Renewable Energy and Energy Reductions or Solar Geoengineering for Climate Change Mitigation?

Abstract

This review explores the question: should the world rely wholly or partially on solar geoengineering (SG) to mitigate climate change (CC), or on renewable energy, together with deep energy reductions? Recent thinking is for SG to only supplement more conventional climate change mitigation methods. However, we first show that conventional mitigation methods are not working., given that global annual CO₂ emissions are still rising, so it is far more likely that SG will be called upon to counter most anthropogenic CC, as early research proposed. The paper next examines the various SG proposals that have been considered and their objectives. Future choices could be between an increasingly unpredictable climate, and SG, with its own risks and unknowns, or deep energy reductions and RE. The claim is that SG has far lower costs for a given climate forcing reduction compared with more conventional methods, and equally important, could be quickly implemented, producing temperature reductions in a year or so, compared with decades needed for more conventional mitigation approaches. SG implementation would affect not only the technical potential for key RE sources but also the actual uptake of RE and energy reductions. However, a fair comparison of RE and SG must recognise that the SG option also requires a solution to rising ocean acidification (OA). Because the material quantities needed annually to counter OA are orders of magnitude larger than for SG, its costs and energetic requirements will also be far higher, as will the time for implementation.

1. Introduction

Our planet already faces a variety of equally serious environmental and ecological crises. These include: the risks of and the risk of catastrophic climate change [1,2]; global biodiversity loss [3,4]; the rise of oceanic hypoxic regions [2], and loss of ocean phytoplankton and ocean acidification [4,5]. These risks to the planetary future are also evidenced by the actual or imminent transgression of planetary boundaries [6], or the conclusion of Ecological Footprint (EF) researchers that Earth has already overshot its ‘biocapacity’ by 70% [7]. Regarding climate change, global CO₂ levels are already too high; even at the 1.1–1.2 °C rise above pre-industrial temperatures that we currently experience; serious changes are already occurring as evidenced by the alarming rise in extreme weather in recent years [8,9] and mass loss from polar icecaps and Arctic sea ice [10].

Climate intervention in the form of solar geoengineering (SG) can be defined as the human manipulation of the physical environment on a very large or even global scale to achieve some defined aim. (SG is used here because the more general term ‘geoengineering’ is often used to include large-scale carbon dioxide removal (CDR) methods, and more specific terms such as ‘solar radiation management’ (SRM) or ‘stratosphere aerosol injection’ (SAI) exclude surface albedo modification).

Even though interest in this method goes back many decades [11], recent interest was re-kindled by a 2006 paper by the late atmospheric scientist Paul Crutzen [12], who, observing the slow progress made in more conventional climate mitigation methods, proposed injecting sulphate aerosols into the tropical lower stratosphere. In this much-discussed paper, he noted the immediate global cooling effect of the Mt Pinatubo volcanic eruption in the Philippines in June 1991, and suggested that this provided a natural experiment in global climate modification. It was soon realised that modifying shortwave radiation would also affect global precipitation, a result shown by models and confirmed by the experience with major volcanic eruptions [13]. Other serious side effects concern ocean acidification, discussed later, which will intensify if fossil fuel combustion continues [10,14]. SG must not only avert CC, but must not exacerbate these other serious environmental problems—as must also the more conventional CC mitigation methods designed to reduce emissions.

Like Crutzen, other early proposals for geoengineering usually took a global viewpoint, and focused mainly on how to maintain the present (or desired) average global surface temperature in the face of rising levels of greenhouse gases. SG was often assumed to completely offset anthropogenic climate change (CC). Most recent papers, however, assume that SG need only offset a fraction of global warming, with most global CC to be offset by conventional mitigation methods (e.g., [15,16]). As an example, Irvine and Keith [16] found that ‘halving warming with stratospheric aerosol geoengineering could potentially reduce key climate hazards substantially while avoiding some problems associated with fully offsetting warming.’ As the US National Academies of Sciences, Engineering, and Medicine [15] put it: ‘Solar geoengineering is not a substitute for mitigation’.

These conventional approaches, apart from RE and energy reductions, include nuclear energy and the various carbon dioxide removal (CDR) methods, both mechanical and biological. Nuclear energy is seen as at best maintaining its present minor energy share [17]. Its share of global electricity production peaked in 1996 at 17.5%, but by 2021 it had fallen to 9.8% [18]. Whether the various CDR methods can have an important role in CC mitigation is strongly contested (see [19] for a list of recent articles optimistic and pessimistic about its potential). However, this may not be important if CDR cannot greatly cut net CO₂ emissions over the crucial next decade or so. Although more optimistic forecasts have been published, a set of forecasts from the US Energy Information Agency (EIA) [20] project that global energy-based CO₂ emissions will be about 23% higher in 2050 compared with 2020 in the base case. Clearly, alternative fuels or CDR are not expected to have much of a role in the EIA scenarios.

Globally, modern forms of RE are still a minor energy source. The IEA calculated that global commercial energy use (i.e., excluding traditional firewood) was 565 EJ in 2020. Commercial RE only accounted for 68.6 EJ, or 12.1% of the total [21]. RE could take many decades to replace fossil fuels as the dominant energy source [22,23]. Far from stopping the rise in CO₂ energy-related emissions, these rose from 21.3 Gt (Gt = gigatonne = 10¹⁸ tonne) in 1990 to 33.9 Gt in 2021 [18]. Morris et al. [24], in their ‘Growing Pressures’ (for action on CC) still envisaged global CO₂ emissions continuing to rise until 2045. As Matthews and Wynes [25] put it in an eponymous article: ‘Current global efforts are insufficient to limit warming to 1.5 °C’. In the short term, the only approach capable of rapid implementation would be deep energy reductions, although major political, economic and social changes would be needed.

Nevertheless, an important argument can still be advanced in favour of major climate intervention. By loading the atmosphere with heat-trapping gases, we are already modifying the planetary albedo. As discussed by IPCC [10], a further 0.5 °C rise in planetary temperature will have much more serious effects than the preceding 0.5 °C rise. Climate intervention risk must be compared with those of possible catastrophic CC [26].

We are left with three choices for the near and medium-term future. First, like the EIA, we can simply accept that CO₂ levels will continue to rise, and accept the consequences. Second, we can implement SG as the dominant approach for stemming CC and accept its risks. Third, we can continue with RE growth, but recognize that most reductions will need to come from deep energy reductions, especially in the high-emission countries. Such reductions will not occur quickly, and will meet strong resistance from industry, especially the FF industry. Furthermore, given the continuing strong link between energy and GDP, declines in GDP can be expected, a result in line with degrowth theorists (e.g., [27]). Political and military power are broadly related to GDP and GDP per capita. So, this third option is unlikely to lower global emissions much over the next decade. Today, there are no easy solutions left. The chance was lost in 1990, the year of the first IPCC report, when steady conversion of FF to RE was a real option.

This Introduction has outlined the main thesis of this review. It argued that, contrary to the view that SG need only be an adjunct to conventional mitigation methods, it will come to be regarded as the dominant approach. Section 2 briefly outlines the rational for article selection in this review. Section 3 describes the various SG proposals, ranging from the most ambitious—a multitude of tiny mirrors in space to block some of the insolation reaching Earth—to the least, such as local schemes for painting roofs white. Section 4 considers in turn the effects on each of the major RE types that the various proposed SG interventions could have. It shows that SG will inevitably add great uncertainty to RE planning, and in some cases, will also lower the global potential for RE. It also examines how the promise of an SG fix could undercut momentum for CO₂ emission reductions. Section 5 compares SG with RE and energy reductions, stressing that an SG option may need to continue for centuries, and further, must include countering ocean acidification for a proper comparison. Finally, Section 6 summarises the arguments, and stresses that SG will be implemented in a world simultaneously facing a number of urgent environmental crises, as well as heightened international tensions and mistrust.

2. Methods

As already mentioned, the general term climate intervention is often used to refer to CDR methods, both biological and mechanical, as well as SG [13,28]. This review restricts itself to SG approaches which aim to either reduce the incoming shortwave energy received by Earth, or to raise the Earth’s albedo, thereby increasing the outgoing shortwave energy reflected unchanged to deep space.

The references used to detail the various approaches to solar engineering are given in Table 1. There is already a vast and rapidly growing literature on SG: putting the terms ‘solar radiation management’ OR ‘stratospheric aerosol injection’ OR ‘albedo modification’ into the Scopus database returned over 7700 reviewed documents, with over 700 for 2021 alone. Even these three terms combined would not exhaust SG-relevant papers. A careful selection evidently had to be made, and the papers chosen were mainly from the years 2018–2022.

Obviously, a review such as this would normally include papers on SG field experiments that test the validity of the models, but there are currently none being conducted. As Robock [13] has pointed out, ‘Any test of stratospheric SRM would have to be at full-scale implementation for decades to obtain statistically significant responses (because of the chaotic nature of the climate system, a large signal is needed to overcome the noise […])’. Similarly, the latest IPCC report [10] has concluded: ‘The effect of stratospheric aerosol injection on global temperature and precipitation is projected by models to be detectable after one to two decades, which is similar to the time scale for the emergence of the benefits of emissions reductions.’ Work on modelling is progressing, especially under the international Geoengineering Model Intercomparison Project (GeoMIP) and many papers have now been published (see, e.g., [29,30]).

3. Solar Geoengineering Proposals

The various proposals can be divided into those with an unambiguous global impact, such as reflective mirrors in space, and those with mainly a regional or even local impact, such as painting urban surfaces white. A distinction can also be made for Earth-based systems between those that modify atmospheric albedo, such as sulphate aerosol placement in the lower stratosphere, and those that merely modify surface albedo, whether of the ocean, vegetation, desert or urban surfaces. A further distinction is possible: one proposal aims to increase cirrus cloud formation and enhance outgoing long-wave radiation; all the other proposals discussed here work by increasing outgoing shortwave radiation.

Table 1 gives details for each option, as well as references. Most methods proposed to date aim at increasing the albedo of the atmosphere, Earth’s surface or the oceans. The most ambitious proposal to date would not attempt to change Earth’s albedo, but rather to decrease the insolation at the top of the atmosphere. One idea, among others, is to place a swarm of tiny reflective mirrors at about 1.5 million km from Earth [31]. This undertaking would be an immense engineering project, and would take decades to plan and implement, yet would have a lifetime of a few decades only. The cost of the project would run into trillions of dollars [31]. Further, much of this money would need to be spent upfront, with no guarantees of success. The mirrors themselves would need an area of some four million km² to counteract a doubling in CO₂ in the atmosphere, and need to deflect about 1.8% of Earth-bound insolation.

The most-favoured option today (SRM) is the injection of sulphate aerosols into the lower stratosphere. Injection could be done, for example, by naval guns, by very long hoses held aloft by balloons, or by aircraft releasing the aerosols as they travel. The injection could be done with existing equipment, and—it is claimed—would be relatively cheap by the standards of more conventional mitigation approaches. Sub-micron particle sizes are more effective than the larger sized aerosols usually emitted by volcanoes [32]. First, their lifetime in the atmosphere is much longer, requiring much lower annual replenishment for a given atmospheric loading. Second, larger particles, unlike smaller ones, not only scatter radiation but also absorb and re-emit radiation in the long-wave portion of the spectrum, thereby offsetting some of the cooling effect coagulation of the injected aerosols, but would both increase the annual injection amounts, and also increase the offsetting heating effects of larger particles. To counter this, Akasamit et al. [33] have proposed more injection points, and pulsed injections to help solve the coagulation problem.

Placing aerosols in the stratosphere, while more difficult than at lower heights, would reduce the annual tonnage of aerosols needed, because their residence time in the stratosphere is 1–2 years, compared with only about a week in the troposphere, because of rainout. Their lifetimes would be even longer if lofted into the mesosphere [13]. For global cooling, the preferred location for injection would be near the equator, as global coverage of the aerosols is maximised. Minimising the annual aerosol placement needed is important, not only for cost reasons, but because the aerosols will add to acidification of both land and oceans when they eventually descend. Tropospheric sulphate aerosols are already a serious pollution problem for humans.

Another proposal to increase radiation scatter in the atmosphere would brighten marine stratocumulus clouds (marine cloud brightening (MCB)) by seeding ‘low-altitude clouds over the ocean with salt particles to produce more and smaller cloud droplets, leading to brighter clouds’ [34]. The overall effect would be to decrease mean droplet size in the clouds, so raising their albedo, and possibly, their lifetime and areal extent. The authors caution that feedback processes may be at work which could negate albedo increases. Generating micro-bubbles in ocean surface waters to change the ocean surface albedo, has also been suggested as a climate intervention [35].

A different approach would also modify the properties of clouds, but rather than attempting to change albedo, the aim would be to modify cirrus clouds (cirrus cloud thinning (CCT)) to increase the longwave radiation leaving Earth. This climatic intervention would thus more closely mimic the effects of reducing atmospheric greenhouse gas concentrations. The ice crystal sizes in these high clouds would be modified by cloud seeding agents such as silver iodide, or the cheaper (and non-toxic) bismuth tri-iodide. The resulting larger crystal sizes would raise the fall velocities for ice particles, so decreasing global cover of cirrus clouds, which tend to trap outgoing longwave radiation. The residence time of the seeding agents is only 1–2 weeks, so continuous replenishment would be needed. However, as the US National Academies [15] concluded: ‘Existing climate model simulations of CCT have yielded contradictory results’.

Modifying the Earth’s land surface properties is yet another approach to albedo enhancement. Akbari and colleagues [36], and Pearce [37] have assessed the climate implications of painting white all roofs and pavements in the world’s urban areas. One finding is that such action would only delay global warming by 11 years. Further, rainfall patterns could change for the worse [37]. However, in high-insolation tropical countries, urban roofs away from the inner-city area are sometimes still thatched, and even many urban roads remain unsealed. In neither case would a stable surface be available for painting white. Another possible problem would be the effect of highly reflective roads on traffic accidents. Although global albedo reductions may be minor, because of the limited areas treated, the cooling reductions would mainly be in the urban areas themselves, and could benefit a significant share of the global population by reducing the urban heat island (UHI) effect.

Other surface albedo modification schemes would potentially have much larger effects, because the areas treated are so much larger. Seneviratne et al. [38] have examined the potential benefits of modifying the albedo of vegetation. They assume that both cropland and grassland albedo could be feasibly increased. They concluded: ‘The net effects may reach an albedo increase of about 0.05 up to 0.2 in the case of crops with high reflectivity residues, such as wheat, but would be smaller for other crops.’ The type of vegetation would need to be changed, or the plants could be selected or genetically modified for higher albedo. The potential for major changes to ecosystems is unknown, but could be important.

The world’s deserts presently have albedos ranging from 0.2 to 0.5, but this value could possibly be raised to 0.8 through albedo enhancement. One proposal is to cover vast areas of the desert with a highly reflective surface of white polythene with an aluminium foil backing. Desert areas with sparse vegetation, low population and a stable surface layer, covering a total area of about 11.6 million km², are thought suitable for modification [39]. Not only would fragile desert ecosystems be obliterated, but the feasibility of the project is doubtful. There are questions about the durability of the plastic under the desert sun, and how it would be anchored against winds. It may well be that in many areas the sheets would be soon covered by wind-blown dust and fine sand, which might solve the durability and anchoring problems, but would rapidly reduce the albedo of the covering. Crook [35] and Robock [13] are both sceptical that any surface modification approaches can make much difference.

Only some of these have been subject to even limited field testing, so their engineering feasibility is unknown. One geoengineering scheme already enjoys official support—using light-coloured roofs. The US government has a ‘Cool Roofs’ program, and a similar scheme is being introduced in Europe.

Table 1. Summary of solar geoengineering proposals.

SG Proposal	Comments	References
Place giant mirrors in space	Most expensive of all proposals, would take decades to implement; difficult to dismantle if serious side effects found.	[13,28,31,39,40]
Inject aerosols into lower stratosphere	Major volcanic eruptions show the feasibility of this method. Problems include possible particle coagulation, reductions in global precipitation, delays to ozone layer repair.	[12,13,28,30,39,40,41,42,43]
Brighten maritime clouds.	No chemicals used, and impacts would quickly disappear if in-tervention stopped.	[28,34,42,43,44]
Modify properties of cirrus clouds	Unlike other proposals, should not affect precipitation patterns. May not work as planned, and chemical additives needed.	[13,15,41,42,43]
Pump water onto sea ice, icecaps.	Vast and expensive engineering construction in a fragile envi-ronment.	[45,46,47]
Place floating reflec-tors in the ocean.	Could target specific areas. Potential adverse effects on ocean ecosystems if large-scale.	[13]
Modify desert albedo.	Large areas with low populations available. Durability of materials, and effectiveness of approach questionable; adverse effect on desert ecosystems; possible land competition with solar en-ergy schemes; high costs.	[35,39]
Paint urban roofs and pavements white.	Would produce local climate benefits, and politically and technically easy to implement. Global climate mitigation impacts would be minor; would compete with roof solar energy systems; high costs.	[28,37,41,42,43,48]

Table 1. Summary of solar geoengineering proposals.

SG Proposal	Comments	References
Place giant mirrors in space	Most expensive of all proposals, would take decades to implement; difficult to dismantle if serious side effects found.	[13,28,31,39,40]
Inject aerosols into lower stratosphere	Major volcanic eruptions show the feasibility of this method. Problems include possible particle coagulation, reductions in global precipitation, delays to ozone layer repair.	[12,13,28,30,39,40,41,42,43]
Brighten maritime clouds.	No chemicals used, and impacts would quickly disappear if in-tervention stopped.	[28,34,42,43,44]
Modify properties of cirrus clouds	Unlike other proposals, should not affect precipitation patterns. May not work as planned, and chemical additives needed.	[13,15,41,42,43]
Pump water onto sea ice, icecaps.	Vast and expensive engineering construction in a fragile envi-ronment.	[45,46,47]
Place floating reflec-tors in the ocean.	Could target specific areas. Potential adverse effects on ocean ecosystems if large-scale.	[13]
Modify desert albedo.	Large areas with low populations available. Durability of materials, and effectiveness of approach questionable; adverse effect on desert ecosystems; possible land competition with solar en-ergy schemes; high costs.	[35,39]
Paint urban roofs and pavements white.	Would produce local climate benefits, and politically and technically easy to implement. Global climate mitigation impacts would be minor; would compete with roof solar energy systems; high costs.	[28,37,41,42,43,48]

Although there is a clear distinction between very localised initiatives on the one hand, such as a municipal government implementing a white roofs program in its area, and on the other, ones with necessarily global and so international effects, such as ‘sunshades’ in space, there is a risk that any modest foray into SRM will place us on a slippery slope [13].

The realisation that both serious side-effects occurred with global interventions and that regional interventions were possible with primarily regional effects, has led to the proliferation of objectives for interventions. Interventions can now conceptually be varied in both time and location, to maximise a particular objective and minimise known side effects [44]. Researchers have also proposed using two or more interventions simultaneously. As shown above, even the regional effects of a single global intervention are not known with any precision. At least for one intervention, Zampieri and Goessling [49] offered the following blunt conclusion: ‘Our results cast doubt on the potential of sea ice targeted geoengineering to mitigate climate change.’

4. Renewable Energy and Energy Reductions

The future path RE will take is already beset with many uncertainties. The list of factors affecting its fortunes include uncertainties about fossil fuel depletion, the future of nuclear energy and CDR, and for RE itself, both the energy return on energy invested for the various forms, and their technical potential [8]. Furthermore, the growth of the various RE outputs will be affected by their environmental impacts—including any inputs of scarce minerals—as they are greatly scaled up to replace fossil fuels, and any resulting citizen opposition.

4.1. Effects of Solar Geoengineering on Renewable Energy Potential

Apart from solar satellite power systems, all forms of RE depend on regional values of parameters such as wind speeds, direct insolation, precipitation and runoff. The specific effects of geoengineering on RE will depend on what particular form geoengineering takes. This section considers how the various forms of geoengineering could impact on the potential for each RE source. Most RE sources will be affected to some extent, but the following sub-sections look in turn at solar, biomass, hydro and wind energy, the four most important RE types. Tidal energy and geothermal energy, because they are not derived from solar energy, will not be directly affected. Wave energy would be affected in a manner similar to wind, but is not yet generating any commercial energy. It needs to be kept in mind that on-going climate change will also affect RE potential [50].

4.1.1. Solar Energy

Solar energy from PV arrays is already widely used, and has great potential, especially in high insolation countries [51,52]. As already discussed, placing giant arrays of reflecting mirrors in space is the most expensive and most inflexible option and would have a very long lead time (several decades) from approval to effective operation. It thus negates the two main benefits claimed for geoengineering over other mitigation methods, low cost and rapid climate modification action. Since the mirrors would block some solar radiation before it entered the Earth’s atmosphere, there would be no scattering. This proposal is the least likely to be implemented, but its effects on solar energy potential would be expected to be minor, in line with the reduction in the solar radiation constant, which may need to be 2–4% [29,53], depending on what level of GHGs need to be neutralised.

Proposals that would reduce the planetary albedo by placing aerosols in the lower stratosphere would have the largest negative impact on solar energy potential, because the aerosols would scatter some of the incoming radiation, reducing the direct sunlight component. Murphy [54] has shown that for each 1% reduction in total sunlight reaching Earth, the output of solar thermal energy conversion plants, and any other solar systems which rely on focusing mirror arrays, would be reduced by 4–10%. Furthermore, in contrast to the modelled effects discussed in Section 2, there is empirical evidence for this claim: the output of the Californian SEGS plants were appreciably reduced in the aftermath of the Mt Pinatubo volcanic eruption in June 1991, and recovered after a year or so, in line with the recovery in direct solar radiation.

Production of solar electricity would be strongly affected in another way: some methods for albedo reduction conflict with area requirements for solar energy capture. Painting roofs white would obviously reduce the area available for roof-top solar hot-water or PV systems (Further, in cooler climates, reducing albedo would increase winter heating energy requirements. Having higher winter energy needs is even more of a problem if solar-energy is to be the major energy source in the future, since output of solar plants is then at a minimum). Covering deserts with high-albedo plastic would conflict with proposals such as Desertec, which envisage the use of deserts in North Africa and the Middle East for huge solar energy farms [50].

Sulphate aerosols would also adversely affect passive solar energy, although it is difficult to say to what extent. Passive solar energy for regulating building temperatures would be affected, as would lighting techniques that rely on focussing direct sunlight into buildings.

Solar geoengineering could affect precipitation levels in many regions. Nowhere would this be more disruptive than in the already water-stressed regions such as the Middle-East. Elshafei et al. [55] modelled the effect of a massive Pvcell array on Lake Nasser in Egypt. They found that if around 1000 km² of the lake was covered some 20% of the lake’s area, evaporation would be reduced, important since 20–30% of Egypt’s of Nile water is already lost through evaporation.

Abdelal [56] conducted an actual pilot study on experimental ponds using two fresh sources in Amman Jordan, and demonstrated a number of benefits. Power generation was similar to ground-motanted panels, and water quality was improved. Most importantly, a 60% reduction in evaporation was recorded. These studies also show that new technologies can help increase RE output to combat the twin problems of CC and water scarcity.

4.1.2. Biomass Energy

Major volcanic eruptions in the past have not only cooled the planet, they have also led to declines in globally averaged precipitation and global runoff. Large volcanic eruptions in the past have led to widespread drought and famine [57]. In a geo-engineered world, there would be no reduction in the need for non-energy uses of biomass—food, fibre, forestry products and forage for livestock. In a ‘business-as-usual’ world, demand for all these products could be expected to rise, because of both expected population increase [58] and rising per capita incomes. Even with a stable climate, as a residual use of biomass, bioenergy production would be threatened by rising demand for these competing biomass products [59]. Any fall in global Net Primary Production (NPP) that resulted from decreases in globally averaged precipitation, would further adversely affect bioenergy production.

As discussed by Li and Yang [60], the increase in diffuse radiation from aerosol light scattering would act to enhance plant growth, because more solar energy would be available to initiate photosynthesis on normally shaded leaves. There should also be further gains from steadily rising levels of atmospheric CO₂. On the other hand, because of the reduction in direct compared with indirect radiation, it is claimed that ‘solar geoengineering would reduce by at least 2–5% the growth rates of phytoplankton, trees, and crops’ [61]. In both cases, the gains will only occur if water and essential nutrients such as phosphorus and available nitrogen are not limiting factors. Changing plant types to increase vegetation albedo could also be expected to have some effect on NPP, but to an unknown extent.

4.1.3. Hydro Energy

Hydroelectric dams are expected to have lives of up to a century, which is fortunate given their huge monetary and energy costs for construction. As shown earlier, average global precipitation is anticipated to decrease in a geo-engineered world. Furthermore, research teams modelling the effects of a given geoengineering intervention using different global circulation models have found widely divergent changes from present precipitation averages for a given region. For instance, both Alamou et al. [61] and Bhowmick et al. [29] have argued that SRM could adversely affect the West African and Indian summer monsoons respectively. The differences found also depend to some extent on the location of aerosol placement or cloud brightening.

Even if the global models advance to such an extent that we can predict with confidence the precipitation and runoff changes that would occur for a given geoengineering intervention, the problem of future uncertainty still remains. The level of intervention needed would be expected to rise over time as the positive climate forcing from GHGs increases, and both the location, timing and level of intervention will change depending on what objectives the intervention is targeted to meet. The outcome is that if major geoengineering policies are implemented, we will have little idea of the hydrology patterns that will result in any river basin. Some basins will experience both rises in precipitation and temporally more uniform flows; others will experience the opposite. However, river basin planners will not know with any certainty into which category their basin falls. The result is that they will not be able to predict with any confidence either electricity output from existing hydro schemes, or the likely energy and monetary returns from proposed new hydro schemes. Without such knowledge the expansion of hydroelectricity, the present major source of non-intermittent electricity, would be in jeopardy.

4.1.4. Wind Energy

SG will probably only have minor effects on wind potential, but it could lead to its increased use, as wind energy has been promoted as one means of preserving the polar icecaps and stopping rising sea levels. One proposal is to pump huge amounts of sea water onto the West Antarctic Ice Sheet. An estimated 145 GW of wind power capacity would be needed [45], compared with 111 GW global new wind power capacity installed in 2020 [18]. It would be a huge engineering undertaking in a hostile and fragile environment. A similar plan has been proposed for the Arctic, where the aim would be to pump vast quantities of seawater onto surface ice during the winter. When frozen, it would thicken sea ice and extend its coverage. This solution would need ‘10 million windmills across the entire Arctic to refreeze it, at a cost of $500 billion’ [46]. Moon [47] has also strongly criticised geoengineering solutions as a technical fix to Arctic glacial melt.

4.2. Effects of Geoengineering on Renewable Energy Uptake and Energy Reductions

At first glance, geoengineering could be expected to have a general adverse impact on the prospects for future use of RE, regardless of type, because it promises to reduce the need to lower carbon emissions. In fact, all conventional methods of reducing fossil fuel use (through energy efficiency or conservation measures, nuclear energy, CDR, as well as RE), could be greatly affected, although some researchers have disputed this [48]. Thus, it could delay both the transition to large-scale use of RE and energy reductions, or if geoengineering is permanently maintained, to remove entirely the need for large-scale RE, at least as a climate mitigation measure. Has the promise of geoengineering already had an adverse effect on RE uptake? No precise answer can be given, but the slow growth in RE compared with what is needed for climate stability is consistent with this view. According to the IEA [21], in 2010 RE accounted for 13.6% of all global primary energy; by 2020, this figure had only grown to 15.7%.

Already, corporate interest in geoengineering is growing [62], as this is seen as the only option capable of maintaining a global growth economy, continued FF use, and addressing climate change. (Harris [63] titled his 2020 New Scientist article ‘Billionaire geoengineering’).

Despite all the cautionary words about SRM being merely a stop-gap measure to buy us more time to implement more conventional methods of mitigating climate change, its potentially low cost and rapid action compared with the alternatives, could make it very attractive to policy-makers [64]. A further attraction for geoengineering, whether global or regional, is that, unlike RE, up-front costs are a small share of the total—except for space-based mirrors.

Although, in general, fossil fuel interests and corporations in general could be expected to back solar intervention, matters are not always so straightforward. In the Arctic, there is interest in both exploration for oil and natural gas, which will prove far easier in ice-free waters. Further, an Arctic passage for shipping could potentially save both fuel and money for freight ships [65]. However, any benefits here would be vastly outweighed by the losses from stranded FF assets if FF production is seriously curtailed in future [66].

Conversely, in the unlikely event that RE or CDR grow very rapidly, it is at least possible that they would blunt the drive for solar geoengineering—as well as other options for climate mitigation.

5. Solar Geoengineering and RE Comparison

There are several vital differences between geoengineering and RE for climate mitigation. RE provides energy, geoengineering does not. So, we need to compare the costs of geoengineering plus fossil fuels with RE only. Furthermore, following the geoengineering path would eventually require alkalinity enhancement (e.g., calcium oxide or hydroxide) to the oceans to reduce progressive acidification, as much of the CO₂ released finishes up in the oceans. Most importantly, any use of RE that replaces some fossil fuels permanently prevents both the release of the resulting CO₂ and its climate forcing, but with the geoengineering plus fossil fuels option, CO₂ levels in both atmosphere and oceans continue to rise, and the climate forcing is suppressed only for as long as geoengineering measures are applied.

5.1. Solar Geoengineering Requires Ocean Alkalinity Enhancement

The world’s oceans, which account for half of our planet’s primary biomass productivity, are in trouble [4,67]. Rousseaux et al. [68] have reported that ocean phytoplankton mass is declining by around one percent each year. Oceans are also experiencing increasing acidity. Seventy years ago, ocean pH was 8.20, but has fallen to 8.04 in 2021. More than half of marine life have shells containing aragonite, ‘a mineral form of calcium carbonate that will dissolve by the time the acidity drops to pH7.95 […]’ [5]. A pH of 7.95 could happen as early as the second half of the 2040s. In short, OA is at least as serious a problem as global climate change. To remedy this, much research is now being done on ocean alkalinity enhancement (OAE). OAE would entail spreading finely ground alkaline minerals on much of the ocean surface, either by ship or aircraft. Mongin et al. [69], have also suggested it as a way of mitigating local acidification along the Australian Great Barrier Reef.

Caserini et al. [70] explored the potential for spreading slaked lime from the global cargo ship fleet, and concluded that several Gt of capacity was possible each year. Release from a point source such as a ship inevitably means high concentrations near the release point, which could have adverse environmental effects [59]. Releasing alkaline minerals from ships may be far cheaper than by aircraft, but with aircraft spreading there is less risk of high local concentrations. However, Gentile et al. [71] performed a detailed analysis of the costs for aircraft dispersal, varying factors such as dispersal time and height. Anywhere between around 3000 and 42,000 aircraft were assumed to be needed. Aircraft fuel energy use involved a CO₂ removal penalty of between 28–77%. The result was a cost per tonne of CO₂ removed of between USD 31 and USD 1920, even ignoring the alkaline materials cost. Fakhraee et al. [72] found that application of olivine was ineffective compared with CaO or MgO, and that 6–30 Gt of the latter feedstocks would be needed annually, depending on assumptions. Their overall conclusion was: ‘Taken together, our results highlight distinct challenges for ocean alkalinity enhancement as a CDR strategy and indicate that mineral-based ocean alkalinity enhancement should be pursued with caution.’

5.2. Solar Geoengineering May Be Needed for Centuries

For how long would we need to maintain SG to negate the forcing from a given amount of fossil fuel use? Swartz [73] has concluded that ‘cessation of anthropogenic emissions atmospheric would result in substantial recovery of CO₂ toward its preindustrial value in less than a century’. On the other hand, several studies (e.g., [39,53,74,75]) have argued that full recovery could take centuries, even millennia. The conclusion is that SG cannot be used merely as a stop-gap measure: the climate consequences of FF CO₂ emissions will continue for at least the rest of this century, and perhaps many more centuries, long after FF use has ceased, even if employed at a low level.

This differs from RE (or all forms of conventional CC mitigation), which provide a permanent reduction in CO₂ emissions equal to the total cumulative amount of CO₂ emissions either avoided or removed. Thus, these approaches incur only a single cost, but the full costs of geoengineering depend on the present value of all future annual costs, which is in turn heavily dependent on the discount rate used.

5.3. Discussion: Comparative Costs

Some cost estimates for SRM alone are now available. Robock [13] has given rough costs of producing a radiative forcing of about −2 W/m² or about that needed to ‘offset half the climate change that would result from doubling atmospheric CO₂’. The annual costs for the placement of 12 Mt of sulphur into the lower stratosphere he estimated as roughly USD 20–200 billion. An advantage of SG is that it has up-front costs that are a far smaller share of the total than for RE sources, where they can represent most of the costs. If greater levels of negative forcing are needed, costs will rise disproportionately, because particles would coagulate and grow larger, making them less effective and decreasing their atmospheric residence time [13].

Costs of placement rise with altitude, but in the case of aerosols, effectiveness also rises both because of longer residence time and because aerosols are above clouds; clouds reduce the light that can be potentially scattered. In contrast, surface and tropospheric interventions will primarily affect the modified region, although global effects will occur if the region of albedo enhancement is sufficiently large.

Estimates for costs per tonne of CO₂ avoided are available for various RE sources and CDR methods, although in some cases, like SG, CDR methods are untried at large scale. For various RE sources, the IPCC estimated costs are all under USD100. For CDR, estimated costs cover a huge range: USD 65–2120 [76,77]. Costs even lower than for RE are reported by the IPCC for energy reductions [65].

As already discussed, the OAE cost for air dispersion alone was estimated as USD 31–1920 per tonne of CO₂ neutralised. The costs for producing the CaO or MgO, grinding it and transporting it to the aircraft takeoff points would also be costly. Hence, the costs for OAE, even if ships were used, are likely to dwarf those for SG alone, given that the material requirements are some three orders of magnitude higher.

One important difference between all conventional methods of mitigation and SG is that the climate forcing from a reduction in CO₂ emissions is permanent. In contrast, with stratospheric SG, climate forcing reductions will need continued annual applications of aerosol for many decades. SG would thus transfer much of the costs of climate mitigation to future generations. The present generation would leave future generations with two choices, both unattractive: bear the annual costs of the intervention, perhaps for one or more centuries, or stop the intervention. However, as researchers have stressed, if we discontinue SG (perhaps because of the discovery of unexpected adverse side effects), the consequences of such an abrupt cessation could prove catastrophic to ecosystems [78]. The climate forcing from the steadily rising atmospheric CO₂ concentrations would be suddenly restored, resulting in an unprecedented surge in global temperatures, with possibly catastrophic consequences, as ecosystems would have little time to adjust.

More regional efforts, specifically albedo enhancements of plants, desert, or urban surfaces within a given nation’s territory, are least likely to generate international opposition. In fact, for urban surfaces, such albedo enhancement is difficult to separate from passive solar engineering. However, the costs could be much higher than for stratospheric aerosol placement. The Royal Society [41] estimated a cost of $0.3/m²/yr for painting urban surfaces white (with repainting every 10 years), which would need only proven technology. For 1.5 million km² of urban surface [39], the cost in the first year would be $450 billion, presumably in 2009 USD values, but this could be partly offset by lower air conditioning costs. However, such local interventions could at best slow, not reverse, climate change.

6. Conclusions

For dealing with climate change alone, one possible approach is to simply continue ‘business-as-usual’, which has largely been the response over the past three decades, as evidenced by ever-rising emissions (and temperatures) over this period. However, given the climate-related damages that are already occurring (see Section 1), this ‘do nothing’ (or, at best, do little’) approach is no longer an option. SG advocates are correct about this important point. Its impact and risks must be compared with those from unchecked further climate change.

Section 1 also showed that conventional approaches to CC mitigation are not working, as energy related emissions (and overall GHG emissions) continue to grow. Implementing widespread use of RE in the long-term is essential for replacing depleting FFs. Global cuts in energy will also help in ameliorating the other environmental challenges such as biodiversity loss. However, deep cuts in energy would need fundamental political changes; such changes look unlikely at present. It follows that SG will likely be required to neutralize the forcing from nearly all anthropogenic CC, not just a minor part, given that even the RE and energy reduction option will take a decade or more to make a real difference.

SG, whether the major or auxiliary means for CC mitigation, will likely cause both decreases in the potential of some RE sources, or will reduce the drive for RE and energy reductions (see Section 4), particularly given the importance of FFs in the global economy, and the crucial economic dependence of many nations on oil exports. Supporters will point to its lower upfront costs (if OAE costs are ignored), given the arguments advanced in this review, the lack of progress on conventional mitigation methods.

Politics is also important for the prospects of major SG implementation. We presently live in a time of heightened international tensions and mistrust. SG could be viewed as yet another means for implementing countries to gain competitive, even military, advantage. A much-discussed intervention, sulphate aerosol placement in the lower stratosphere, could require the use of either naval artillery or military planes, which would strengthen suspicion that the interventions had military purposes, and that the real purpose was to seek geopolitical advantage for the initiating country (or group of countries). Unfortunately, there are precedents for military attempts to modify weather conditions [79,80].

Biermann [81] has bluntly stated: ‘Solar geoengineering is neither necessary nor desirable. A global moratorium is needed’. An open letter calling for such a moratorium has already been signed by hundreds of scientists [64,82]. Even in an ideal, peaceful, world, global SG implementation would find difficulty in gaining international agreement. The problem is that global average temperatures and precipitation levels do not necessarily reflect regional levels. As we have seen, monsoons in India and west Africa could be adversely affected; there will be winners and losers among nations. One possible response is for the winners to compensate the losers. However, ongoing CC has disproportionately affected those living in the tropics, and little compensation has been offered. Local SG measures, such as ‘white roofs’ or crop albedo changes (Section 3) should not cause problems, but will only be of minor use for CC mitigation.

Implementation of a major geoengineering project, or several projects simultaneously in different regions, will add yet another layer of uncertainty, for two reasons. First, albedo engineering will change Earth’s energy flows to some extent. The changes from the average of the recent past in any region will depend on the timing, type, and location of the intervention, as well as its extent, as measured for example by the stratospheric loading of submicron sulphate aerosols. Even if all these were known with certainty, as already seen, the various models often produce conflicting results for their regional effects. Furthermore, models will prove less useful if the timing and extent of various geoengineering interventions are not known in advance, as could happen if ‘counter-geoengineering’ measures were undertaken by a country or countries dissatisfied with the result of the original SG measure—or group of measures.

The limitations of the present study arise from pervasive uncertainties. These include, first, the potential for conventional mitigation approaches to deliver deep emission cuts in the limited time available. Such reductions seem unlikely, but cannot be ruled out. If conventional mitigation approaches were successful in delivering strong GHG reductions, only a minor SG response would be needed. A second source of uncertainty concerns the regional responses to both large-scale and more modest GC interventions. The risks, both known and unknown, are likely to be greater for large-scale SG interventions. A third uncertainty is how CC will unfold if the ineffective policies of the past three decades are continued for some time.

Where, then, do we go from here? For effective mitigation, the world will need both RE and drastic reductions in global demand for energy [8,19], as recoverable FFs reserves will eventually deplete. Further research on both the technical aspects of energy reductions, including finding which energy sectors will yield the greatest cuts, and how the socio-economic obstacles may be overcome, are urgently needed. Such an approach would enable us to avoid the great risks of major geoengineering projects for the present generation, and the possible heavy costs for future ones. It is not only the safest, but in the long run, may even prove the cheapest option. It will not, however, be an easy option.