carbon (Gt C), with more at depth and in permafrost and peat, against roughly 870 Gt C in the atmosphere today. Soil organic carbon (SOC) is the carbon in soil organic matter -- plant and microbial residues at every stage of decomposition, living microbes, and organic compounds bound to minerals. It underpins soil fertility, structure, water-holding and biological activity, and small changes in this very large stock are significant for the climate: losses from cultivation and drainage have added CO2 to the atmosphere for centuries, and well-managed soils can take some of it back.
Carbon sequestration means the removal of CO2 from the atmosphere and its storage in a reservoir -- here, soils and the vegetation growing on them. This subject explains how soil carbon is organised into pools that turn over at different rates, what controls how much carbon a soil holds, which practices build it and by how much, why gains slow down and can be reversed, and how changes are measured, modelled, reported and verified. It also covers the international '4 per 1000' initiative and national programmes, and the inorganic side of soil carbon, including carbonates and enhanced rock weathering.
The subject sits between soil science, climate policy and carbon markets. The inventory equations that turn stock changes into CO2 (Tier 1 stock-change factors, the 20-year default, 44/12) are taught in Carbon Accounting and GHG Inventories; the crediting of soil carbon -- additionality, buffers, uncertainty deductions -- in Carbon Markets, Offsets and Payments for Ecosystem Services; and the other agricultural gases, which can offset or exceed soil carbon gains, in Agricultural GHG Emissions and Mitigation. Soil fundamentals (texture, organic matter chemistry, nutrient cycling) are in Soil Science for Sustainability.
Soil carbon appears in the ICAR PhD paper in Environmental Sciences (carbon sequestration), UGC NET, GATE Environmental Science, ICAR soil-science papers, the Certified Crop Adviser Sustainability specialty and soil-scientist certifications, and in the daily work of carbon-project developers, verifiers and researchers. Sequestration rates quoted here are broad ranges from reviews; any figure for a particular field depends on climate, soil, history, depth sampled and duration, and should be treated as an estimate with uncertainty, not a constant.
Soil organic matter (SOM) is about half carbon by mass; the traditional conversion multiplies organic carbon by 1.724 (the van Bemmelen factor, which assumes SOM is 58% carbon), though a factor nearer 2 fits many soils better. SOM includes fresh and partly decomposed plant residues, roots and root exudates, living microbial biomass (typically a few per cent of SOC), microbial residues (necromass), and organic compounds adsorbed on mineral surfaces or held inside aggregates.
The older view that most SOM is 'humus' -- large, chemically recalcitrant humic molecules that resist decay -- has given way to a view of persistence as an ecosystem property. Organic compounds persist mainly because they are physically protected inside aggregates, chemically bound to clay and silt minerals and to iron and aluminium oxides, or kept in conditions (cold, waterlogged, deep) where decomposers are inactive; intrinsic chemical recalcitrance matters less than once thought. Microbial necromass is now recognised as a large source of mineral-associated carbon: microbes process plant inputs and their remains bind to mineral surfaces.
A practical distinction separates particulate organic matter (POM), fragments of plant material that turn over in years to decades and respond quickly to management, from mineral-associated organic matter (MAOM), smaller molecules bound to fine particles that persist for decades to centuries and are limited by the soil's fine-mineral surface area. Building MAOM gives more durable storage; building POM gives faster but less secure gains.
Models and teaching describe SOC as conceptual pools with different turnover times: an active or labile pool (months to a few years), a slow or intermediate pool (decades) and a passive or stable pool (centuries to millennia). The mean residence time (MRT) of carbon in a pool at steady state equals the stock divided by the input flux (or the output flux, which is the same at steady state). Radiocarbon dating and the natural 13C shift after switching between C3 and C4 vegetation are used to estimate turnover.
Most models treat decomposition as first order: the rate of loss is proportional to the stock, dC/dt = I - kC, where I is the annual carbon input and k the decay constant (1/k is the turnover time). At equilibrium, dC/dt = 0 and C* = I/k: the equilibrium stock is the product of inputs and turnover time. Doubling inputs doubles the equilibrium stock; halving the decay rate (for example by reducing disturbance) also doubles it. After a change, the stock approaches its new equilibrium exponentially, C(t) = C* + (C0 - C*) e^(-kt), so most of the change happens in the first few turnover times. This simple model explains much of what follows: why gains slow, why losses after disturbance can be fast, and why inputs are the main lever.
Each year plants take up of the order of 120 Gt C from the atmosphere by photosynthesis, and roughly as much returns through plant and soil respiration; soil respiration -- roots plus decomposers -- is one of the largest fluxes in the carbon cycle. Because these gross fluxes are so large compared with the net change, a small imbalance between inputs and decomposition across the world's soils translates into a significant net source or sink. Historically, conversion of natural land to agriculture has released a substantial amount of soil carbon, often estimated at tens to over a hundred Gt C, which defines a 'carbon debt' that better management can partly repay.
Carbon enters soil as leaf and root litter, root exudates and, on farms, crop residues and manure; it leaves mainly as CO2 from respiration, and smaller amounts leave as dissolved organic carbon, methane from waterlogged soils, and eroded particles. Roots and their exudates contribute disproportionately to stable soil carbon compared with above-ground residues, because they are released within the soil matrix close to mineral surfaces. This is one reason why deep-rooted perennials and cover crops are effective at building soil carbon.
Climate sets both inputs and losses. Plant production (and so carbon input) rises with rainfall and temperature where water is not limiting; decomposition also speeds up with temperature, roughly doubling for each 10 degrees C rise in many studies (a Q10 near 2), and is slowed by both drought and waterlogging. The balance explains broad patterns: cold and wet regions accumulate large stocks (boreal soils, peatlands, permafrost); hot humid tropics have high inputs and fast decomposition; arid soils have low inputs and low stocks. Waterlogging excludes oxygen and allows peat to form, which is why organic soils hold far more carbon per hectare than mineral soils.
Texture and mineralogy determine protection. Clay and fine silt provide surfaces that bind organic compounds; soils with more fine particles generally hold more SOC under the same inputs, and their capacity to store mineral-associated carbon is greater. Allophane and iron and aluminium oxides in volcanic and highly weathered soils bind carbon strongly. Aggregation, promoted by roots, fungal hyphae and microbial products, protects carbon inside aggregates; disturbance by tillage breaks aggregates and exposes it. pH, nutrient supply and depth also matter: subsoils often hold older, more stable carbon at lower concentrations, but because they are thick they hold a large share of the profile total.
Land use is the strongest human control. Converting native forest or grassland to cropland typically reduces SOC in the topsoil by a substantial fraction -- often tens of per cent over decades in reviews -- through reduced inputs, residue removal, tillage-induced aggregate breakdown and erosion. Converting cropland back to grassland or forest usually raises SOC, but more slowly than it was lost. Draining organic soils for agriculture causes continuing losses as peat oxidises.
Within a land use, management changes inputs and losses: crop yields and residue return, cover crops, rotations with perennials, manure and compost, fertilisation (which can raise inputs through higher yields), irrigation (raising both production and sometimes decomposition), grazing intensity and timing, burning, tillage, and erosion control. IPCC Tier 1 methods summarise these as factors for land use, management (tillage) and input level multiplied by a reference stock for each climate and soil class, applied to the top 30 cm (see Carbon Accounting and GHG Inventories). The reference stock and factors differ by climate zone and soil type, which is why the same practice can give very different changes in different places.
Because the equilibrium stock is proportional to inputs, practices that add more organic carbon to the soil are the most reliable. Returning crop residues instead of removing or burning them adds carbon directly. Cover crops grown between cash crops add root and shoot carbon and protect soil; a widely cited global meta-analysis estimated an average gain of around 0.3 t C per ha per year in the topsoil, with large variation and a decline over time. Perennial crops, leys and grass in rotations add deep, continuous root inputs. Higher yields from better nutrition and water management increase residue and root inputs, though the fertiliser's own N2O must be counted. Manure, compost and biosolids add carbon, but at landscape scale they often move carbon rather than create it: if the manure would otherwise have been applied to another field, the net sequestration is small. Biochar (covered in the mitigation subject) adds a very stable form of carbon.
Agroforestry -- trees in fields, hedgerows, silvopasture, shelterbelts -- adds both soil carbon and biomass carbon, often more in the trees than in the soil, and can deliver relatively large removals per hectare where trees grow well. Converting cropland to permanent grassland or forest gives the largest soil gains but takes land out of crop production.
Reduced tillage and no-till reduce disturbance, keep residues on the surface and increase carbon in the top few centimetres. However, many sampling studies that went deeper found that no-till often redistributes carbon -- more near the surface, less below the plough layer -- so the whole-profile gain is smaller than surface measurements suggest and sometimes not significant. No-till still has clear benefits for erosion control, water infiltration and fuel use, and gains are more likely when it is combined with residue retention and cover crops.
Grazing management can increase or decrease SOC. Overgrazing reduces plant cover and inputs and increases erosion; improved management -- appropriate stocking rates, rest periods, sowing productive or deep-rooted species and legumes -- can raise SOC, particularly on degraded grasslands. Claims for very large gains from particular rotational grazing systems are contested, and results vary greatly with climate and starting condition. Grazing land also emits enteric methane and N2O, which may outweigh soil gains.
Preventing erosion keeps carbon-rich topsoil in place; whether eroded carbon is a net source depends on what happens to it on deposition. Avoiding burning of residues and grassland retains inputs.
After a change in management, SOC moves towards a new equilibrium set by the new balance of inputs and decomposition. Because the approach is roughly exponential, the annual gain is greatest at first and declines towards zero as the new equilibrium is reached, typically over a few decades in temperate topsoils; IPCC Tier 1 methods assume the change is completed over 20 years. Sequestration is therefore finite: a field cannot keep gaining carbon at its first-year rate indefinitely. Long-term experiments such as Broadbalk at Rothamsted (since 1843), where plots given farmyard manure every year built up SOC rapidly for decades and then much more slowly, illustrate the pattern.
Saturation adds a ceiling on mineral-associated carbon: the fine mineral fraction has a finite capacity to bind organic matter. Hassink (1997) related the maximum carbon in the fine fraction to its proportion, C_sat (g C per kg soil) = 4.09 + 0.37 x (percentage of particles smaller than 20 micrometres). The 'saturation deficit' -- the gap between current and maximum MAOM -- indicates where additional inputs are most likely to be stabilised; soils far below saturation, such as degraded cropland soils, have the greatest potential. Particulate organic matter is not limited in the same way but turns over faster and is more vulnerable.
Reported sequestration rates for improved practices are usually a few tenths of a tonne of carbon per hectare per year in topsoils, with wide ranges, and are larger for land-use change to grassland or forest. A rate measured over the first five years after a change should not be extrapolated to 50 years. Comparisons between treated and control plots measure the difference between two trajectories: if the control is losing carbon, a practice that merely slows loss shows a 'gain' relative to the control, which is a real benefit for the climate but is not net removal from the atmosphere. Studies must also state the depth sampled, whether stocks were compared on an equivalent soil mass, and the duration.
The finite and slowing nature of sequestration matters for policy: soil carbon can contribute meaningfully for a few decades while other emissions are cut, but it cannot offset continuing fossil emissions indefinitely.
Consider a cropland soil holding 40 t C/ha in equilibrium with annual inputs of 2 t C/ha. Its turnover time is 40/2 = 20 years and k = 0.05 per year. If cover crops raise inputs to 2.5 t C/ha with no change in k, the new equilibrium is 2.5/0.05 = 50 t C/ha. The gain of 10 t C/ha is approached exponentially: after 20 years (one turnover time) the stock has moved 1 - e^-1, about 63%, of the way, reaching about 46.3 t C/ha; after 40 years about 86%. The average rate over the first 20 years (about 0.32 t C/ha/yr) is much larger than over the next 20 (about 0.12 t C/ha/yr), and the first-year rate (0.5 t C/ha) is larger still.
The same reasoning shows that if the cover crops stop, inputs fall back and the soil returns towards 40 t C/ha. It shows too why reducing decomposition (lower k) is as effective as increasing inputs, and why a soil that is still losing carbon after conversion will show a 'gain' relative to an unimproved control even if its absolute stock is falling. Real soils have several pools with different k values, which models such as RothC represent, but the single-pool calculation captures the essential behaviour and is widely examined.
Carbon built up in soil remains only as long as the practices and conditions that built it continue. Ploughing grassland or a long-term no-till field, removing residues, stopping cover crops, draining wetland soils, overgrazing, fire and erosion can all return carbon to the atmosphere. Losses are often faster than gains: carbon accumulated over 20 years can be lost in a few years after disturbance, an asymmetry that follows from first-order kinetics (disturbance raises the decay constant immediately) and from the vulnerability of particulate organic matter.
Climate change itself is a reversal risk. Warming speeds decomposition, and drought reduces inputs; projections suggest that soils, especially in high latitudes and permafrost regions, could become net sources of carbon, though the magnitude is uncertain and depends on how much plant inputs also increase. Practices that build mineral-associated carbon in deeper layers offer more security than those that build surface particulate carbon.
For inventory purposes, reversals are simply counted as emissions when they occur. For carbon credits and corporate removal claims, permanence must be managed explicitly (see Carbon Markets, Offsets and Payments for Ecosystem Services): contracts commit landholders to maintain practices and monitoring for a defined period, often decades; programmes withhold a share of credits in buffer pools to cover reversals; some issue temporary credits or use tonne-year accounting; and reversals must be reported and compensated. Permanence periods vary between programmes, from a few decades to 100 years, which raises the practical problem of obligations outlasting a landholder or a change of ownership.
The climate value of temporary storage is debated. Storing a tonne for 20 years delays warming and buys time, but a tonne later released returns the warming. For this reason, net-zero frameworks increasingly distinguish durable removals from biological removals with reversal risk, and many argue that soil carbon is best treated as a contribution to mitigation and soil health rather than a one-for-one offset for fossil emissions.
A mixed farm converts an arable field to a grass-clover ley grazed by sheep and, over eight years, builds topsoil SOC by an estimated 6 t C/ha. The farm sells credits for the gain under a contract with a 25-year permanence period. In year nine, a drop in lamb prices and a rise in wheat prices make ploughing the ley attractive. If the field is ploughed and returned to continuous arable cropping, much of the gained carbon -- particularly the particulate fraction -- will be lost over the following years, and the reversal must be reported; under typical rules the buffer pool covers unavoidable losses but an intentional change of land use is an avoidable reversal that the farm must compensate. The example shows why permanence is an economic as well as a biophysical question: the carbon stays only while the practice remains more attractive than the alternative, and contracts, payments and penalties must be designed with that in mind. It also shows why contracts must say what happens if the farm is sold.
A SOC concentration (g C per kg soil or %) must be converted to a stock (t C per ha) for climate purposes. For a layer: SOC stock (t C/ha) = SOC concentration (g/kg) x bulk density (g/cm3) x layer thickness (cm) x 0.1 x (1 - coarse fragment fraction), or equivalently SOC (%) x bulk density x thickness. For example, 1.2% carbon at a bulk density of 1.3 g/cm3 over 30 cm gives 1.2 x 1.3 x 30 = 46.8 t C/ha before any coarse-fragment correction. Bulk density -- dry mass of soil per unit volume -- is measured with cores of known volume dried at 105 degrees C; errors in bulk density propagate directly into stocks, and coarse fragments (stones and gravel above 2 mm) must be excluded because they hold no organic carbon.
Depth matters. IPCC Tier 1 and many programmes use 0-30 cm, but management changes can redistribute carbon within the profile, and subsoils hold large stocks; sampling deeper (often to 60 cm or 1 m) gives a more complete picture but is costly. A study reporting surface gains alone may overstate whole-profile change.
When management changes bulk density -- for example, no-till compacts the surface while tillage loosens it -- comparing stocks to a fixed depth compares different masses of soil. A denser soil packs more mass, and hence more carbon, into the same 30 cm, creating an apparent gain even if no carbon was added. The equivalent soil mass (ESM) method compares stocks in the same mass of mineral soil per hectare, adjusting the depth of each sample to match a reference mass. Most current standards for soil-carbon crediting require ESM or an equivalent correction.
Detecting change is statistically demanding because SOC varies greatly across a field while changes are small. Good designs stratify the area by soil type and management history, take many cores composited within strata or individual cores at georeferenced points, re-sample the same points or strata after several years (commonly five or more), and calculate the number of samples needed: n = (t x s / d)^2, where s is the standard deviation, d the minimum change to detect and t the Student t value for the chosen confidence. Paired designs (re-sampling the same locations) reduce the variance of the change. Sample handling -- drying, sieving to 2 mm, removing roots and stones, archiving -- must be consistent between rounds.
Dry combustion is the reference method for total carbon: a finely ground sample is burned at high temperature (typically above 900 degrees C) in oxygen in an elemental analyser, and the CO2 produced is measured, often together with nitrogen (the Dumas principle). It measures total carbon, so in soils containing carbonates the inorganic carbon must be removed first by acid pre-treatment, or measured separately (for example with a calcimeter that measures CO2 released by acid) and subtracted, to obtain organic carbon. Dry combustion is precise and fast but needs equipment and careful sample preparation.
The Walkley-Black method oxidises organic carbon with potassium dichromate in sulphuric acid and measures the dichromate consumed by titration or colorimetry. It is cheap and widely used, especially in India and other countries for soil-testing programmes, but oxidation is incomplete: a recovery of about 77% is commonly assumed, so results are multiplied by a correction factor of about 1.3, although actual recovery varies with soil type and organic matter form. It also generates hazardous chromium waste. Loss on ignition (heating at 360-550 degrees C and weighing the loss) estimates organic matter rather than carbon and is affected by water lost from clays; converting it to carbon needs locally calibrated factors. For carbon accounting, dry combustion is preferred, and results from different methods should not be mixed in a time series without calibration.
Visible-near-infrared (VNIR) and mid-infrared (MIR) diffuse reflectance spectroscopy predict SOC from the way a sample reflects light, using statistical calibration (for example, partial least squares regression or machine-learning models) against samples analysed by dry combustion. Once calibrated, spectroscopy is much cheaper and faster per sample, needs no chemicals, and can estimate several properties at once; MIR generally predicts SOC more accurately than VNIR on dried, ground samples. Large spectral libraries now exist at national and global scales.
The weaknesses are the need for good local calibration (predictions degrade when samples differ from the calibration set), the need to report prediction error, and sensitivity to moisture and particle size, especially for in-field probes. For MRV, spectroscopy is used to increase sample numbers cheaply, with a subset analysed by dry combustion to check accuracy. Bulk density still has to be measured separately, though field instruments and pedotransfer functions are being developed for it.
The Rothamsted Carbon Model (RothC), developed by Jenkinson and colleagues from the long-term experiments at Rothamsted, simulates the turnover of organic carbon in non-waterlogged topsoils on a monthly time step. It divides SOC into five compartments: decomposable plant material (DPM), resistant plant material (RPM), microbial biomass (BIO), humified organic matter (HUM) and a small inert organic matter fraction (IOM) that does not decompose. Plant inputs are split between DPM and RPM by a ratio that depends on vegetation type; each active pool decays by first-order kinetics modified by rate factors for temperature, soil moisture deficit and plant cover, and the decomposed carbon is partitioned to CO2, BIO and HUM according to clay content.
RothC needs relatively few inputs -- monthly temperature, rainfall and evaporation, clay content, plant carbon inputs, and soil cover -- which makes it widely used for national and project assessments. Because inputs are often unknown, users run the model 'in reverse' to find the inputs that reproduce a measured stock at equilibrium. It does not simulate plant growth or nitrogen.
The Century model (Parton and colleagues, 1987) simulates carbon, nitrogen, phosphorus and sulphur in plant-soil systems on a monthly step, with litter split into structural and metabolic pools and SOM into active, slow and passive pools whose turnover depends on temperature, moisture, texture and cultivation. It includes plant production, so management affects both inputs and decomposition. DayCent is its daily version, which adds simulation of trace gases including N2O and methane and is used as a Tier 3 method in some national inventories, notably the United States. Other models include process-based ecosystem models and country systems such as Australia's FullCAM.
Models are only as good as their calibration and validation against long-term experiments. Good practice is to calibrate with independent data, validate against sites not used in calibration, report model uncertainty (for example by Monte Carlo analysis of parameters and inputs), and check that the model is applicable to the climate, soil and practices of interest. In carbon crediting, models are often used to estimate change between measurement rounds (measure-model-remeasure), with periodic soil sampling to keep them honest. Models also project future change and test the plausibility of claimed sequestration rates against equilibrium behaviour.
Measurement, reporting and verification (MRV) is the system that turns a claim of soil-carbon gain into a number someone else can check. Three approaches are used, alone or combined. Measurement-based MRV samples soils at baseline and after several years, using stratified designs, equivalent soil mass and dry combustion; it is direct but costly and slow to detect change. Model-based MRV uses calibrated models driven by management records, which is cheaper and gives annual estimates but depends on the model's validity for each field. Hybrid measure-model-remeasure uses models between periodic measurement rounds that 'true up' the estimates. Activity-based approaches simply verify that practices were adopted and apply default factors, as in IPCC Tier 1; they are simple but imprecise for individual farms.
Remote sensing supports MRV by verifying practices -- cover crops, tillage, residue cover, crop types -- and by mapping strata, but satellites cannot yet measure SOC change reliably below the surface. Digital tools that record management on each field reduce the cost of collecting activity data.
A credible soil-carbon report states the baseline, the depth and method of sampling, the bulk-density and ESM approach, the laboratory method, the model and version if used, all emission sources that change with the practice (N2O from added nitrogen, fuel, methane from livestock), leakage (for example, reduced yields shifting production elsewhere), and the uncertainty of the estimated change. Many crediting programmes deduct credits according to the uncertainty, so that the credited amount is a conservative lower estimate, and withhold buffer credits for reversal risk (see Carbon Markets, Offsets and Payments for Ecosystem Services).
Independent verification checks sampling locations and procedures, laboratory chain of custody, model inputs against farm records, and calculations, often re-sampling a subset of points. National inventories report soil carbon change in the cropland and grassland categories using Tier 1, 2 or 3 methods; companies report it in their land-sector accounting, separately from emissions (see Carbon Accounting and GHG Inventories). Consistency matters: the same practice change should not be claimed by the farmer for credits, by a food company in its Scope 3 removals and by the country in its NDC without clear rules on who claims what.
Verifiers and reviewers see the same errors repeatedly. Fixed-depth comparisons after a change in tillage, without equivalent soil mass correction, create spurious gains. Bulk density is assumed or estimated by pedotransfer function in one round and measured in another. Sampling locations drift between rounds, so the measured change includes spatial variation. Laboratory methods change between rounds -- Walkley-Black at baseline, dry combustion later -- without calibration, producing an apparent increase of the order of the method's incomplete recovery. Carbonate carbon is not removed in calcareous soils. Models are applied outside the conditions for which they were calibrated, or are given yield and residue inputs that do not match farm records. Emissions that rise with the practice, such as N2O from more nitrogen or from legume residues, are left out. Manure imported from elsewhere is counted as net sequestration. Early-year rates are extrapolated over the whole crediting period. Each of these errors inflates the claimed removal, and most can be avoided by following an established protocol such as FAO's GSOC-MRV protocol or the requirements of a reputable crediting programme.
The '4 per 1000: Soils for Food Security and Climate' initiative was launched by France at COP21 in Paris in 2015. Its name comes from a simple calculation: increasing the carbon stock in the world's agricultural (or all) soils by 0.4% per year in the top 30-40 cm would, in principle, be comparable in magnitude to the annual increase in atmospheric CO2 from fossil fuels. The initiative brings together governments, research bodies, farmer organisations and companies around voluntary action and research on soil carbon, emphasising food security and soil health as well as climate.
The initiative has been widely welcomed for raising the profile of soils, but scientists have criticised taking the 0.4% figure as a realistic target. A uniform 0.4% annual increase on all soils is not achievable: many soils are already near equilibrium or saturation, gains slow over time, some regions lack the biomass inputs needed, building SOC requires nitrogen and phosphorus (which carry their own emissions), and gains are reversible. The initiative itself frames the figure as an aspirational indicator rather than a binding target. It remains a useful teaching example of how a large stock times a small percentage gives a large number -- and why such arithmetic must be checked against biophysical limits.
Many countries support soil carbon through policy. Australia's carbon crediting scheme includes soil-carbon methods under which farmers can earn credits (ACCUs) for measured or modelled increases. The European Union adopted a Carbon Removals and Carbon Farming certification framework in 2024, setting quality criteria for removals including carbon farming. In the United States, conservation programmes administered by the Natural Resources Conservation Service fund practices such as cover crops and reduced tillage, and the national inventory uses DayCent to estimate soil carbon change. India's Soil Health Card scheme (launched in 2015) provides farmers with soil test results, including organic carbon, and fertiliser recommendations, and programmes on natural and organic farming promote organic inputs. FAO hosts the Global Soil Partnership, produced the Global Soil Organic Carbon Map (2017) from national contributions, and runs the RECSOIL (recarbonisation of global soils) programme.
These programmes differ in whether they pay for practices or for measured outcomes, how they verify, and how they handle permanence. Practice-based payments are simpler and fairer to small farmers but give less certainty about tonnes; outcome-based payments give more certainty but shift measurement cost and risk onto farmers.
Soil inorganic carbon (SIC) is carbon in carbonate minerals, chiefly calcite (CaCO3) and dolomite. It is abundant in arid and semi-arid soils, where it accumulates as pedogenic carbonate in calcic horizons, and in soils formed from limestone. Global estimates put SIC in the top metre in the order of several hundred Gt C, comparable in size to SOC in drylands. Lithogenic carbonate inherited from parent rock does not represent new sequestration; pedogenic carbonate formed by reaction of CO2 with calcium from silicate weathering or from dust and irrigation water can represent net capture, while carbonate re-precipitated from dissolved limestone does not.
Carbonates also matter for management and measurement. Acidification -- for example from nitrification of ammonium fertilisers -- can dissolve soil carbonate and release CO2. Agricultural lime (limestone or dolomite) applied to correct acidity releases CO2 as it dissolves, which inventories count at 0.12 or 0.13 t C per tonne. In calcareous soils, organic carbon analysis must exclude carbonate carbon, otherwise SOC will be overestimated.
Enhanced rock weathering (ERW) spreads finely crushed silicate rock, such as basalt or rocks rich in olivine, on farmland to speed up the natural weathering reaction that consumes CO2. As the minerals dissolve, CO2 dissolved in soil water is converted to bicarbonate, which can be carried to rivers and the ocean, where much of it remains for very long periods. For the magnesium silicate forsterite the reaction is Mg2SiO4 + 4CO2 + 4H2O -> 2Mg2+ + 4HCO3- + H4SiO4, so in theory one tonne can capture about 1.25 t CO2; basalts have lower theoretical potentials, and realised capture is lower still because weathering is slow and partial and some bicarbonate may degas or precipitate. Weathering of carbonate rock (CaCO3 + CO2 + H2O -> Ca2+ + 2HCO3-) captures CO2 temporarily, but over ocean timescales precipitation returns it, which is why liming is treated as a source.
ERW may bring co-benefits -- raising soil pH, supplying nutrients such as calcium, magnesium and potassium, and possibly increasing yields -- and its storage is far more durable than SOC. Its challenges are MRV (tracking dissolution and the fate of cations and bicarbonate through soils and rivers), slow rates in cool or dry soils, the energy and emissions of grinding and transport, and potential release of trace metals such as nickel and chromium from some ultramafic rocks. Field trials are underway in several countries, and carbon-removal programmes have begun issuing credits under early methodologies; the evidence base is still developing.
Master the stock calculation (concentration x bulk density x depth), the first-order model and its equilibrium (C* = I/k), mean residence time, the Hassink saturation equation and the difference between POM and MAOM. Know why no-till gains often shrink with deeper sampling, why manure may only move carbon, why sequestration slows and can reverse, and how dry combustion, Walkley-Black and spectroscopy differ. Be able to explain the 4 per 1000 arithmetic and its critics, and write the forsterite weathering reaction.
As a soil-carbon project officer, verifier or agronomist you design sampling plans, collect and process cores, choose laboratories and methods, run or check models, and prepare MRV reports. Daily decisions include stratification, sample numbers, depth, how to handle stones and carbonates, whether a model is valid for the region, and how to explain to farmers what a contract requires and what gains are realistic. You also advise on practices -- residue retention, cover crops, grazing, agroforestry -- with honest expectations.
As a programme manager, senior scientist or policy adviser you decide whether to pay for practices or outcomes, which MRV standard to accept, how to price permanence and uncertainty, and how soil carbon fits national targets and corporate claims. You weigh soil carbon against N2O and methane, judge claims from new ventures, protect farmers from unrealistic contracts, and explain to boards and ministers why soil carbon is valuable for soil health, water retention and adaptation but limited as an offset for fossil emissions.
As a researcher you work on persistence mechanisms (MAOM formation, microbial necromass, deep carbon), saturation and its measurement, cheaper and more accurate measurement (spectroscopy, in-field sensors, bulk density), model improvement and uncertainty, the climate value of temporary storage, and the evidence for enhanced rock weathering. You maintain and analyse long-term experiments, contribute to IPCC methods and national inventories, review MRV standards, advise governments designing carbon-farming programmes, and train the soil scientists and verifiers who will apply them in the field.
Sequestration rates are given as broad ranges; the manual does not offer a single rate for any practice because rates depend on climate, soil, history, depth and duration, and reviews disagree. It does not reproduce IPCC stock-change factor tables or reference stocks, which vary by climate zone and soil class. Peatland and wetland carbon are mentioned but not developed (see wetlands-coastal-blue-carbon), nor permafrost. Programme details (Australian methods, the EU certification methodologies, national schemes) change and are described as of 2024-2026. Enhanced rock weathering evidence is early-stage and summarised cautiously. Biochar is treated in Agricultural GHG Emissions and Mitigation.
Original practice questions written to the syllabus and pattern of the exams named in each question's examTags. They are not past papers and do not reproduce any board's questions. Every numerical answer was computed by script; Rates, factors and deductions are stated in each stem because they differ between soils, climates, programmes and guideline versions.
These notes are free to read here, with no account.
What the Laisha app adds:
Free for ten days, then a subscription. We would rather write that here than have you find it later.