1. The impact on the ozone layer and the ozone treaties

The deployment of SRM would have impacts on the chemistry and dynamics of the stratosphere, with potentially adverse effects. For instance, stratospheric aerosol injection, particularly using sulphate aerosols, may lead to increased stratospheric ozone depletion.Footnote ⁴ Effects on ozone due to aerosol injections are not limited to sulphate aerosols. Other aerosols could also impact stratospheric ozone via “changes in heterogeneous chemistry and dynamics and transport.”Footnote ⁵ The specific extent of these effects remains uncertain, partly due to factors such as injection strategy. In any case, to assess these specific effects of SRM on ozone and other atmospheric processes, expertise in atmospheric dynamics is required, also to understand the implications of these for existing international obligations that may address these effects.

The treaty texts of the 1985 Vienna Convention for the Protection of the Ozone Layer and its 1987 Montreal Protocol on Substances that Deplete the Ozone Layer, both ratified by 198 parties, demonstrate considerable relevance to solar geoengineering.Footnote ⁶ The potential impact of use of SRM on ozone could bring such activities in conflict with the obligations in Article 2 of the Vienna Convention, which require states to take measures to protect the environment and human health against “adverse effects” resulting, or likely to result, from human activities that modify the ozone layer.Footnote ⁷ Further, Article 3 of the Vienna Convention obliges states to undertake and cooperate in research and scientific assessments related to activities that may affect the ozone layer. These obligations extend not only to understanding climate impacts,Footnote ⁸ but also to the exchange of socio-economic and commercial information,Footnote ⁹ as well as legal information concerning laws, bilateral agreements, administrative measures and patent availability.Footnote ¹⁰ As such, the Vienna Convention obliges states to assess solar geoengineering techniques and their geophysical and chemical impacts on the atmosphere and report on socio-economic and legal developments. This is underlined by the observed practices of the regime: the 2022 report by the UN-backed Scientific Assessment Panel to the Montreal Protocol included a chapter dedicated to stratospheric aerosol injection and its potential impacts on ozone.Footnote ¹¹ Later that year, the Parties to the Protocol acknowledged the need for further assessment while “[n]oting also the potential for negative effects that stratospheric aerosol injection may have on the ozone layer.”Footnote ¹² These obligations show that there is an institution in place that is legally mandated to assess and guide these central impacts of solar geoengineering.

2. Weather and climate impacts and the WMO

The injection of aerosols in the stratosphere would impact weather and climate patterns with potentially adverse effects. An obvious impact is the potential reduction in global temperature, which would be unevenly distributed both spatially and temporally.Footnote ¹³ Moreover, the implementation of SRM methods would likely result in alterations in the atmospheric hydrological cycle that may result in a decrease in global mean precipitationFootnote ¹⁴ and alter regional precipitation patterns.Footnote ¹⁵

The World Meteorological Organization (WMO), with 187 Member States and 6 Member Territories, is one of the rare examples of an international organisation in international environmental governance that not only assesses impacts on weather, climate, meteorology, hydrology and related environmental issues, but also promotes international coordination in these areas. Article 2 of the Convention of the World Meteorological Organization, adopted in 1947, states that one of the purposes is “to further the application of meteorology to aviation, shipping, water problems, agriculture and other human activities.”Footnote ¹⁶ As such, Article 2 suggests that the WMO is the primary international institution for assessing the impacts of weather dynamics and climate on human life. Since SRM intends to modify weather and climate, the WMO thus has the responsibility to evaluate the central impact of solar geoengineering. Importantly, however, and in terms of governance mandate, no regulatory capacity beyond describing and assessing the weather has been conferred on the WMO.

Beyond the legal finding that the WMO is mandated to assess the impact of solar radiation modification on the weather, it should be emphasised that in the WMO, the necessary expertise to assess and evaluate weather and climate dynamics is present.Footnote ¹⁷ Consequently, and not only from a legal but also from a functional perspective, any decision-making scheme on deployment would have to integrate this knowledge or duplicate it, risking incoherencies.

3. Biodiversity impacts and the mandate and practice of the CBD

Ecosystems and biodiversity may be adversely impacted by both the implementation and sudden termination of solar geoengineering. Deployment of SRM may influence global and regional weather and climate changes, hydrological cycles and soil and vegetation, potentially disrupting ecosystems and negatively impacting biodiversity.Footnote ¹⁸ Abrupt termination of any deployment scheme could lead to rapid and unprecedented increases in ocean and land temperatures, which may trigger unparalleled climate changes that may result in substantial losses of biodiversity.Footnote ¹⁹

The international institution mandated to assess the ecosystem impacts related to solar geoengineering is the 1992 Convention on Biological Diversity (CBD).Footnote ²⁰ The CBD enjoys near-universal participation, with 196 states having ratified it, except for the United States, which has only signed. Nonetheless, the United States did not submit a formal note of non-ratification and remains an active observer at the annual Conferences of the Parties (COP) to the CBD. The objective of the CBD is to conserve biological diversity and its sustainable use. Furthermore, Article 3 of the Convention states that member states have “the responsibility to ensure that activities within their jurisdiction or control do not cause damage to the environment of other States or of areas beyond the limits of national jurisdiction”.Footnote ²¹ Areas beyond national jurisdiction are global commons such as high seas, the atmosphere and space, thus relevant for solar geoengineering. Moreover, the preambular paragraphs underline that “where there is a threat of significant reduction or loss of biological diversity, lack of full scientific certainty should not be used as a reason for postponing measures to avoid or minimise such a threat” and that “it is vital to anticipate, prevent and attack the causes of significant reduction or loss of biological diversity at source”.Footnote ²² In Article 7, states further commit themselves to monitor processes and activities that are likely to have adverse effects on the “conservation and sustainable use of biological diversity.”Footnote ²³ Given the likely impacts of solar radiation modification on biodiversity, member states are obliged under the CBD to monitor but also avoid or minimise adverse impacts on biodiversity.

This interpretation is reinforced in the observed practices of the regime. Consistent with its mandate, the 2010 decision X/33 paragraph 8(w)—reaffirmed in 2012, 2016 and 2024—invites parties to ensure that “no climate-related geo-engineering activities that may affect biodiversity take place” under the provisos that the restriction applies until “an adequate scientific basis on which to justify such activities and appropriate consideration of the associated risks for the environment and biodiversity and associated social, economic and cultural impacts.”Footnote ²⁴ Although the decision is not directly legally binding on states, it is seen as a “normative precedent”Footnote ²⁵ and represents a consensus decision of all member states that interpret the binding treaty. Although the United States did not ratify the CBD, it should be noted that it did not object to references to CBD X/33 decisions at the United Nations Environment Assembly (UNEA) negotiations on SRM in 2019 and 2024. Even when assuming a lack of United States engagement in the CBD, any new multilateral treaty on SRM cannot be designed without taking into account existing CBD obligations that apply to its very large number of Parties. Moreover, the global expertise on biodiversity represented in the CBD cannot simply be overlooked or duplicated by a new SRM regime—this functional perspective is further emphasized in Section III.1 below. While the CBD has a broad mandate to assess and monitor activities that have an impact on biodiversity and to set guidelines to do so, it does not have effective control and regulatory mechanisms in place to govern SRM use.

4. Security and geopolitical impacts: UNSC and ENMOD

Solar geoengineering also presents various security and geopolitical risks. The unilateral deployment of SRM could have adverse impacts on the territories, economies and security of other states. Such unilateral deployment, involving a state or group of states deploying the technology without the consent or approval of the international community, therefore, represents a significant security concern.Footnote ²⁶ Unlike current emissions of greenhouse gases, the deployment of SAI would be intentional and could be perceived as deliberate and politically motivated acts of aggression.Footnote ²⁷ Some have also raised concerns about the potential militarisation of solar geoengineering due to the involvement of military and security institutions in research, the integration of militarisation assumptions in deployment scenarios, and the strategic potential of solar geoengineering in geopolitics, energy and climate change interlinkages.Footnote ²⁸

Various treaties and international institutions are potentially relevant for addressing security aspects of solar geoengineering. The United Nations Security Council (UNSC), operating within a framework of universal applicability to all UN member states, holds the “primary responsibility for the maintenance of international peace and security,” as specified in Article 24 of the UN Charter.Footnote ²⁹ It can determine “threats to the peace” and take measures under Articles 41 and 42 to maintain peace.Footnote ³⁰ Thus, the UNSC could impose sanctions in cases of unilateral deployment if viewed as a threat to international peace and security. It is important to acknowledge that practical implementation might be hindered by factors such as veto rights and lack of enforcement mechanisms.

Another potentially relevant regime is the 1977 Convention on the Prohibition of Military or Any Other Hostile Use of Environmental Modification Techniques (ENMOD), with seventy-eight parties, including China and the US.Footnote ³¹ The main substantial obligation enshrined in the text of the 1977 treaty is not to undertake military or other hostile use of environmental modification techniques having severe, widespread, and long-lasting effects.Footnote ³² Thus, “hostile” deployment of solar geoengineering, causing such outcomes, is prohibited. The line between peaceful and hostile use remains debated.Footnote ³³ Since ENMOD was agreed upon in an arms control context during the Cold War, some argue that only acts in “armed conflicts” are covered by ENMOD.Footnote ³⁴ However, since the methods and techniques of warfare are changing rapidly and do not only include traditional weapons, discussions are ongoing to broaden the concept of “armed conflict” to include new technologies.Footnote ³⁵ Moreover, in case the militarisation of solar geoengineering technologies proceeds, as some view as likely,Footnote ³⁶ the interpretation of “hostile use” could well encompass SRM deployment scenarios, even if no weapons are deployed in a conventional manner. While both the UNSC and ENMOD are not entirely adequate for preventing or managing potential conflicts arising from deployment, these existing frameworks still offer some relevance in addressing security aspects and may restrict some solar geoengineering activities.

5. Impacts on human rights and human rights law

Human rights law becomes relevant when specific solar geoengineering activities violate specific human rights. Any solar geoengineering deployment scheme would need to comply with existing human rights obligations and standards. The Universal Declaration of Human Rights (UDHR) provides a foundational normative framework for human rights, with many of its provisions reflected in binding international and regional human rights instruments, such as the International Covenant on Economic, Social and Cultural Rights (ICESCR), the International Covenant on Civil and Political Rights (ICCPR), the African Charter on Human and Peoples’ Rights (ACHPR), the European Convention on Human Rights (ECHR) and the American Convention on Human Rights (ACHR).

Some have already pointed to solar geoengineering activities as presenting a significant risk of violating specific human rights – some of which recognised in the above mentioned instruments – such as the right to a clean, healthy, and sustainable environment, the right to an adequate standard of living, the right to life, and the rights associated with consultation and free, prior, and informed consent.Footnote ³⁷ These and other potential violations to human rights from any future use of SRM have also been recognised by the UN Human Rights Council Advisory Committee, which noted that “the development of any such technologies and policies to support them would not be in accordance with the protective standards of the human rights regime.”Footnote ³⁸ Although not all relevant human rights instruments and protective standards noted above are binding in a strict legal sense, principles of international law deriving from human rights law and protective standards—as well as environmental law—remain applicable to any assessment of, or decision-making on, solar geoengineering technologies. In this context, the human rights regime provides important tools for assessing and monitoring this key impact area of any future SRM governance regime.Footnote ³⁹

We have outlined several existing regimes in this section, with relevance for governing potential impacts associated with solar geoengineering (see Figure 1).

Figure 1. A fragmented institutional architecture with extensive coverage of SRM impact areas.

Our analysis is not exhaustive, given the existence of over 1300 multilateral environmental agreements, many of which could also be affected by (unilateral or multilateral) solar geoengineering activities.Footnote ⁴⁰ Our discussion above does highlight, however, that most of the core impact areas linked to SRM are already covered by a wide array of international legal obligations, even if solar geoengineering is not explicitly mentioned or directly regulated as a technology or practice herein.

While certain governance “deficits” are present in this fragmented institutional landscape, particularly regarding key regime attributes such as lack of universal participation, effective decision-making or enforcement capacities, such deficits are common to many international environmental regimes,Footnote ⁴¹ and are also likely to hold for any future new SRM regime, should one be devised. Distinguishing such “deficits” from the notion of governance “gaps” is important in our view.

In contrast to deficits, by governance gaps, we understand here a gap in coverage of specific SRM impacts. A governance gap would imply an absence of regulations, norms or institutional coverage of key impact areas associated with SRM. As our analysis above has shown, multiple areas of SRM-related impacts are already covered within the international legal architecture. It is in this light that we argue that solar geoengineering, when approached from a multi-dimensional impact perspective, as we do in this article, is not an entirely ungoverned space. Hence, we challenge the notion of governance “gaps,” given that a diverse array of impacts of SRM are already covered by existing international regimes, which cannot be overridden or ignored. As such, this web of relevant international obligations, even if scattered, would need to be leveraged in any future governance of SRM, to which we turn next.

1. A fragmented architecture: implications for governing SRM non-use

In terms of a coordinated response for a restrictive, non-use regime on SRM, this fragmented institutional landscape does not pose an insurmountable hurdle. The legal obligation guiding this coordination is the principle of systemic integration in Article 31(3)(c) of the Vienna Convention on the Law of Treaties (VCLT). This article instructs treaty interpreters to take into account “any relevant rules of international law applicable in the relations between the parties.”Footnote ⁴² Accordingly, any new SRM regime would have to take into account the treaties, norms and obligations outlined in Section II.

This legal obligation is not only relevant for norm collisions and normative divergencies between regimes in a strictly legal sense, but also for operations and assessments within a regime.Footnote ⁴³ Since any formal or informal decision or operational application of a treaty inevitably relies on scientific and operational assessments conducted under its mandate, coordinating research across these is proposed as one tool to manage legal fragmentation.Footnote ⁴⁴ From this, it follows that (scientific) assessments conducted within one regime and from a specific scope and mandate must not draw any operative conclusion on SRM measures without taking into account the assessments of adjacent regimes. Hereby, “taking into account” would be more than mentioning a finding and then setting it aside. Setting aside or prioritising impact dimensions of neighbouring regimes without mutual referencing and substantial reasons would not be in line with the principle of systemic integration.

The more restrictive the mandate of any new SRM governance regime, the easier it will be to meet the requirements of Article 31.3.c VCLT on systemic integration. This is because only assessments and decisions that are directed to active measures and, therefore, actively risk frustrating the goals of another regime can trigger the legal obligations of systemic integration. In this light, the existing web of obligations to avoid or minimize diverse potential adverse impacts of SRM will result in restricting or ruling out some (or all) SRM deployment options. In theory, fragmentation may block the possibility of effective decision-making within a single new SRM regime. But instead of coming to quick conclusions at this point, it will be illuminating to have a closer look at the literature on fragmentation of international law. Early literature on fragmentation focused on pointing out the negative effects of fragmentation, such as producing conflicts and inefficiencies due to diverging international obligations.Footnote ⁴⁵ This corresponds to most of the findings in the SRM governance literature, which suggest governance gaps due to incoherence and incomprehensiveness of the existing governance landscape.Footnote ⁴⁶

However, the report of the Study Group of the International Law Commission on fragmentation in international law does not suggest the need to overcome fragmentation, but it rather analyses how to deal with it.Footnote ⁴⁷ In line with this, a new generation of contributions on fragmentation in international law has taken a more positive view, focusing on conceptualising how to deal constructively with fragmentation. This “new and optimistic view” on fragmentation, as Giorgetti and Pollack have noted, “has become the conventional wisdom in the international legal community today”.Footnote ⁴⁸ Peters points out that fragmentation can be viewed as an adequate response to the complexities of contemporary life and that fragmentation plays a role in preventing “abuse” by establishing a separation of powers, enabling checks and balances.Footnote ⁴⁹ Young and her contributors demonstrate how the diversification of legal regimes can lead to “productive friction.” This term refers to the dynamic interactions between different legal systems that can foster innovation and adaptation within international law. Such interactions may encourage regimes to learn from each other, leading to more effective and context-specific legal solutions.Footnote ⁵⁰

Following this, a fragmented governance architecture holds the potential to enhance the quality of restrictive SRM governance. As examined in Section II, SRM assessment and deployment presents a complex and multi-dimensional challenge, and would have direct implications for many fields of international cooperation and law. Consequently, the foremost governance challenge lies in encompassing all these diverse impact areas and specialized domains, delineated above. Given that SRM means an unprecedented human intervention into planetary systems and intends to manage the basis of life on Earth, the quality of decision-making should be central to global governance deliberations and arrangements. In this context, fragmentation does not necessarily weaken the unity of international legal systems if applied to governing non-use of SRM. Instead, it has the potential to increase comprehensiveness and coherence of a restrictive, non-use regime.Footnote ⁵¹

2. A fragmented architecture: implications for governing SRM use

In governing the use (as opposed to the non-use) of SRM, i.e. developing an SRM deployment regime, the presence of an already existing landscape of institutional obligations poses more significant difficulties. This is particularly true for proposals suggesting a new regime that should have SRM or geoengineering “in mind” in order to increase “coherence” and efficient decision-making.Footnote ⁵² A more in-depth legal examination of how decisions on SRM deployment would relate to existing obligations is still missing in the SRM governance literature.

Creating a new governance system or extending an existing one would not occur in “clinical isolation” from existing frameworks.Footnote ⁵³ According to Article 31.3.c VCLT, a new regime designed to govern SRM deployment would have to align with other international treaties and legal obligations as well. This is for two reasons. First, as we elaborate below, this is because the rules for norm collisions do not apply in this setting; and second, because fragmentation of international law is the regular finding, whereas superior norms and regimes that trump others are rare occurrences.

First, the rules for norm collisions are not applicable, since the regimes involved would not regulate the same subject matter. The lex specialis rule states that if a “same subject matter” is regulated by a general and a more specific rule as well, the latter shall prevail. A rule is only more specific in regulating “the same subject matter” if both rules are in some respect similar. Therefore, regimes that have different purpose mandates (e.g. protection of biodiversity, climate protection, human rights, management of SRM) are not, per se, regulating the same subject matter. Were that not the case, environmental protection regimes would always be legally subordinated to regimes regulating other “more specific” human activities. But none of these diverse purposes enjoys intrinsic priority over the other.Footnote ⁵⁴

Instead, the different purpose mandates of existing regimes have to be reconciled bottom up and not top down, and different regimes have to inform each other.Footnote ⁵⁵ In line with this, the International Court of Justice, when examining an extreme circumstance of self-defence in “Legality of the Threat or Use of Nuclear Weapons”Footnote ⁵⁶ did not set aside either the law on the use of force or the humanitarian law. Instead, both had to contribute relevant considerations to the advisory opinion. Our finding that the lex specialis rule does not apply between regimes having different purpose mandates does not mean that neither of the two conflicting rules from the different regimes can be applied or take precedence at all; in line with Article 31.3.c VCLT, such a rule should only be applied after thorough consideration of the specific setting. The point is that none of the existing relevant regimes is automatically set aside.

In order to avoid these complexities, states could, in theory, agree on a new regime that should take precedence over existing regimes. States can set aside existing obligations only if they decide to establish a self-contained and superior regime. Such an approach would position a new governance system on top of the hierarchy within the relevant international law. However, this is highly unrealistic and would represent a constitutional leap. This is evidenced by the fact that international law rarely allows for a “trumping” impact of hierarchically superior norms over the norms of other regimes and treaties. Some of the rare examples of such prevailing and quasi-constitutional rules are Article 103 of the UN Charter or the ius cogens prohibition of torture.

Turning to international environmental law, even the much-discussed World Environmental Organisation failed to come into existence.Footnote ⁵⁷ Instead, it was agreed that the United Nations Environment Programme (UNEP) shall galvanise actions by other institutions and agencies; it was not given the authority to steer or authoritatively coordinate such actions.Footnote ⁵⁸ It seems unlikely that states will agree on an overarching SRM institution to have the capacity to make decisions on its own or steer other relevant institutions.

Therefore, even if there were a new regime having the mandate to decide on SRM deployment, it would not be in the position to set aside existing regimes and obligations, but Article 31.3.c VCLT and the principle of systemic integration must be applied in any SRM governance regime or architecture.