Biodiversity and Geoengineering

01Learning objectives

After completing this chapter, you should be able to:

• Explain the scale and nature of contemporary biodiversity loss, distinguishing between population declines, local extinctions, and global extinctions.
• Evaluate competing arguments for why biodiversity matters, drawing on instrumental, intrinsic, and Singapore-specific perspectives.
• Distinguish between Solar Radiation Management and Carbon Dioxide Removal, and assess where each currently sits on the spectrum from research to deployment.
• Argue a position on whether geoengineering should be pursued, accounting for moral hazard, termination shock, and governance gaps.
• Connect the biodiversity and geoengineering debates to the deeper question of who has the power to remake the natural world, and who bears the consequences.

02Key terminology

Biodiversity

The variety of life on Earth, including diversity within species (genetic), between species, and of ecosystems.

Sixth Mass Extinction

The proposed name for a current human-caused extinction event comparable to the five major extinction events in the geological record.

Ecosystem services

The benefits humans receive from functioning ecosystems, including pollination, water purification, climate regulation, and disease control.

Local extinction (extirpation)

The loss of a species from a specific area, while the species persists elsewhere.

Geoengineering (climate engineering)

Deliberate, large-scale intervention in the Earth's climate system to counteract climate change.

Solar Radiation Management (SRM)

Techniques that reflect sunlight back into space, including stratospheric aerosol injection and marine cloud brightening.

Carbon Dioxide Removal (CDR)

Techniques that draw carbon dioxide out of the atmosphere, including direct air capture, afforestation, and ocean fertilisation.

Termination shock

Rapid temperature rebound that would occur if SRM were deployed and then suddenly stopped without underlying emissions reductions.

Moral hazard

The concern that the perceived availability of a technological climate fix could weaken political will to cut emissions.

03Introduction: a note from the author

Mount Pinatubo erupted in June 1991. Twenty million tonnes of sulphur dioxide reached the stratosphere. For the next two summers, global temperatures ran about half a degree cooler than they should have. No treaty. No vote. No consent. The planet had run an unintended experiment on its own atmosphere, and it had worked.

Thirty-one years later, an American entrepreneur named Luke Iseman launched two helium weather balloons over Baja California, Mexico, each containing a few grams of sulphur. Six months after that, he incorporated the work as a company called Make Sunsets and began selling versions of the same effect online for ten dollars a gram. Mexico banned him within months.

In Singapore, meanwhile, contractors are tunnelling seventy metres beneath the country's last serious patch of rainforest. The depth was chosen as a kind of penance, an expensive apology for a piece of infrastructure that had to be built somewhere.

Three stories, three different answers to the same question. How much should humans remake the natural world, and on whose authority?

The diagnosis: how much are we actually losing?

In May 2024, the seas around Singapore's southern islands hit 31.7°C, the highest temperature recorded by the country's marine monitoring network. Within weeks, the corals around Pulau Hantu, Pulau Satumu, Kusu Island, and Sisters' Islands began to turn white. By July, around 44 per cent of monitored colonies had bleached. Most recovered by November. But the species Pocillopora acuta and Pachyseris speciosa did not recover evenly. The reefs survived this round.

The frequency is the problem. Mass bleaching events in Southeast Asia once happened every twenty-five to thirty years. They now happen on average every six. The corals will not always recover. At some point, a year's bleaching event will be followed by another one before recovery is possible, and reefs that survived 2024 will not survive 2030.

Now widen the camera.

The WWF's 2024 Living Planet Report put the figure at 73 per cent: the average decline in monitored wildlife populations between 1970 and 2020. The IPBES Global Assessment estimated in 2019 that one million plant and animal species face extinction. The IUCN Red List Index has now deteriorated by more than 12 per cent since 1993.

These numbers carry weight. They are also routinely misunderstood, including by the people who quote them.

Take the 73 per cent figure. It does not mean that 73 per cent of all wildlife has disappeared. It is the geometric mean of population change across roughly 35,000 monitored populations of 5,495 vertebrate species. A small population that crashed by 90 per cent counts the same as a large population that crashed by 90 per cent. Plenty of populations in the dataset are stable or growing, but the steep declines drag the average down. The figure is real. It describes abundance loss rather than species loss.

The species-loss numbers are estimates rather than counts. About 1.7 million species have been formally described. The actual total is somewhere between five and twenty million; nobody knows. When IPBES says one million species are at risk, it is extrapolating from groups that have been thoroughly assessed (mostly birds and mammals, where over 99 per cent of species have been catalogued) onto groups that have barely been studied (most invertebrates, plants, and fungi). Sceptics call this fishy. Supporters reply that excluding the unstudied groups would understate the problem far more than including them overstates it.

If you cite "the sixth mass extinction" in your essay, treat the phrase as a contested label rather than an established fact. The box below explains why.

The Singapore data tell a sharper story because we have the records. A 2023 study in the Proceedings of the National Academy of Sciences compiled over 50,000 biodiversity records covering ten major taxonomic groups from 1819 to the present. The estimated overall extinction rate was 37 per cent. More than two-thirds of native orchid species are gone. Less than 0.2 per cent of original primary forest cover remains.

If the species-by-species count is contested, the trajectory is not. Climate change and biodiversity loss reinforce each other, and the speed is what threatens the ability of ecosystems to adapt. The figure that should anchor your essays is every six years, rather than 73 per cent or one million.

04Why it matters: the argument for conservation

Conservation has a public relations problem. Almost everyone supports it in the abstract. Almost nobody wants to pay for it in the specific.

Why save species? The question is not as silly as it first sounds. Conservation has costs, and those costs land somewhere. Land set aside for nature reserves cannot be developed. Money spent on environmental impact assessments is money not spent on hospitals. The Cross Island Line decision boiled down to roughly two billion Singapore dollars and six minutes. Multiplied across every infrastructure decision, every land-use plan, every agricultural policy, the cumulative cost runs into trillions globally. Why should that be paid?

The strongest answer is instrumental. Functioning ecosystems provide services that humans depend on but cannot easily replace. Pollinators are involved in crops responsible for roughly 35 per cent of global crop production volume; in their absence, an estimated five to eight per cent of that production would actually be lost outright. Forests regulate water cycles. Mangroves and reefs absorb storm surges and protect coastlines worth billions in property. Roughly half of pharmaceutical compounds in modern clinical use trace back to natural products. When IPBES warned that biodiversity loss undermines progress toward 80 per cent of the assessed Sustainable Development Goals, this is what it meant. Lose the supporting infrastructure of nature, and the things humans care about start to fail.

The argument from intrinsic value is harder to put on a spreadsheet but easier to defend morally. Other species exist independent of their use to us. Driving them to extinction for our convenience is a kind of theft from the future and from the rest of life on Earth. This is the philosophical position behind much religious environmental thought, behind the deep ecology movement, and behind a great deal of conservation in practice. It does not yield a benefit-cost number, but it shapes how people respond when those numbers come due.

There is a third argument, and it is the one most likely to win an essay: extinction is irreversible. Almost any other policy mistake can be corrected. An over-built road can be torn up. A bad tax can be repealed. A species that goes extinct cannot be brought back. This asymmetry should weight the calculation toward caution in a way that ordinary cost-benefit analysis does not capture.

The trade-offs are where the argument gets interesting. Consider the Cross Island Line.

Case study: the Cross Island Line, three options and the price of conservation

The Cross Island Line (CRL) is a roughly 50-kilometre east-west MRT line currently under construction. The full project is expected to cost around S$40.7 billion. Phase 1 began construction in January 2023. Phase 2, the section that runs through the centre of the island, is the one this case study is about.

The original engineering studies considered three alignments through the central catchment area. The third was eliminated quickly: a route running directly under MacRitchie Reservoir was ruled out because it would have crossed a geological fault line. That left two real options.

The Nature Society (Singapore) had been arguing for the skirting alignment since 2013. Even underground tunnelling, they argued, would risk disturbing the hydrology that keeps the rainforest alive. Vibrations could affect wildlife. Soil investigation work would damage forest habitat above the tunnel route. And, more philosophically, once the precedent was set that infrastructure could pass beneath a nature reserve, future projects would be harder to refuse.

Two environmental impact assessments later, in December 2019, then Transport Minister Khaw Boon Wan announced the decision. The government chose Option 1, the direct alignment. But not in its original form.

The tunnel would now be sunk to seventy metres, almost double the original depth and the deepest MRT tunnel ever built in Singapore. The previous record was 43 metres. The deeper alignment was supposed to mitigate environmental risk by placing the tunnel within solid granite rock, well below the root systems and hydrology of the surface forest. No surface structures. No ventilation shafts within the reserve. Khaw described the decision on Facebook as agonising. The tunnel would cross the reserve, but it would do so as gently as engineering could manage.

This is where the case becomes more interesting than the standard "development versus environment" framing makes it look.

The seventy-metre tunnel is a third option rather than a victory for either of the first two. The skirting route would have protected the reserve absolutely but cost two billion dollars and disrupted hundreds of homes. The original direct route at 40 metres would have been the cheapest path but with the highest environmental risk. The deep tunnel is an engineering compromise: most of the time savings of the direct route, most of the environmental protection of the skirting route, but at substantially higher engineering cost than a normal-depth tunnel would have required. Granite is hard to bore through. Tunnels at seventy metres need stronger ventilation systems, more complex station designs, and more sophisticated emergency egress.

Singapore could afford that compromise. Most countries cannot.

The deep-tunnel option is essentially a way of buying biodiversity protection with money. For a country with Singapore's fiscal headroom, that is available. For Malaysia, Indonesia, the Philippines, Vietnam, India, Bangladesh, or any of the dozens of fast-developing tropical countries that are now building out their own metro and rail systems, that option largely is not. A new metro line in Hanoi or Jakarta will not be built at seventy metres beneath a primary forest, because the engineering cost would simply consume the entire project budget. The cheaper direct route would be chosen, with whatever environmental damage it entails.

This is the deeper argument the Cross Island Line exposes. Conservation through engineering is a luxury good. Singapore can preserve its 0.2 per cent of remaining primary forest by building expensive tunnels around and beneath it. A country at a quarter of Singapore's GDP per capita facing the same trade-off will not. It will choose the cheap direct alignment, or it will not build the line at all, but either way the engineering compromise that lets Singapore have both transport and trees is not on its menu.

This matters for two reasons. First, most of the world's remaining biodiversity is in tropical countries that are not as wealthy as Singapore. The mathematical fact is that protection-through-money cannot scale globally because most of the relevant places do not have the money. Second, when Singapore presents its conservation record on the international stage, it tends to omit the size of the wallet doing the work. The achievement is real but not exportable. A talking point about Singapore as a model for "balancing development and biodiversity" carries an implicit footnote: balancing requires capital that most of the world does not possess.

You should be careful with this point in essays. It does not invalidate the Singapore approach. Buying environmental protection with money is genuinely better than not protecting it at all, and the deep tunnel may well work as intended.

But the strongest student answer recognises that what looks like a clever Singaporean trade-off is in fact a function of fiscal capacity, and that "do what Singapore did" is not advice you can give to most countries.

05Geoengineering: the climate switch

Geoengineering has two image problems. The first is that it sounds like science fiction. The second is that it is not.

Mount Pinatubo did, by accident, what some scientists now propose to do on purpose. Twenty million tonnes of sulphur dioxide reached the stratosphere in 1991. Sulphate aerosols formed and stayed aloft for roughly two years, reflecting sunlight and lowering global average temperatures by about half a degree until they settled out. This is the natural analogue for stratospheric aerosol injection (SAI), the most-discussed form of solar geoengineering. The technical idea is to deliberately inject sulphur dioxide or other reflective particles into the upper atmosphere using high-altitude aircraft.

The economics are striking. A 2018 Harvard study by Smith and Wagner estimated that a deployment programme using purpose-built tanker aircraft would cost about US$2 to $2.5 billion per year averaged over the first fifteen years of operation. A 2020 follow-up by Smith put the longer-term cost at roughly US$18 billion per year for each degree Celsius of warming avoided. By comparison, the global mitigation pathway compatible with limiting warming to 1.5°C costs trillions of dollars per year for the electricity sector alone.

This combination, large potential effect at low monetary cost, is what makes SAI both attractive and dangerous. Cheap planetary thermostats are a tempting toy. Cheap planetary thermostats with poorly understood side effects, no governance, and no off switch are something else entirely.

Geoengineering divides into two broad families with different profiles.

The current state of play is uneven. SRM has not been deployed at scale; the most ambitious proposed academic field test, Harvard's SCoPEx, was cancelled in March 2024 after sustained opposition from Indigenous and environmental groups in Sweden. Commercial DAC is operating, but at small scale and below expectations.

Consider Climeworks' Mammoth plant in Iceland, the largest operational direct air capture facility in the world. Its nameplate capacity is 36,000 tonnes of CO₂ per year. In 2024, its first year of operation, it captured roughly 105 tonnes total. That is not 105 tonnes a day, or a week. That is 105 tonnes for the entire year. As of mid-2025, only 12 of Mammoth's 72 collector containers were fully operational. The filter modules deployed at scale were not performing the way the prototypes had suggested they would. Climeworks laid off around a quarter of its workforce in 2025. The plant manager has acknowledged that, given the embedded emissions of construction and operation, it could take up to a decade for the facility to "pay off" its own carbon footprint.

For comparison, global energy and industry CO₂ emissions reached 37.4 billion tonnes in 2024. Mammoth's nameplate capacity removes the equivalent of roughly thirty seconds of global emissions. Its actual 2024 capture removed the equivalent of about a tenth of a second.

The cost picture is sobering. Climeworks captures CO₂ at around US$1,000 per tonne today. The company targets US$250 to $350 per tonne by 2030. Independent ETH Zurich research suggests US$230 to $540 per tonne by 2050 is more realistic. For comparison, Singapore's carbon tax is currently S$45 per tonne in 2026 and 2027, with a planned trajectory toward S$50 to $80 by 2030. The gap between what carbon costs to emit and what it costs to remove is the gap CDR has to close before it can scale meaningfully. At present rates, removing a tonne costs roughly twenty times what it costs to emit one. A market that prices removal at these levels does not exist.

06The risks and politics of engineering the planet

The technical case for geoengineering can be summarised in one sentence: we may need it. The political case against it can be summarised in another: but who decides?

Three risks dominate the debate.

Termination shock is the cleanest argument. SAI suppresses warming as long as you keep injecting aerosols. The aerosols settle out within a year or two of any injection ending. Start a programme and then stop it suddenly, and the suppressed warming returns rapidly. A planet that had been buffered against five years of accumulated emissions might experience those five years' worth of warming in five months. Ecosystems and societies that adapted to the cooled climate would face thermal whiplash with very limited time to respond. Models suggest termination shock could be more damaging than the warming the SAI was preventing in the first place.

This is a deeper problem than it sounds. Once a programme begins, it cannot easily stop. A future government that decides geoengineering is too dangerous, or too politically toxic, or simply too expensive, faces a choice: keep paying or accept catastrophic rapid warming. That is barely a choice. It is a hostage situation, voluntarily entered into.

Moral hazard is harder to quantify and more contested. The original argument is straightforward: if governments, voters, and corporations come to believe that geoengineering can save them from climate change, the political pressure to cut emissions weakens. Why decarbonise expensively when you can spray aerosols cheaply?

Some climate researchers argue the moral hazard concern is overstated. Public opinion surveys generally find that learning about geoengineering does not reduce support for emissions cuts. But this misses the relevant audience. The audience that matters for moral hazard is oil and gas company executives, fossil-exporting governments, and politicians who would prefer not to legislate hard choices. For those actors, the prospect of a future technological fix is genuinely useful as a delaying argument, whether or not the fix ever materialises.

The governance vacuum is the most acute. There is no functioning international framework for SRM. The 2010 moratorium under the Convention on Biological Diversity is widely cited but is non-binding and contested in scope. The technology is cheap enough that a wealthy individual, a corporation, or a small coalition of states could deploy it unilaterally. The effects would be global. The case below shows how thin the governance is at both ends, against rogue private deployment and against legitimate academic research.

Case study: Make Sunsets and SCoPEx, or how to get geoengineering wrong from both directions

In April 2022, an American entrepreneur named Luke Iseman launched two helium weather balloons over Baja California, Mexico. Each balloon contained a few grams of sulphur dioxide. The balloons were intended to burst in the stratosphere and release the gas, where it would form sulphate aerosols and reflect a microscopic amount of sunlight. He had no atmospheric science training. He had previously been a hardware director at the start-up incubator Y Combinator. Six months later, in October 2022, he formally incorporated the work as a company called Make Sunsets, with co-founder Andrew Song, and began selling "cooling credits" online for ten dollars per gram, claiming each gram offset the warming effect of one ton of CO₂ for one year.

Two things were true at once. First, the actual climate effect of two balloons of sulphur dioxide was negligible. The atmospheric scientist James Haywood called it a public relations stunt and said it would not "make a blind bit of difference." Second, the political effect was substantial. The Mexican government had not been consulted. It announced in January 2023 that it would prohibit solar geoengineering experiments within Mexican territory, citing the absence of international agreements and the failure of Make Sunsets to seek consent. The US Environmental Protection Agency demanded information from the company in April 2025.

Now consider the inverse. Harvard's Stratospheric Controlled Perturbation Experiment (SCoPEx) was a serious, peer-reviewed scientific project. It aimed to launch an instrumented balloon over northern Sweden to study how aerosols actually behave in the stratosphere. The researchers worked with an independent advisory committee of nine experts. They sought partnerships with Swedish scientific institutions. The proposed test would have released no more than two kilograms of calcium carbonate, which lead researcher David Keith pointed out was less than a commercial airliner releases in a single minute of normal flight.

SCoPEx was cancelled in March 2024.

The proximate cause was a 2021 letter from the Saami Council, the representative body of the Indigenous Saami people of Sweden, Norway, Finland, and Russia. Åsa Larsson Blind, then vice-president of the Council, told Grist that the project violated the Saami worldview, in which respect for nature is foundational. "We have a very clear position," she said, "that we do not approve of the development of solar geoengineering technology in Sápmi." The advisory committee recommended further public engagement before any test. The project's principal investigator, Frank Keutsch, eventually wound it down. Keith, the other lead researcher, left Harvard for the University of Chicago.

These two cases sit at opposite ends of the same problem.

Make Sunsets demonstrated that a private actor with no scientific credibility, no peer review, and no institutional oversight can conduct stratospheric experiments without consent. The legal frameworks are not strong enough to stop them. Mexico responded after the fact. The EPA responded years after the fact. The Atlantic Ocean does not need anyone's permission to receive falling sulphur particles, and neither did the upper atmosphere over Baja California.

SCoPEx demonstrated that credible academic researchers, working transparently, with peer review and Indigenous consultation, find it nearly impossible to conduct even small-scale tests. The legal frameworks are not even relevant. Public opposition is enough.

Neither outcome serves the world well. If we want to know whether geoengineering works, we need research. If we want research to be legitimate, it needs governance. We currently have the worst of both worlds. The cowboys can fly. The careful scientists cannot.

07Stewards or engineers? The justice question

The two halves of this chapter look like different topics, but they ask the same question. Biodiversity loss is what happens when humans alter ecosystems faster than ecosystems can adapt. Geoengineering proposes to alter the entire climate system on purpose. Both sit on a continuum of human intervention in the natural world. The debate about each is at root a debate about what kind of authority humans should claim over nature.

The conservation argument says: stewards. Pull back where we can. Restore what we have damaged. Accept slower development in exchange for ecological intactness. The Cross Island Line decision was a steward decision in form; the seventy-metre depth was an attempt to walk gently. Whether it worked depends on whether the deep tunnel actually limits hydrological disruption, which we will not know for decades.

The geoengineering argument, in its strongest form, says: we are already engineers, whether we admit it or not. Atmospheric CO₂ is now around 422 parts per million, more than 50 per cent above pre-industrial levels. Long-term global warming is currently estimated at around 1.3°C above the 1850 to 1900 baseline, with 2024 the first calendar year to exceed 1.5°C as a single-year average. The real choice lies between engineering the planet badly by accident and engineering it deliberately with some attempt at control.

Both arguments point at something true. The steward position is right that the costs of further intervention are uncertain and potentially catastrophic, and that the existing damage is itself evidence we are not very good at predicting consequences. The engineer position is right that pure restraint is not actually on the table; the question is what kind of intervention, on what authority, and at what scale.

But there is a deeper question, and this is the one your essays should reach for: who decides, and who pays?

The countries that have done the most to cause the climate crisis are not the ones that will suffer most from it. The G7 economies have emitted roughly half of cumulative CO₂ since 1900 with under ten per cent of the world's current population. The countries facing the most severe consequences from warming, including Bangladesh, the Maldives, the Pacific island states, the Sahel, and large parts of South and Southeast Asia, are largely those that contributed least.

The same asymmetry holds for solutions. The countries that could plausibly deploy SRM are very few: the United States, China, and possibly a wealthy coalition. The countries that could fund a deep MRT tunnel under a nature reserve are only a handful. The countries that could build a Climeworks plant the size of Mammoth are fewer still. Almost every climate technology that exists is held by a small set of wealthy actors. Almost every climate consequence that matters lands somewhere else.

Singapore sits in an unusual position within this asymmetry. It is wealthy enough to pursue expensive conservation and expensive carbon capture domestically. It cannot deploy SRM and would not benefit from doing so unilaterally even if it could. It is also acutely vulnerable as a low-lying island state; if a major emitter decides geoengineering is too risky and pulls back from a deployed programme, the resulting termination shock would arrive in Singapore's harbours before it reaches Washington's. The country is simultaneously a net beneficiary of global wealth and a net price-taker on global climate decisions made elsewhere.

This is the deepest point of the chapter, and it is worth ending on. Whatever we conclude about how humans should treat the natural world, the conclusion has to account for the fact that not all humans have the same power over it, and the ones with the most power are usually not the ones who will bear the consequences. That asymmetry, more than any technical argument about ecosystem services or termination shock, is what makes environmental policy a question of justice rather than only a question of management.

08Key statistics

Memorise these for essays.

09Debate map

Six positions, their evidence base, their Singapore application and the vulnerability of each.

Scroll the table sideways on a phone.

10Chapter summary

Biodiversity loss is real and accelerating, although the headline figures are widely misunderstood. The 73 per cent figure measures average population change, not species loss. The species-loss numbers are estimates rather than counts. The strongest figure to use in essays is a trajectory rather than a number: Southeast Asian mass coral bleaching events used to occur every twenty-five to thirty years and now occur every six.

The case for conservation rests on three legs: instrumental value (ecosystem services), intrinsic value (the moral standing of other species), and reversibility (extinction is permanent). The Cross Island Line case shows what these arguments look like when they meet a real-world infrastructure decision. The deep-tunnel compromise is genuinely impressive engineering, and it is also a luxury good. Most countries facing similar trade-offs cannot afford the same solution.

Geoengineering divides into Solar Radiation Management (fast, cheap, contentious) and Carbon Dioxide Removal (slow, expensive, more accepted). Singapore is committed to CDR but, like most countries, has no realistic agency over SRM. The Make Sunsets and SCoPEx cases show the governance vacuum at both ends: rogue private deployment is barely deterred, while serious academic research is stalled. Even the most-celebrated DAC plant is operating at a fraction of its design capacity.

The two debates share a question. Should humans be stewards of the natural world, pulling back where possible, or engineers of it, intervening more deliberately? The honest answer is that we are already both, and the question is on what authority and at what scale, rather than whether to intervene. The asymmetry of agency between countries, and the asymmetry between who causes harm and who suffers it, is the deepest justice question in environmental policy. It is also the one most likely to give your essays a Band 1 spine.

11References

The twenty-six sources the chapter is built on, as printed in the booklet.

1. WWF Living Planet Report 2024 (73% decline; regional and ecosystem breakdowns; species counts). World Wildlife Fund, October 2024.
2. IPBES Global Assessment Report 2019 (1 million species at risk; SDG impact). Intergovernmental Science-Policy Platform on Biodiversity and Ecosystem Services.
3. UN Sustainable Development Goals Extended Report 2025, Goal 15 (Red List Index 12% deterioration; tree assessments; freshwater data). United Nations Statistics Division.
4. Cowie, R.H., Bouchet, P., & Fontaine, B. (2022). The Sixth Mass Extinction: fact, fiction or speculation? Biological Reviews, 97(2).
5. Wiens, J.J. (2025). Questioning the sixth mass extinction. Trends in Ecology & Evolution.
6. Ceballos, G. et al. (2015). Accelerated modern human-induced species losses: entering the sixth mass extinction. Science Advances.
7. End-Permian extinction (about 96% marine species; about 252 million years ago). MIT News, 2018; Britannica entry on Permian extinction.
8. Kristensen, N.P. et al. (2023). Two centuries of biodiversity discovery and loss in Singapore (37% extinction rate; 18% Southeast Asia projection). PNAS.
9. Singapore 2024 mass coral bleaching (44% bleaching at peak; 31.7°C sea-surface temperature; recovery by November 2024). Wild Shores of Singapore reporting on St John's Island National Marine Laboratory data, April 2025; The Star (Asian) reporting, May 2024.
10. Mass bleaching frequency (every 25 to 30 years historically, every 6 years now). Ballard Brief on Southeast Asian coral degradation, May 2025.
11. Pollinator contribution to crop production volume (35%; 5 to 8% direct contribution). Klein et al. (2007), Proceedings of the Royal Society B, 274:303 to 313, summarised in Aizen et al. (2009), Annals of Botany.
12. Cross Island Line decision and three-options arc (December 2019; 4km direct versus 9km skirting; 70m depth; 6 minutes longer travel time; geological fault rejection of MacRitchie route; Khaw Boon Wan statement; Yew Lian Park 216 homes). Ministry of Transport official statement, 4 December 2019; Land Transport Guru, October 2024; Mothership.SG, December 2019; Land Transport Authority via The Straits Times, February 2016.
13. Cross Island Line total cost (S$40.7 billion) and Central Catchment Nature Reserve area (3,043 hectares). Wikipedia compilation citing LTA project data; NParks.
14. Singapore Green Plan 2030 (City in Nature; 7,800 hectares; over 40% green cover). Government of Singapore.
15. Singapore CCUS targets (2,500 kt CO₂ equivalent by 2030; Jurong Island). Singapore Economic Development Board on Singapore's UN biennial transparency report, December 2024.
16. Singapore carbon tax trajectory (S$25 per tonne 2024 to 2025, S$45 per tonne 2026 to 2027, S$50 to $80 by 2030). National Climate Change Secretariat (NCCS), 2025; Ministry of Sustainability and the Environment.
17. Climeworks Mammoth plant: nameplate 36,000 tonnes per year capacity (Climeworks press release, May 2024); actual 2024 capture of about 105 tonnes; 12 of 72 collector containers operational as of mid-2025; 25% workforce layoffs 2025. Latitude Media, June 2025; CleanTechnica, May 2025; Chemical & Engineering News, July 2025.
18. Global fossil CO₂ emissions 2024 (37.4 billion tonnes); atmospheric CO₂ about 422.5 ppm, 52% above pre-industrial. Global Carbon Project 2024 Carbon Budget; NOAA Climate.gov.
19. ETH Zurich DAC cost projections (US$230 to $540 per tonne by 2050). ScienceDaily reporting on ETH Zurich research, March 2024.
20. Mt Pinatubo 1991 climatic effect (about 0.5°C cooling for about 2 years; about 20 million tonnes SO₂). US Geological Survey via CNBC reporting, January 2023; MIT Technology Review, April 2024.
21. SAI cost estimates. Smith, W. & Wagner, G. (2018), Stratospheric Aerosol Injection Tactics and Costs in the First 15 Years of Deployment (about US$2.25 billion per year average), Belfer Center / Environmental Research Letters; Smith, W. (2020), The cost of stratospheric aerosol injection through 2100 (about US$18 billion per year per °C).
22. Make Sunsets timeline (April 2022 first launch by Iseman in Baja California; company incorporated October 2022 with Andrew Song; January 2023 Mexican ban; April 2025 EPA action; Haywood "PR stunt" quote). MIT Technology Review, December 2022 and January 2023; CNBC, January 2023; Climate Home News, January 2023; Washington Post, January 2023; US EPA, April 2025.
23. SCoPEx (cancelled March 2024; Saami Council 2021 letter; Åsa Larsson Blind quote; under 2 kg calcium carbonate; comparison to commercial airliner emissions per minute per Keith). Keutsch Group / Harvard SCoPEx site; MIT Technology Review, March 2024 and February 2021; Harvard Crimson, March 2024; Grist, April 2021.
24. Geoengineering moral hazard, termination shock, governance framework, and country deployment capabilities. Brookings Institution, 2022; Carnegie Endowment for International Peace, 2025; SRM360, 2025.
25. G7 cumulative CO₂ emissions and population share (about 54% of cumulative emissions since 1900; about 10% of world population). Gil-Alana et al. (2017), reported in PMC; IEA via Wang et al. (2024) Energy; Wikipedia G7 entry; Statista (UN DESA data, 2024).
26. Long-term warming about 1.3°C; 2024 single-year warming 1.55°C above 1850 to 1900. World Meteorological Organization, January 2025.

Source: "The Environment: Biodiversity and Geoengineering", GP Essentials II content crashcourse booklet, ETG GP. Reproduced in full, with the reference URLs omitted and the printed section numbering re-set for the web.