This is in addition to steep and immediate emissions cuts.<\/p>\n
Scientists expect carbon emissions to push warming beyond 1.5C<\/a> in the decade ahead, meaning that the target can only be achieved \u201cfrom above<\/a>\u201d via large-scale CDR that brings down global temperatures.<\/p>\n These temperature trajectories are known as \u201covershoot\u201d pathways.<\/p>\n The third \u201cstate of CDR\u201d report, written by more than 50 scientists, says that countries\u2019 current CDR plans would fall short of what is needed to limit warming to 1.5C by more than 5bn tonnes of CO2 (GtCO2) per year by 2050.<\/p>\n Global CDR would have to increase fourfold \u2013 from 2.2GtCO2 in 2026 to 8.75GtCO2 by 2050 \u2013 to have a chance of meeting the 1.5C target by 2100, according to the report.<\/p>\n It adds that deploying CDR can be a \u201cgradual process\u201d, making the period 2026-30 \u201ccrucial\u201d for \u201cestablishing CDR\u2019s role in limiting climate damages\u201d in the future.<\/p>\n Below, Carbon Brief covers the key findings of the third state of CDR report. (This follows from Carbon Brief\u2019s coverage of the first<\/a> report in 2023 and second<\/a> report in 2024.)<\/p>\n What is CDR?<\/p>\n According to the report, the definition of CDR is:<\/p>\n \u201cHuman activities capturing CO2 from the atmosphere and storing it durably in geological<\/a>, terrestrial or ocean reservoirs, or in products. This includes human enhancement of natural removal processes but excludes natural uptake not directly caused by anthropogenic [human-caused] activities.\u201d<\/p>\n<\/blockquote>\n In addition to this, the report includes \u201cthree key principles\u201d for CDR, which are:<\/p>\n In this report, a CDR method is considered durable if it is able to lock up carbon for \u201cdecades or more\u201d.<\/p>\n The report classifies CDR techniques as either \u201cconventional\u201d or \u201cnovel\u201d.<\/p>\n \u201cConvential\u201d CDR techniques are \u201cwell established, already deployed at scale and widely reported by countries as part of [land-use] activities\u201d.<\/p>\n The methods included in this group are tree-planting, ecosystem restoration, agroforestry (trees in agriculture), improving soil carbon in croplands and natural lands, and durable wood production.<\/p>\n \u201cNovel\u201d CDR techniques have \u201clower level of readiness for deployment and, as a consequence, are currently deployed at smaller scales\u201d, says the report.<\/p>\n Some examples of different CDR methods are listed on the graphic below.<\/p>\n The graphic also shows whether carbon is captured through biological or chemical processes, as well as how \u201cready\u201d the method is and for how long it can store carbon, among other features.<\/p>\n The report says that CDR is \u201cneeded alongside deep and rapid emissions reductions\u201d to give Earth a chance of limiting global warming to 1.5C. It continues:<\/p>\n \u201cIt should play a smaller role than emissions reductions given uncertainty around the feasible levels of scaling, sustainability limits, storage availability and the risk of reversal, among other constraints.\u00a0<\/p>\n<\/blockquote>\n \u201cIn general, CDR should be seen as a limited resource that will need to be used prudently.\u201d<\/p>\n<\/blockquote>\n It adds that CDR can \u201cfulfil three major functions\u201d.\u00a0<\/p>\n In the near term, CDR can help reduce \u201cnet emissions\u201d, it says.<\/p>\n In the medium term, CDR can \u201ccounterbalance residual emissions\u201d to achieve net-zero CO2 or net-zero greenhouse gas emissions, the report continues.\u00a0<\/p>\n (\u201cResidual emissions\u201d are those that cannot be eradicated through technologies or societal changes, such as methane emissions from rice production.)<\/p>\n Research<\/a> suggests that global warming is likely to stop, more or less, once net-zero is achieved globally.<\/p>\n In the long term, CDR can \u201chelp achieve net-negative emissions<\/a>\u201d, a state where CO2 removal exceeds emissions, says the report.<\/p>\n In this state, humans could lower global temperatures. This may allow the world to limit global warming to 1.5C by 2100, even if the temperature target is surpassed earlier on in the century.\u00a0<\/p>\n Future trajectories where temperatures exceed the 1.5C limit before being brought back down again through CDR techniques are known as \u201covershoot<\/a>\u201d pathways.<\/p>\n Back to top<\/a><\/p>\n What are current levels of CDR?<\/p>\n The report says that, at present, \u201c99.9%\u201d of existing CDR is conventional, land-based techniques such as tree-planting and ecosystem restoration.<\/p>\n The world currently removes 2.2GtCO2 per year, equivalent to around 5% of gross global CO2 emissions, it continues.<\/p>\n The largest contributors to removing CO2 from the atmosphere are China, the US, the EU, Brazil and Russia.<\/p>\n The chart below shows the amount of CO2 removed each year over 2014-23 by the largest contributors, through tree-planting (afforestation) and forest restoration (reforestation).<\/p>\n \u201cNovel\u201d CDR, such as biochar<\/a> and direct air capture<\/a>, currently removes just 2m tonnes of CO2 annually at present, according to the report.<\/p>\n However, these methods have been growing at a rate of 40% per year \u2013 \u201csimilar to successful technologies like solar energy, but insufficient for the scale-up required to meet the Paris temperature goal\u201d, says the report.<\/p>\n The graphic below illustrates how the contribution of conventional CDR currently dwarfs novel CDR, but how the latter techniques are quickly growing.<\/p>\n The report says that investment in CDR companies recovered in 2025 following a dip \u2013 and its \u201cshare of all climate-tech funding\u201d grew to 2.6%.<\/p>\n The report also notes that, at present, most CDR efforts are unevenly distributed across the world.<\/p>\n For example, two-thirds of conventional CDR in voluntary carbon markets is in Latin America, according to the report. (Voluntary carbon markets<\/a> are where companies can buy credits for carbon-reducing or removing projects, such as tree-planting, to claim that they have \u201coffset\u201d some of their own emissions.)<\/p>\n In addition, most pilot projects that aim to demonstrate novel CDR methods are located in only a few countries, such as Sweden, Denmark and the US, says the report.<\/p>\n The chart below shows the location and timeline of demonstration projects that have been announced, are under construction or in operation globally.<\/p>\n The report continues:<\/p>\n \u201cWhile first-movers play important roles, if their actions do not diffuse more widely, vulnerability emerges, as evidenced by the impact of US climate policy dismantling.\u201d<\/p>\n<\/blockquote>\n (For more, see: How is policy impacting CDR demand?<\/a>)<\/p>\n Back to top<\/a><\/p>\n How much CDR is needed to reach net-zero goals?<\/p>\n The report examines three scenarios where global temperature rise is limited to \u201cwell below\u201d 2C by 2100:<\/p>\n The pledges considered in the report are \u201cnationally determined contributions\u201d, or NDCs, which countries submit periodically to the UN Framework Convention on Climate Change<\/a> (UNFCCC). NDCs lay out a country\u2019s climate ambition.<\/p>\n Under the current ambition scenario, the report projects a total of 5.9GtCO2 of CDR by 2050 and 12GtCO2 by 2100.\u00a0<\/p>\n This scenario would result in end-of-century warming of 1.7-2.7C. Importantly, the report says, this scenario does not result in the world reaching net-zero CO2 levels, \u201cmeaning that global temperatures would continue to rise, albeit at a much more gradual pace, beyond 2100\u201d.<\/p>\n Under the highest-possible ambition scenario, CDR scales up to 8.8GtCO2 by mid-century and 15.3GtCO2 by the end of the century.<\/p>\n This scenario assumes \u201cfull buy-in by all nations\u201d, with economics, scale-up and sustainability providing the main constraints on CDR deployment, the report says.\u00a0<\/p>\n The highest ambition scenario results in global temperatures peaking at 1.7-1.8C around 2050 and the world achieving net-zero emissions around that time.\u00a0<\/p>\n Under the delayed ambition scenario, CDR would scale up to 7GtCO2 by 2050 and 23.6GtCO2 by 2100. This scenario shows global temperatures peaking between 1.7C and 2.0C.\u00a0<\/p>\n This scenario requires larger CDR deployment in the long term than the highest-ambition scenario does, due to the larger cumulative emissions caused by delaying deep emissions reductions.<\/p>\n In both the high ambition and delayed ambition scenarios, the world reaches \u201cdeeply net-negative CO2 emissions\u201d by 2100, the report says. This continued deployment of CDR will further draw CO2 from the atmosphere, lowering global temperatures back down to 1.5C.<\/p>\n The chart below shows annual global greenhouse gas emissions through the end of the century under current ambition (red), highest ambition (green) and delayed ambition (blue) scenarios.<\/p>\n While global CDR capacity scales up more slowly in the first and third scenarios, the report notes that, in all three cases, \u201cnovel CDR reaches gigatonne-scale deployment by 2050\u201d.<\/p>\n Back to top<\/a><\/p>\n What does the science say about the potential and costs of CDR?<\/p>\n There is a wide range of both carbon-removal potential and associated costs between different methods of CDR, according to the report.<\/p>\n However, it also notes that these numbers \u201crange widely\u201d in the scientific literature.\u00a0<\/p>\n The discrepancies in estimates of carbon-removal potential are due to a number of factors, the report says, including a lack of available scientific data, inconsistencies in the assumptions made in assessing technical feasibility and a lack of agreement on what, exactly, \u201cpotential\u201d means.<\/p>\n These elements also influence the cost of different CDR methods, but additional factors \u2013 such as deployment costs in different areas, technological approaches and scope \u2013 also play a role in establishing price differences. Because of this, the report says, \u201ccost estimates are often difficult to compare across methods, complicating design and policy decisions\u201d.<\/p>\n The chart below shows the reported range of mitigation potential (left) and reported range of costs (right) for different CDR methods. The top four rows indicate conventional CDR methods, while bottom 11 rows show novel CDR methods. The chart refers to \u201cmitigation potential\u201d, rather than removal potential, because some estimates do not distinguish between removals and avoided emissions.<\/p>\n (Avoided emissions<\/a> refers to the difference in emissions from carrying out a project, compared to a hypothetical alternative \u2013 such as the reduced emissions from halting deforestation.)<\/p>\n The darker colours indicate estimates that are more constrained, meaning that they are either based on stricter assumptions or there is more agreement between different estimates.<\/p>\n The report notes that for most removal methods, the low end of the potential is around 1GtCO2 per year, while the upper limit of costs is more than $200\/tCO2.<\/p>\n The least expensive CDR approaches are forestry-based methods, soil-carbon sequestration and biomass burial. For forestry-based methods, the report puts the cost of CDR at $5-$53 per tonne of CO2 removed. Soil-carbon sequestration costs reach as high as $150 per tonne of CO2 removed, but could have negative overall costs \u201cwhen accounting for crop yield increases potentially resulting\u201d from changed farm-management practices, the report says.<\/p>\n However, it adds that \u201cthese CDR methods are typically associated with lower levels of permanence\u201d than other methods.<\/p>\n Other relatively low-cost methods include coastal wetland restoration, biochar, bioenergy with carbon capture and storage<\/a> (BECCS) and enhanced rock weathering<\/a>, while ocean alkalinity enhancement is a medium-cost option.\u00a0<\/p>\n The most expensive methods include direct air carbon capture and storage (DACCS) and direct ocean carbon capture and storage (DOCCS).<\/p>\n The report also notes that a total estimate of CDR removals cannot be obtained by adding up the removal potential of all of the separate methods, since different methods can compete for scarce resources. For example, BECCS, biochar, biomass burial and biomass sinking all rely on the same base input \u2013 biomass \u2013 and therefore cannot all be maximised at the same time.<\/p>\n Back to top<\/a><\/p>\n What have governments pledged on CDR?<\/p>\n While many countries include some amount of CDR in their national climate plans, there is currently a large gap between the amount of CDR pledged in these plans and the amount that will be needed to limit global temperature rise to 1.5C by the end of the century, says the report.<\/p>\n This quantity is referred to as the \u201cCDR gap\u201d \u2013 the difference between what is pledged and what is needed.\u00a0<\/p>\n The size of the CDR gap is dependent not just on the pledges made by countries, but also the choice of the \u201cbenchmark\u201d scenario against which the pledges are measured. Lower \u2013 or delayed \u2013 emissions reductions lead to larger shortfalls in the long term, meaning \u201cCDR must subsequently be scaled to very high levels\u201d, says the report.<\/p>\n Current NDCs and other country submissions to the UNFCCC total 2.5GtCO2 per year of removals in 2030, 2.7GtCO2 per year in 2035 and 3.6GtCO2 per year in 2050.\u00a0<\/p>\n This gives a CDR gap of 0.3GtCO2 in 2030, 1.2GtCO2 in 2035 and 5.2GtCO2 in 2050, according to the report. These figures are obtained using assumed \u201cimmediate, ambitious action at all levels to reduce emissions\u201d and the most-ambitious estimates of CDR set out in national pledges. Together, this provides a \u201clower bound\u201d for the CDR gap, says the report.<\/p>\n By comparison, a 10-year delay in implementing ambitious emissions reductions will result in the need to remove at least an additional 150GtCO2 from the atmosphere, compared to the most ambitious scenario. (See: How much CDR is needed to reach net-zero goals?<\/a>)<\/p>\n The report says that the CDR gap has widened since the second state of CDR report<\/a> was released in 2024, due to the US leaving the Paris Agreement<\/a>. It adds that other countries have \u201cnot delivered a step change in ambition\u201d in their latest round of climate pledges<\/a>.<\/p>\n It also cautions that \u201ccredibility issues with national pledges may mean that the CDR gap is actually larger than what we assess here\u201d.<\/p>\n The report notes that current CDR pledges by companies are \u201csubstantially higher than country pledges\u201d, at 5GtCO2 per year in 2050. However, it adds, \u201ccredibility in these announcements is low\u201d.<\/p>\n Back to top<\/a><\/p>\n What is the current funding and research landscape for CDR?<\/p>\n Funding of CDR research and development \u2013 as well as investment in CDR companies \u2013 has continued<\/a> to increase in recent years.<\/p>\n In total, there has been around $5.6bn in grant funding distributed to CDR research since 2005, according to the report\u2019s analysis. Roughly one-third of this has come in the past three years.<\/p>\n Funding for CDR research grants grew 13% each year between 2022 and 2025, the report says, and the corresponding number of research publications grew at a similar rate.<\/p>\n Funding was largely targeted at a handful of key areas, notably soil carbon sequestration, biochar and forest-based CDR.\u00a0<\/p>\n DACCS and BECCS only make up a small number of active grants, but together account for around two-fifths of all funding due to \u201csubstantially larger\u201d project sizes.<\/p>\n Despite the growth of research grants and scientific publications, the report concludes that early-stage innovation in CDR is \u201cuneven\u201d and says there is \u201cno strong evidence of a step-change\u201d.\u00a0<\/p>\n It notes that much of the support for CDR has come from projects with a broader focus, rather than those that focus specifically on CDR.<\/p>\n The authors also point to a decline in \u201cinventive activity\u201d, as measured by patenting of CDR-related innovations. While patenting for emissions-cutting technologies in general has been on an upward trajectory, CDR patenting peaked in 2011.<\/p>\n Meanwhile, the report highlights the \u201cremarkable\u201d sustained investment in CDR companies, against a backdrop of falling investment in climate-related technologies. It notes that CDR now accounts for around 3% of overall \u201cclimate-tech funding\u201d.<\/p>\n Yet, again, it says future developments remain \u201cuncertain\u201d. Since the previous<\/a> 2024 \u201cstate of CDR\u201d report, companies have scaled back their ambitions and policy reversals \u2013 notably in the US \u2013 \u201cunderscore that funding uncertainty remains a key barrier\u201d. (See: How is policy impacting CDR demand?<\/a>)<\/p>\n An upward tick in funding in 2025 was driven primarily by a \u201csurge\u201d in grants from predominantly public institutions, as well as $0.5bn in debt financing for a single BECCS project<\/a> in Sweden.\u00a0<\/p>\n Reliance on such funding sources \u201chighlight[s] the volatility of the CDR innovation ecosystem\u201d, according to the report.<\/p>\n The report also has a chapter focusing on the voluntary carbon market, which it describes as \u201cpropelling most of the current demand for novel CDR\u201d.<\/p>\n The scale of this market remains fairly small, with contracts for 0.04GtCO2 of removals signed last year.\u00a0<\/p>\n Moreover, the concentration of sales within a small number of buyers \u2013 particularly Microsoft<\/a> \u2013 remains a \u201ccritical vulnerability\u201d, the authors note.\u00a0<\/p>\n Back to top<\/a><\/p>\n How is policy impacting CDR demand?<\/p>\n The report analyses CDR policies in G20 nations \u2013 which together account for three-quarters of global emissions \u2013 to assess how they are acting to support CDR across their economies.<\/p>\n In total, 140 countries have announced<\/a> net-zero targets, including virtually all of the world\u2019s major emitters. In doing so, the report points out that the governments of these nations have \u201cimplicitly included a role for CDR in their climate plans\u201d.<\/p>\n However, this does not always translate into measures specifically designed to scale up CDR.\u00a0<\/p>\n Only the EU has adopted a binding, quantified removals target<\/a> into law \u2013 namely, the goal to reach 310m tonnes of CO2 equivalent (MtCO2e) of annual net removals in the land sector by 2030.<\/p>\n Overall, conventional CDR is the main focus of policy, with various governments focusing on tree planting to absorb CO2 from the atmosphere.<\/p>\n Among G20 nations, only the UK<\/a> and Australia have set specific goals to scale up novel CDR, such as BECCS and DACCS, over the coming decade.<\/p>\n The report highlights some nations, including Canada, Germany, Switzerland and the UK, as taking proactive steps to incentivise CDR.\u00a0<\/p>\n The authors point to national strategies, financial support for CDR and efforts to integrate it into emissions trading systems (ETS) as examples of effective policy making.<\/p>\n (The report also stresses that the US, which was previously a \u201cleader\u201d on CDR, has now \u201cfrozen or dismantled funding and support\u201d for CDR under the Trump administration.)<\/p>\n Most of the successful policies highlighted in the report focus on supporting the supply of CDR, with \u201cless attention so far on creating demand\u201d.\u00a0<\/p>\n This is significant because CDR \u201cgenerally lacks a natural market\u201d, meaning there are not automatically buyers willing to spend money on emissions removals. Therefore, the authors say, policy interventions are important to create markets and boost demand.<\/p>\n \u201cCompliance\u201d carbon credits<\/a> \u2013 referring to<\/a> credits that can be used to meet legally mandated emissions targets \u2013 provide a way to support demand, according to the report authors.\u00a0<\/p>\n Only some ETSs, such as those used in New Zealand and Australia, allow the use of credits based on forest-related removals for compliance. (It is worth noting that such credits are controversial<\/a>, as removals by forests are not always permanent.)<\/p>\n The report also highlights the need for \u201cfoundational policies to create a governance framework for CDR, including rules for quantification of removal, guidelines for community engagement and the minimisation of negative environmental impacts\u201d.<\/p>\n\n
\n
![\"Infographic](\"https:\/\/www.europesays.com\/ie\/wp-content\/uploads\/2026\/06\/Screenshot_2026-06-01_at_17.19.11.png\")
CDR techniques and their characteristics. Credit: Edwards et al. (2026)<\/p>\n\n
\n
![\"Chart](\"https:\/\/www.europesays.com\/ie\/wp-content\/uploads\/2026\/06\/Screenshot_2026-06-02_at_12.10.33-1024x607.png\")
CO2 removed via afforestation and reforestation each year by the world\u2019s largest contributors to current CDR. Credit: Edwards et al. (2026)<\/p>\n![\"Infographic](\"https:\/\/www.europesays.com\/ie\/wp-content\/uploads\/2026\/06\/Screenshot_2026-06-02_at_09.57.47-1024x727.png\")
A graphic illustrating the contribution of \u201cconventional\u201d and \u201cnovel\u201d to current CDR methods. Credit: Edwards et al. (2026)<\/p>\n![\"Chart](\"https:\/\/www.europesays.com\/ie\/wp-content\/uploads\/2026\/06\/Screenshot_2026-06-02_at_10.20.46-1024x750.png\")
Location and timeline of demonstration projects that have been announced, are under construction or in operation globally. Credit: Edwards et al. (2026)<\/p>\n\n
\n
![\"Annual](\"https:\/\/www.europesays.com\/ie\/wp-content\/uploads\/2026\/06\/Screenshot_2026-05-29_16.13.26.png\")
Annual emissions, in GtCO2e per year, for the three scenarios: current ambition (red), highest ambition (green) and delayed ambition (blue). Source: Edwards et al. (2026)<\/p>\n![\"Annual](\"https:\/\/www.europesays.com\/ie\/wp-content\/uploads\/2026\/06\/Screenshot_2026-06-01_10.06.45.png\")
Annual mitigation potential (left) and cost range per tonne of CO2 (right) for conventional and novel CDR methods. Orange bars indicate the range of values reported, with darker colours indicating less uncertainty about the estimates. Source: Edwards et al. (2026) <\/p>\n