Introduction
The likely insufficiency of conventional climate mitigation policies to prevent dangerous climate change has prompted increasing interest in technological interventions to cool the planet. Most of the scholarly and policy literature on climate engineering has focused on interventions aimed at influencing the global climate—for example, by injecting reflective particles (aerosols) into the stratosphere, which would reduce the amount of sunlight reaching the surface of the Earth. Our focus here is on a less-studied area: potential Arctic-specific interventions to thicken sea ice, reduce ice melt, or stabilize glaciers, which we argue raise fewer legal and governance issues than global interventions.
Conventional policies to limit (and ultimately halt) global warming have focused on greenhouse gas emissions reductions. But emissions reductions, although essential, cannot by themselves save the Arctic. Even if greenhouse gas emissions were completely eliminated tomorrow, sea ice would simply stabilize at its current (low) levels, not recover anytime soon.Footnote 1 And, of course, eliminating emissions tomorrow is a complete impossibility. Global emissions are still increasing and, even on the most optimistic assumptions, will continue for decades. According to current estimates, the world, at best, will warm another half degree and, more likely, will warm much more, causing additional harms to the Arctic.Footnote 2
Opinions vary widely about whether Arctic-specific interventions are technically feasible and, if so, would be scalable, cost-effective, and safe.Footnote 3 We express no opinion on these issues, which are outside our areas of expertise. Instead, our focus is on the legal issues raised by Arctic-specific interventions, including the relevant environmental norms.
Overview of Arctic InterventionsFootnote 5
Arctic interventions fall into four general categories: (1) interventions to thicken sea ice; (2) interventions to slow snow and ice melt on land; (3) interventions to stabilize glaciers, to slow their slippage into the ocean; and (4) so-called “Pleistocene parks.”
Arctic Climate Interventions.Footnote 4
Thickening sea ice. Sea ice reflects five to ten times more sunlight than water, so when it melts and is replaced by open water, more solar radiation stays in the Earth system and warms the planet. The goal of sea ice thickening would be to add ice during the winter so the ice does not melt through completely in the summer. One proposal is to pump water from beneath the ice and spray it onto the ice surface, where it would be exposed to the air and freeze more quickly.Footnote 6 Even if this did not delay global warming,Footnote 7 it could at least help ice-dependent species and communities.Footnote 8
Slowing ice and snow melt. Reducing ice and snow melt has the same goal as sea ice thickening: namely, to prevent reflective snow and ice from being replaced by water and dry land, in order to preserve the Arctic’s albedo. In theory, the melting of ice and snow could be inhibited by increasing their reflectivity—for example, by placing reflective materials on the surface of the ice or snow in order to reduce the amount of solar radiation it absorbs.Footnote 9 One project, which proposed using glass microspheres as the reflective substance, shut down in 2025, but researchers are continuing to experiment with other reflective materials.Footnote 10
Stabilizing glaciers. In contrast to snow and ice management, which aims to preserve the Earth’s albedo, glacial engineering has as its primary goal to limit sea-level rise.Footnote 11 Between 1992 and 2020, Greenland is estimated to have lost more than 5 billion tons of ice, adding a half-inch to global sea level.Footnote 12 If Greenland’s ice sheet was to melt completely, this would raise sea levels globally by more than 7 meters. Proposals to slow the loss of Greenland’s ice sheet include:
• Erecting underwater curtains to block warmer sea water from penetrating the base of the ice where it reaches the sea.
• Buttressing ice to prevent it from slipping.
• Pumping out water beneath glaciers that lubricates their slippage downhill.Footnote 13
Pleistocene park. Finally, some scientists have suggested that reintroducing large herbivores into the Siberian tundra could slow the thawing of permafrost.Footnote 14 The theory is that large herbivores such as reindeer, horses, and bison would help pack down snow, thereby reducing the snow’s insulating effect and accelerating the freezing of the permafrost below. One modeling study found that this approach could reduce the thawing of permafrost by almost 40 percent.Footnote 15
Distinctive Features of Arctic InterventionsFootnote 16
Proposed global climate interventions are highly controversial and raise difficult governance issues, given their global effects and potential risks.Footnote 17 By contrast, Arctic-specific interventions raise fewer legal and governance issues. They differ from global climate interventions in several important respects:
First, Arctic interventions arguably constitute adaptation measures. For example, building a sea curtain to slow the slippage of glaciers into the ocean is not dissimilar to building a sea wall to prevent coastal flooding. Like a sea wall, the function of a sea curtain would be to prevent the movement of seawater that causes damage to the environment—in the case of a sea wall, damage to coastal areas; in the case of a sea curtain, damage to glaciers. Similarly, refreezing seawater would be analogous to reforestation or the development of drought-resistant crops: the goal would be to preserve or restore the status quo ante, rather than to affect the global climate system.
Second, Arctic-specific interventions would have primarily local or regional effects. For example, spraying reflective particles on an ice-covered lake in Canada to inhibit melting would take place within Canadian territory and would not have significant transboundary effects.
Third, Arctic-specific interventions would clearly not be a substitute for emission reductions. They would thus not be subject to one of the objections to global climate interventions raised by critics: namely, that the prospect of being able to limit global warming through technological interventions might undermine efforts to reduce emissions.Footnote 18
Finally, some Arctic interventions, such as Pleistocene parks, can be framed as “natural” rather than technological solutions. In essence, they aim to re-wild regions that were transformed by the hunting of large herbivores by humans.
Legal and Governance Issues
The legal and governance issues raised by Arctic-specific interventions differ for land versus ocean-based interventions. Land-based interventions would be governed primarily by international environmental law, while sea-based interventions would be governed primarily by the law of the sea.
Land-Based Interventions
As discussed above, land-based interventions include spraying reflective particles onto ice and snow and reintroducing large herbivores into permafrost regions. Because they occur on land, these interventions would be within the jurisdiction of the territorial state. Like other such activities, they would be subject to the general principles of international environmental law, which impose rather modest obligations to limit adverse transboundary effects. As spelled out by the International Court of Justice, the primary duty of states is to use due diligence to prevent significant transboundary harms.Footnote 19 As part of that duty, states must conduct environmental impact assessments of activities that may have significant adverse transboundary effects and must notify and consult with potentially affected states.Footnote 20 If an intervention might affect an Indigenous People or a local community, this would also implicate Indigenous rights and human rights law. In addition, parties to the Polar Bears Agreement would be obliged to “take appropriate action to protect the ecosystems of which polar bears are a part” from any adverse impacts of land-based interventions.Footnote 21
It is not clear why Pleistocene parks or reflective particles would have significant adverse transboundary effects. But if scientists thought that they might, then the territorial state would have a duty to assess and possibly notify, consult, and take reasonable measures to mitigate. In addition, if an intervention might impact Indigenous Peoples, a state would need to consult and cooperate with the affected group.Footnote 22
Beyond their local and regional effects, Arctic-specific interventions would have global impacts, since what happens in the Arctic does not stay in the Arctic. Indeed, that would be one of their main motivations: glacial engineering aims to limit sea-level rise and slowing snow and ice melt aims to maintain the Arctic’s albedo. But compared to global interventions such as stratospheric aerosol injection, which affect the climate system pervasively, Arctic interventions are much more modest. Whether they would have any significant effect on the global climate is unclear. To the extent they did, they would be no different from many other measures states take within their territory to address climate change, such as emission reductions, reforestation, and painting roofs white, none of which are seen as raising difficult governance questions.
Ocean-Based Interventions
Ocean-based interventions in the marine Arctic would be subject to the environmental norms set forth in the UN Convention on the Law of the Sea (UNCLOS)Footnote 23 as well as in customary international law. Under Article 192 of UNCLOS, states are obligated to “protect and preserve the marine environment.” This obligation is now generally accepted as reflecting a rule of customary international law.Footnote 24 As the International Tribunal for the Law of the Sea (ITLOS) stated, the obligation is “one of due diligence.”Footnote 25 Thus, states would be required to exercise due diligence to protect and preserve the marine environment, including marine species and ecosystems, from environmental harms stemming from ocean-based interventions in the marine Arctic.
Like land-based interventions, states conducting ocean-based interventions would be subject to the no harm principle. For ocean areas, this principle is set out in Article 194(2) of UNCLOS, which has been interpreted as a due diligence obligation.Footnote 26 This obligation is linked to an obligation to conduct an environmental impact assessment under Article 206 of UNCLOS because Article 206 constitutes a “particular application” of the more general obligation enunciated in Article 194(2).Footnote 27 If ocean-based interventions might affect the rights of Indigenous Peoples, it would be crucial to ensure the participation of Indigenous Peoples in the environmental impact assessment process.Footnote 28
Although UNCLOS contains no express obligation to apply the precautionary approach, ITLOS, in its advisory opinion on climate change, interpreted the due diligence obligation under Article 194 as requiring application of the approach.Footnote 29 By identifying potential risks that may trigger the application of the precautionary approach, environmental impact assessments of ocean-based interventions would facilitate its application.Footnote 30
It is beyond serious argument that effective implementation of an environmental impact assessment is of critical importance in the protection of the Arctic environment.Footnote 31 However, Article 206 of UNCLOS does not specify the content of an environmental impact assessment. Accordingly, as the ICJ stated in Pulp Mills on the River Uruguay:
[I]t is for each State to determine in its domestic legislation or in the authorization process for the project, the specific content of the environmental impact assessment required in each case, having regard to the nature and magnitude of the proposed development and its likely adverse impact on the environment as well as to the need to exercise due diligence in conducting such an assessment.Footnote 32
As concerns marine spaces beyond national jurisdiction in the marine Arctic, however, an environmental impact assessment relating to ocean-based intervention would be subject to the rules and procedures set out in the Biological Diversity Beyond National Jurisdiction (BBNJ) Agreement.Footnote 33 Under the BBNJ Agreement, the environmental impact assessment of a planned activity that is to be conducted in marine areas beyond national jurisdiction is to be supervised by the Agreement’s Scientific and Technical Body.Footnote 34 Thus ocean-based interventions on high seas ice sheets would be subject to a more institutionalized assessment procedures.
It must also be noted that follow-up and monitoring are crucial to reviewing the impacts of ocean-based interventions on the Arctic environment. An obligation of monitoring is provided by Article 204 of UNCLOS and Article 35 of the BBNJ Agreement.
Refreezing marine areas could potentially conflict with navigational rights. The legal status of navigational routes in the marine Arctic is not free from controversy, but it is relevant to note that the coastal state is obliged not to hamper the innocent passage and transit passage pursuant to Articles 24(1) and 44 of UNCLOS and must give “due regard” to freedom of navigation in its exclusive economic zone and on the high seas pursuant to Articles 56(2) and 87(2).Footnote 35 The coastal state is also required to give appropriate publicity to any danger to navigation under Articles 24(2) and 44 of the Convention.
Conclusion
Arctic-specific interventions are still quite speculative. Major questions exist regarding their feasibility, effectiveness, scalability, and safety. But from a law and governance standpoint, they are more like conventional mitigation and adaptation than like potential global interventions. Given the magnitude of the threats posed by global warming for the Arctic, we cannot afford to exclude, ex ante, any options.
Open access
