The Case for Carbon Dioxide Removal Technologies

This is a controversial subject especially when it involves spending public money on technological solutions. This week’s Climate Uncovered takes a look at why it’s needed to meet the Paris goals.

The only way to limit temperatures to +1.5ºC, or even +2ºC, by the end of the century is to remove and permanently store a significant quantity of the greenhouse gases accumulating in the atmosphere. To manage an overshoot, where temperatures exceed these limits and are brought back down by 2100, will require even more carbon dioxide removal.

Estimating how much needs to be removed is a similar exercise to calculating remaining carbon budgets. They are therefore susceptible to changes in the observed sensitivity of the climate to the level of greenhouse gases that have accumulated and at the mercy of evolving feedback mechanisms.

How much removal is required?

Let’s start at the beginning. In the pre-industrial period, typically defined as 1850-1900, the average annual atmospheric concentration of CO₂ was 278ppm. Through a combination of fossil fuel burning and land use change, we have increased this to 424ppm by 2024. The rate of yearly increase is accelerating as we burn more and more fossil fuels and pressures on nature turn previous carbon sinks (absorbers) into carbon sources (emitters).

According to the IPCC, the ‘safe’ limit is 350ppm¹. This level would stabilise temperatures at the +1.5ºC level required to avoid the worst impacts of climate change. The last time the level was this low was 1990. So basically we not only need to stop making new emissions but we also need to remove all the CO₂ emitted over the last 35 years and permanently lock it away.

To give a financial analogy of the problem, imagine you have been growing your career for 35 years, earning more money each year and becoming accustomed to a comfortable lifestyle. Out of the blue, the Revenue Service realises you have not been paying any income tax for the whole time and sends you two demands. First you need to start paying tax on your future earnings and second you need to pay back-taxes which have accumulated over the 35 years. Taxing your current earnings is equivalent to achieving net-zero. Paying your back taxes is equivalent to carbon dioxide removal (CDR).

There is a slight issue with this 350ppm level however. Recent data has strongly suggested that the climate’s sensitivity to CO₂ is higher than the IPCC’s best estimate. They use a response value of 3ºC for a doubling of atmospheric CO₂ (termed ECS) whereas the current data is suggesting a value of 4.5ºC. This would imply that the safe limit is actually closer to 325ppm. The last time the level was this low was 1970, 55 years ago.

We could take this even further. A recent paper suggested that in order to prevent ice sheet collapse and catastrophic sea level rise, temperatures need to return to just +1ºC. This would set the limit at 325ppm for an ECS of 3ºC or 310ppm for an ECS of 4.5ºC (1970 or 1950 levels).

Using 350ppm as the target is mind boggling enough for now though, so we’ll use this target from now on, baring in mind assumptions for its effectiveness may be quite optimistic.

The IPCC scenarios known as shared socio-economic pathways (SSPs) suggest a range of futures depending on both future policy and emissions. SSPs are identified by two numbers, firstly a socio-economic family from 1 to 5 with 1 being sustainability and 5 being heavy fossil fuel development. 2 is the middle of the road pathway. The second number refers to the level of mitigation deployed and roughly equates to the resultant climate forcing in W/m2 (but don’t worry about that too much, just note that the smaller the number the greater the mitigation effort implied and the larger the number the warmer the planet will become).

This could produce a massive number of scenarios so just 5 tend to be concentrated on. The full scenarios are highly detailed and consider other GHGs as well as biogenic feedbacks such as permafrost response and land use implications (you can read more here²), but in a nutshell:

SSP1-1.9 is all out rapid global decarbonisation and removals for +1.5ºC with limited overshoot (already blown so no longer relevant).

SSP1-2.6 is a +2ºC compatible scenario with a lot of mitigation and carbon removal. It is the only viable scenario for staying within the Paris Agreement commitment (highly unlikely unless policies radically improve)

SSP2-4.5 is the scenario we are most closely tracking to date and points towards the widely reported +2.7-3.2ºC rise by 2100. It includes peaking of emissions in 2030-35 but not achieving net-zero by 2100 and not stopping the continued warming after 2100.

SSP3-7.0 is where the fossil fuel lobby are urging their bought politicians to head towards, no carbon removal, little reduction in emissions, peaking only in late century and temperatures heading through +3.5ºC within 75 years.

SSP5-8.5 is now thankfully unlikely but would have seen all out continued expansion of all fossil fuel use with no mitigation effort at all and temperatures heading through +4.2ºC within 75 years.

Let’s imagine that the world gets fed up of the increase in climate related loss and damage from storms, heatwaves, droughts and accelerating sea level rise and decides to comply with the Paris Agreement and return temperatures to +1.5ºC by the end of the century following a now unavoidable overshoot. The IPCC estimate that assuming net-zero was first achieved by 2060, a total of 360 gigatonnes of CO₂ would need to be removed from the atmosphere by 2100 to meet this goal.

This estimate was included in the IPCC’s AR6 WG3 report of 2022 but looking at the rest of the scenario is probably an underestimate given the trajectory since then. It has emissions peaking by 2025 (very unlikely), future emissions before net-zero of just 720 GtCO₂ (compared to 2,790 GtCO₂ under the current best fitting SSP2-4.5 scenario), peak warming of 1.7ºC (very low given the last 24 months averaged 1.6ºC) and still only gives a 24% chance of success.

But OK, let’s work with these numbers. 360 GtCO₂ removal from 2060 to 2100 is 9 billion tonnes per year of net removals. Don’t forget that carbon capture will already be a part of net-zero to balance difficult to eliminate emissions from cement, steal etc., which doesn’t count towards removals. Taking that into account, let’s round it up to a total carbon capture and removal requirement of 10 GtCO₂ per year from 2060.

That’s a big challenge, but I’m sorry to say it’s still not quite the whole story. In calculating the requirements, the IPCC made some other assumptions.

They included an assumption that agriculture, including land use change and the whole food system which currently emits about 4 GtCO₂ per year would become a sink over the next 30 years, instead removing 5 GtCO₂ per year. This is incredibly ambitious but lets be generous and meet them halfway. Say agriculture achieved net-zero. That still adds 5 GtCO₂ to our removals requirement.

Current Natural Carbon Sinks

The IPCC also assumed that intact nature on land would pitch in and help by removing an additional 10.6 GtCO₂ per year and that stable oceans would also remove an additional 11.4 GtCO₂ per year.

This was a realistic assumption since its well known that plants grow more and therefore take-up more CO₂ in a richer CO₂ environment. There is also the expanding greening into higher latitudes which should increase vegetation based carbon sinking.

Unfortunately though, forest fires have also expanded with warming temperatures and are outpacing the increased growth and greening. According to the World Resources Institute’s Global Forest Review³, the world lost 6.7 million hectares of primary rainforest in 2024 emitting 3.1 GtCO₂, roughly the fossil fuel emissions of India for the same year. Canadian forest fires in 2025 are already exceeding record levels. A typical forest fire eliminates 100 years worth of carbon sink from the area destroyed.

In an analysis of the Earth’s annual carbon dioxide cycle, since the 1980s natural absorption was increasing as temperatures rose, but it peaked in 2008 and has been declining at an increasing rate since then⁴. Many countries that were relying on forest growth within their Nationally Determined Contributions under the Paris Agreement are having to recalculate as their forest areas become sources rather than sinks. Finland is an example of this natural decline, but is not alone.

Although increased greening is occurring at a global scale, the new green is not as good as the old green at absorbing carbon. While primary forest is being lost, herbaceous and scrublands are increasing. The chart below shows the changes in land use and land cover conversions since 2000⁵⁶. The big winners in terms of increased area are lower grade, less absorbing ecosystems, meaning that overall carbon sinks are still declining despite overall area increase.

On the ocean side, a recent paper⁷ has suggested that as the coral reefs die off, this could lead to an additional take-up of CO₂ from the oceans by as much as 5% this century since as corals grow, they release CO₂ to the atmosphere. I’m sure you’ll agree that this is not a desirable result, but it should be considered in the predictions.

Given the decline in natural carbon sinks observed today, we need to add a further 20 GtCO₂ to our total removal requirement since on its own, nature is not able to make its hoped for contribution. This takes us to 25-35 GtCO₂ per year depending on the contribution of Agriculture.

Storage Durability

One more consideration is required before we can look at the range of potential solutions. A study by Brunner et al⁸ examined the subsequent warming for different durations of carbon storage once net-zero was achieved. In their analysis they used SSP1-2.6 achieving net-zero in 2060 at a temperature of +2ºC. They only looked at the storage of residual CO2 emissions required to reach net-zero, so the hard to abate sectors like steel and cement manufacture, rather than removal, but examined four scenarios.

The worst case or control was direct re-release, effectively not capturing the remaining emissions or very short term storage like planting a tree which is subsequently cut down or burned within a decade or two. They then looked at 100 year, 1,000 year and permanent durability of storage. The 100 year tracked closely to the re-release showing it was effectively a waste of effort, only delaying the subsequent warming. 1,000 years was close to permanent and effectively prevented further temperature rise.

The conclusion is that to be effective, storage of removed carbon dioxide needs to be durable for 1,000 years to be effective from a climate mitigation perspective.

Nature Based Solutions

Firstly, restoring and protecting ecosystems of all types is essential in its own right. The rate of habitat and ecosystem pollution and destruction currently underway can only be described as a crime against both nature and humanity. In considering its role in the context of climate mitigation, I’m treating carbon removal as a co-benefit rather than as the prime reason for doing it.

Reforestation

99% of current carbon credits for carbon capture are traded in Afforestation/Reforestation (AR). In its simplest form this involves re-planting trees where there were previously forests that have been cut down, and planting trees in new areas, including at higher latitudes.

A 2024 paper⁹ examined the effect of ambitious AR on the temperature overshoot and eventual 2100 temperature and returned promising results but also revealed the scale of the task.

Using restoration potential maps and biodiversity constraints they reached a potential AR area of 595 Mega Hectares by 2060 reaching 935 Mha by 2100 above the 2015 baseline. This occurs at the expense of heavily managed pastures, which decrease by 575 Mha and more lightly managed rangelands, which decrease by 360 Mha. The scenario included reforesting 514 Mha of previously deforested areas globally, corresponding to 60% of total historical deforestation since 1850. The remaining 421 Mha are afforestation, which is employed by priority over the less biodiverse rangelands.

Modelling the impact and taking into account other factors such as effects on ocean absorption rates as atmospheric concentrations drop, the net contribution to the AR scenario is 3.32 GtCO₂ per year averaged to the end of the century. The rate starts higher reaching 7.1 GtCO₂ per year during peak growth before the new ecosystems mature and stabilise.

The impact is promising however supporting a reduction of peak temperature by 0.08ºC, 2100 temperature by 0.2ºC, and overshoot duration by 13 years.

Biochar

Biochar is a potential nature based solution. This is a form of carbonised biomass made from crop residues which is used to fertilise fields fixing the carbon that the plants sank from the atmosphere while growing into the soil. A 2023 study¹⁰ looked at the total global potential for Biochar with storage durability of more than 100 years concluding that 2.6 GtCO₂ could be sequestered this way each year. However, when considering limitations on sustainable residue harvesting and competing livestock usage, the global biochar production potential decreases to 1.9 GtCO₂ with 1.3 GtCO₂ per year remaining sequestered after a century. They did not look further to see what remained after 1,000 years.

Ocean Iron Fertilisation & Artificial Whale Poo

Oceanic phytoplankton require iron to help them grow. By treating the ocean waters with added iron, blooms can be encouraged, which when they die, sink to the ocean floor to be permanently stored. Some obviously enters the food chain, but eventually a good proportion is sequestered.

There are some doubts about this approach, principally that the blooms in iron laced waters use up other nutrients more quickly leading to a crash. The depletion of the other nutrients can also inhibit natural blooms elsewhere within the circulating ocean system. Whether the total primary productivity in the whole ocean basin is improved by the intervention has therefore been questioned. As such it’s hard to put a contribution to it in terms of potential GtCO₂ per year.

When whales feed they do so at depth but poo on the surface bringing nutrients with them. Plankton blooms and fish nurseries form which absorb CO₂ then when they die, sink back down to the depths. It’s been proposed that by distributing artificial poo in the form of rice husks for example, similar capture and storage can be achieved.

I’ve included them here for completeness and I’ll let you investigate further if you’re interested.

Nature Based Contribution

In summary, a highly ambitious afforestation campaign transforming pastureland which would involve a global swing away from meat consumption, reversing deforestation and replanting 60% of forests lost since 1850 and a global commitment to Biochar use within agriculture could get us roughly 4.6 GtCO₂ per year towards the 25-35 GtCO₂ per year requirement to meet the Paris Agreement commitments.

The balance will have to be made up with technological solutions.

Technological Carbon Dioxide Removal

I’ll just summarise these here as the objective of this post is to explore the need for rather than describe CDR technologies.

Three core methods are proposed and in some cases implemented, albeit at small scale. Carbon Capture and Storage (CCS) involves removing CO₂ from high concentration gas streams such as flue stacks. The captured CO₂ is then stored in some way, normally through geological storage underground. This can be used for achieving net-zero from, say steel processing, or if used with the burning of biofuels to act as a removal from the atmosphere referred to as BECCS (Bio-energy with carbon capture and storage). Dracs power station in the UK is reportedly pursuing this route.

The photo below shows a CCS prototype built at Ferrybridge coal fired power station in the UK in 2012. It could remove 100 tonnes of CO2 a day at a representative cost of ~8% of the generated power. The scale of the pilot is illuminating especially when you consider it was fed by just 1% of the stack’s flow. It’s probably not scalable for fossil energy generation, but it is suitable for blue hydrogen, steel and cement production.

Direct Air Capture (DAC) involves a similar process but rather than taking CO₂ from a concentrated stream, taking it from the atmosphere directly. It’s much less energy efficient due to the lower concentrations, but is being pursued as a scalable solution. Climeworks in Iceland is the notable prototype.

The third and most controversial is marine CDR. I’ve mentioned iron fertilisation above which is a biological approach, but from a technological perspective, Ocean Alkalinity Enhancement is proposed as a method to allow the oceans to absorb more CO₂.

There are other suggestions like burying trees underground or sinking them in the oceans, but not that can deliver in the billions of tonnes per year scale.

It’s a combination of these main approaches that will be needed to remove the 20-30 GtCO₂ per year requirement if humanities decides to comply with the Paris Agreement. This is in addition to maximal reforestation and biochar implementation.

Conclusions

The scale of the task ahead has very unfortunately reached a point where nature alone, despite highly ambitious scenarios, cannot help us comply with the Paris Agreement commitments and hold temperatures to well below +2ºC this century. We have simply pumped too much CO₂ into the atmosphere.

If we want to stop the warming, we will need to capture and permanently store all the residual emissions which will involve technological carbon capture and storage solutions. If we want to reduce temperatures from an overshoot we will need to involve direct carbon removal technologies.

There is a moral hazard in that developing these technologies may create an excuse for continued fossil fuel burning, but that is a risk we will have to take.

Thirty years ago it may have been possible to drive towards an early net zero and nature based solutions would have been a practical way to achieve our climate aims, but alas this is no longer an option.

Much has been made of Climeworks recent struggles¹¹, but it’s essential that work like this is supported and continues to scale. It’s a prototype, a trailblazer for a vital part of our future. Critics have pointed out that it emitted more CO₂ in its manufacture than it has captured to date, but that’s not the point.

The Wright Brother’s first powered flight in 1903 covered just 37 meters. They could have build hundreds of bicycles in the time it took them to build the flyer which would have done a better job of travelling 36m. It was a further 6 years before Louis Blériot flew across the English Channel. Imagine if critics had won out and flight abandoned as a waste of time - OK with hindsight that might have been a good idea, but you know what I mean.

Love it or hate it, we can’t do without Carbon Capture and Storage or Carbon Dioxide Removal, so we might as well get on with developing it. While it is in a market failure mode, that means spending public money until policies are enacted through carbon taxes or other mechanisms to make it viable economically. The alternative is a 3ºC plus world which will be bad.

1. IPCC Sixth Assessment Report Working Group III: Mitigation of Climate Change. https://www.ipcc.ch/report/ar6/wg3/downloads/

2. Meinshausen, M.et al.: The shared socio-economic pathway (SSP) greenhouse gas concentrations and their extensions to 2500, Geosci. Model Dev., 13, 3571–3605, https://doi.org/10.5194/gmd-13-3571-2020 , 2020.

3. Forest 2024 report : https://gfr.wri.org/latest-analysis-deforestation-trends

4. James C. Curran, Samuel A. Curran, Natural sequestration of carbon dioxide is in decline: climate change will accelerate 2025 https://rmets.onlinelibrary.wiley.com/doi/10.1002/wea.7668

5. Hou, Z., Zhang, L., Peng, J. et al. Radiative forcing reduced by early twenty-first century increase in land albedo. Nature 641, 1162–1171 (2025). https://doi.org/10.1038/s41586-025-08987-z

6. Weiskopf, S.R., Isbell, F., Arce-Plata, M.I. et al. Biodiversity loss reduces global terrestrial carbon storage. Nat Commun 15, 4354 (2024). https://doi.org/10.1038/s41467-024-47872-7

7. Lester Kwiatkowski et al. Declining coral calcification to enhance twenty-first-century ocean carbon uptake by gigatonnes, 2025 https://doi.org/10.1073/pnas.2501562122

8. Brunner, C., Hausfather, Z. & Knutti, R. Durability of carbon dioxide removal is critical for Paris climate goals. Commun Earth Environ 5, 645 (2024). https://doi.org/10.1038/s43247-024-01808-7

9. Moustakis, Y., Nützel, T., Wey, HW. et al. Temperature overshoot responses to ambitious forestation in an Earth System Model. Nat Commun 15, 8235 (2024). https://doi.org/10.1038/s41467-024-52508-x

10. Shivesh Kishore Karan, Dominic Woolf et al. Potential for biochar carbon sequestration from crop residues: A global spatially explicit assessment 2023 https://doi.org/10.1111/gcbb.13102

11. https://www.theguardian.com/environment/2025/may/17/swiss-firm-that-captures-carbon-from-air-to-cut-workforce-by-more-than-10