3.1 Jurisdiction Ratione Personae and the Existence of a Dispute

The Court’s ratione personae jurisdiction refers to its competence to exercise jurisdiction over the States (the persons) named as applicant and respondent to the case. Depending on the constellation of States, norms, and routes to the Court which are involved in any given case, there are numerous ways in which a potential application could either succeed in meeting or fail to meet the threshold. These manifold possibilities cannot be assessed in the abstract; rather, they are dependent on particular combinations of States and legal instruments. As noted above, for this study our working presumption is of a case between two States both of which having entered Optional Clause declarations.⁴⁹ As such, jurisdiction ratione personae over the States concerned is, in principle and prima facie, established by their respective Optional Clause declarations.

There are, however, two important provisos. First, many (though notably not all⁵⁰) States’ Optional Clause declarations carve out exceptions which condition the extent of their acceptance. For example, Australia’s declaration explicitly excludes disputes concerning the delimitation of maritime zones, as well as those in which the parties have agreed to pursue a different method of dispute settlement.⁵¹ Australia’s declaration is reasonably representative: many declarations contain specific subject-matter limitations, as well as requiring reciprocity and excluding dispute settlement where another method has been agreed. Few declarations would specifically exclude a climate case, save those of Romania and Slovakia,⁵² which explicitly exclude disputes “with regard to the protection of the environment”. Setting aside these exceptions, it seems likely that a climate-related case could meet the criteria established in the respective applicant’s and respondent’s declarations, provided that reciprocity exists and no alternative method of dispute settlement has been agreed.⁵³

Secondly, the text of Article 36(2) of the Statute evinces certain minimal requirements which must be met in order to activate the Court’s jurisdiction over any two specific States, ratione personae, even where the Optional Clause jurisdiction is concerned. Article 36(2) requires that a “legal dispute” exist between the parties in order for a State to seise the Court.⁵⁴ The classic definition of the existence of a dispute remains that in Mavrommatis Palestine Concessions: “a dispute is a disagreement on a point of law or fact, a conflict of legal views or interests between parties.”⁵⁵ In other words, for a dispute to arise it would suffice, in principle, that one party raise a point of fact or law which is contested by the other State. In The Gambia v. Myanmar, the Court reaffirmed this principle and noted that as a matter of objective determination, and of “substance, not a question of form or procedure”, a number of points could be considered by the Court, including exchanges directly between the parties, as well as exchanges in multilateral settings.⁵⁶ Taken as a whole, the Court determined them to “suggest the existence of a divergence of views” in respect of the treatment of the Rohingya population, and pointed notably to the lack of response by Myanmar on various occasions as itself indicating the existence of a dispute.⁵⁷ The Court deployed nearly identical reasoning in its March 2022 Order in Allegations of Genocide under the Convention.⁵⁸

Though it is unusual for applicants to fail to clear the threshold of the “existence of a dispute”, (in)famously, it was for lack of a legal dispute that the Court declared inadmissible the Marshall Islands’ claim against the United Kingdom brought under the Optional Clause, one of the Nuclear Disarmament cases which it submitted to the Court in 2014.⁵⁹ Although the decision by the Court was taken at the preliminary objections phase, the logic may be relevant, mutatis mutandis, to the issue of prima facie jurisdiction, as the objection that a “legal dispute” between the parties (within the meaning of Mavrommatis) is lacking may be identified as a viable strategy to counter any request for provisional measures. In that dispute, the United Kingdom had objected that, under customary international law, a State intending to invoke the responsibility of another State must give prior notice of its claim, “such notice being a condition of the existence of a dispute.”⁶⁰ The Marshall Islands could only substantiate such notice indirectly, invoking statements in multilateral fora such as the UN General Assembly. Deciding by casting vote, the Court disagreed that the Marshall Islands had offered sufficient particulars relating specifically to the United Kingdom’s conduct,⁶¹ and declared that the existence of a dispute rested in part on the UK being aware, or being unable to be unaware, that a dispute on an issue of law or fact existed.⁶²

Whatever one might think of the Court’s further tightening of the criteria for the existence of a dispute, any potential climate litigant will be wise to anticipate these when devising its litigation strategy. With foresight, the Nuclear Disarmament standard should be easily satisfied by any sufficiently unambiguous communication addressed squarely by the prospective applicant to the potential respondent State. A diplomatic communiqué or note verbale, even if it does not demand or receive a response, ought to suffice to raise awareness in the respondent State that its conduct is regarded by the prospective applicant State as being in violation of its obligations.⁶³

3.2 Jurisdiction Ratione Materiae

As regards customary international law, the Court’s jurisdiction ratione materiae is broad. This would be different were jurisdiction to be founded under a compromissory clause, which by definition would confine jurisdiction to the scope of the said treaty, or a special agreement, which delineates the four corners of the consent of the parties to a specific dispute. However, the material scope of a dispute on customary international law is limited only by the specific States’ Optional Clause declarations. Provided the requisite declarations under Article 36(2) of the Statute have been entered, the broad enumeration therein covers a vast range of potential disputes pertaining to:

1. The interpretation of a treaty;

2. Any question of international law;

3. The existence of any fact which, if established, would constitute a breach of an international obligation;

4. The nature or extent of the reparation to be made for the breach of an international obligation.

Provided, therefore, that the dispute concerns a question of international law – for our purposes, specifically falling within the meaning of Article 36(2)(c) of the Statute – it may thus concern interpretative questions, factual determinations, or the assessment of reparation due. As we will further explain in Section 4.2, infra, we would encourage an applicant to focus on the legal consequences flowing from a violation of the customary international law obligation not to cause transboundary harm to another State. From this, logically, would flow interpretative questions on the scope of that obligation, assertions on the existence of facts, and the potential reparation owed by a respondent.

3.3 Jurisdiction Ratione Temporis

With regard to jurisdiction ratione temporis, the Court must be satisfied that both parties have accepted the jurisdiction of the Court during the relevant time period, and crucially, by “the date on which the application is submitted to the Court”.⁶⁴ With all due caveats given the hypothesis imagined here, consent will be established through each Party’s Optional Clause declaration on the basis of reciprocity, i.e., only after both States’ declarations have entered into force vis-à-vis one another, with the later declaration crystallising the scope of the mutual consent and, with it, the basis of jurisdiction ratione temporis.⁶⁵

Given the complexities of climate change prevention, two particular paths suggest themselves. One path rests on breaches that are presently occurring or have occurred subsequent to the mutual acceptance of jurisdiction, such as – for example – a failure to transmit a required report, or a decision taken to build new fossil fuel infrastructure that is incompatible with the State’s commitments. The second would be to invoke a continuing violation, in the sense of Article 14(2) of the ARSIWA, such as a failure to enact necessary legislation, to deliver agreed finance for mitigation and adaptation, or, reasoning by analogy from the Court’s Chagos Archipelago Advisory Opinion, a continuing failure to correct an earlier wrong.⁶⁶ Both remain plausible routes, despite the complex series of facts that are relied upon to ground the scientific consensus on climate change.

4.1 Varying Thresholds of Plausibility Applied by the Court

The Court’s engagement with the plausibility threshold is not consistent. In 2009, the Court’s Order in Obligation to Prosecute or Extradite equated the “plausibility” of rights with their “possibility”, as the Court was satisfied that “the rights asserted [are] grounded in a possible interpretation of the Convention [… and] therefore appear to be plausible”.⁷² In its 2011 Order in Certain Activities in the Border Area, however, the Court went further. The rights invoked were determined to be plausible on the basis that both parties asserted rights based in their claimed sovereignty over disputed territory. However, in arriving at this conclusion, rather than merely noting that a possible basis for the claim existed, the Court actively assessed the treaties and other authorities on which the two States had based their respective claims.⁷³ Crucially, however, though the Court required proof that the legal argument was plausible, it assumed the factual conditions asserted by each State would ultimately be found to exist on the merits, and did not go so far as to enquire into the plausibility of those factual conditions themselves.⁷⁴ The approach of focussing on legal, rather than factual, questions was further reinforced in its 2014 Order in Certain Documents and Data, where the Court enquired carefully into the plausibility of the rights asserted. Recalling that the “equality of the parties must be preserved” in international dispute settlement, the Court opined that disputing States “would expect” that the opposing Party not interfere in the preparation of its case; it therefore “would follow that in such a situation, a State has a plausible right” to be free of such interference, including with respect to its communications with counsel.⁷⁵ Under this approach, the primary distinction between the threshold at the interim measures stage and that at the merits is the absence of factors linking the rights in question to the specifics of the case before the Court; that is to say, the applicability of the law to the facts, and consideration of any exculpatory clauses or circumstances precluding wrongfulness.⁷⁶

The ambit of the test was, again, expanded in the Court’s 2017 Order in Application of CERD (Ukraine v. Russia), however: for the first time, the Court held that the plausibility requirement was not met on the basis that factual elements of the claim had not been established.⁷⁷ A further rise in the plausibility threshold can be found in the 2018 Order in Treaty of Amity,⁷⁸ in which the potential breach of Iran’s rights by certain US acts was assessed under the two States’ 1955 Treaty of Amity, Article XX of which carves out certain exceptions to the general regime of the Treaty for “fissionable materials” and for actions “necessary to protect […] essential security interests”.⁷⁹ For the first time, the Court thus considered a potential defence as part of its plausibility assessment.

Taken chronologically, this selective line of case law hints at a rising standard on plausibility, but that is not the whole story. For example, the lower “possibility” standard remains the best reading of the Court’s 2018 Order in Application of CERD (Qatar v. UAE),⁸⁰ despite being rendered during the same period as the higher thresholds used in the Application of CERD (Ukraine v. Russia) and Treaty of Amity orders. In the 2020 Order in The Gambia v. Myanmar,⁸¹ a low standard – only somewhat higher than de minimis – was applied, and the Court did not assess factual elements of the genocide definition, including the presence of intent, thus directly contradicting its Order in Application of CERD (Ukraine v. Russia). It would be more correct, as has been argued by Sparks and Somos, to regard the standard as oscillating wildly.⁸²

The recent Application of the Genocide Convention (Ukraine v. Russia) Order further contributes to this oscillation, and merits further scrutiny. Ukraine’s submission did not contend that there had been a breach of the prohibition against genocide, but instead, invoked a series of interlinked claims: that Ukraine could assert the rights “not to be subject to a false claim of genocide”, “not to be subjected to another State’s military operations on its territory based on […] the Genocide Convention”, and “not to be harmed by [… Russia’s] military action falsely cloaked as one undertaken to prevent and punish genocide”.⁸³ Rather elliptical on the face of it, the Court’s cautious response was to give weight to Ukraine’s claims, pointedly referring to “other means” for a party to fulfil its obligation to prevent or punish genocide than the use of force, particularly in the light of Article 1 of the Charter.⁸⁴ On the factual question, the Court noted curtly that it was “not in possession of evidence substantiating the allegation […] that genocide has been committed on Ukrainian territory”, and that it was doubtful that the Genocide Convention authorises the “unilateral use of force in the territory of another State for the purpose of preventing or punishing an alleged genocide”. Accordingly, in a tightly worded conclusion, “Ukraine has a plausible right not to be subjected to military operations […] for the purpose of preventing and punishing an alleged genocide”.⁸⁵ The Court’s relative elasticity in upholding legal plausibility, in the absence of a factual enquiry, seems to nudge the standard back towards a lower threshold, though future oscillations ought to be expected.

4.2 Plausibility in a Climate Change Case

Turning now to the specifics of a climate change case, the inconstant standard for plausibility poses obstacles that an applicant State will need to anticipate. Out of precaution, we consider the best approach would be to ensure that the more exacting standard in Application of CERD and Treaty of Amity is met. That is to say, the applicant will need to establish to the satisfaction of the Court the existence of the right(s) claimed in the abstract (i); that those rights apply to the factual circumstances at issue in its claim (ii); and the absence of any defences or countervailing principles (iii).⁸⁶ Notwithstanding the 2022 Order in Allegations of Genocide, there remain hints in the Court’s case law that these elements will be given more prominence than the de minimis “possibility” threshold that it appeared to apply in earlier cases.⁸⁷ As the higher threshold raises both legal and factual arguments, and in the case of an application premised on the customary law “no-harm” principle, it is likely to be the factual element (step (ii)) which presents the most significant potential obstacle. We consider, however, that a carefully designed application can be constructed so as to surmount the plausibility threshold.

(i) Existence of the right in the abstract. First, the existence of the customary “no-harm” rule will need to be demonstrated plausibly to exist as a norm of customary international law. Abundant evidence exists to satisfy this requirement. Already in Corfu Channel, the ICJ acknowledged the existence of “every State’s obligation not to allow knowingly its territory to be used for acts contrary to the rights of other States.”⁸⁸ Though not specific to environmental harms, the Court has extended the principle to environmental concerns in the Threat or Use of Nuclear Weapons advisory opinion,⁸⁹ in Pulp Mills,⁹⁰ and most recently to activities “causing significant damage to the environment of another State” in Activities in the Border Area and Construction of a Road.⁹¹ The International Tribunal for the Law of the Sea (ITLOS) has further extended the ambit of the customary law principle to areas beyond national jurisdiction, notably to the deep sea bed and to other areas which it characterised as “the common heritage of [hu]mankind.”⁹²

For all this, an obstacle exists inasmuch as these judgments evince no absolute right to be free of transboundary environmental harms. The “no-harm” rule is not an obligation of result, but one that requires States to comply with the requirements of due diligence and thereby demonstrate their best efforts to prevent significant transboundary harm.⁹³ In using “all means at [their] disposal in order to avoid activities which take place in [their] territory […] causing significant damage to the environment of another state”,⁹⁴ States are to ensure “not only the adoption of appropriate rules and measures, but also a certain level of vigilance in their enforcement and the exercise of administrative control applicable to public and private operators.”⁹⁵ In addition, due diligence encompasses procedural obligations, i.e. that States must “undertake an environmental impact assessment where there is a risk that the proposed industrial activity may have a significant adverse impact in a transboundary context”.⁹⁶ The impact of that qualified, procedural nature of the “no-harm” rule will primarily be felt in the second phase of the Court’s analysis, to which we now turn.

(ii) Applicability of the “no-harm” rule. Once a customary rule is established to restrain – even if only through procedural means – transboundary harms, an applicant will have to establish both that the due diligence threshold has not (prima facie, at least) been met, and that the harms associated with climate change fall within the scope of the rule. These are not merely considerations for the merits, and will need some articulation at this stage to meet the plausibility threshold. Therefore, from the outset, the applicant must assert the principle in Corfu Channel, that States have a right not to endure harms arising from the territories of other States,⁹⁷ as encompassing also the connected obligation to exercise due diligence in assessing whether an action is at risk of causing “significant adverse impact[s] in a transboundary context”.⁹⁸ Once it is established that such harms are threatened, additional obligations engage a respondent State to “notify and consult in good faith with the potentially affected State” in order “to determine the appropriate measures to prevent or mitigate that risk.”⁹⁹

In the context of climate change, the enquiry will necessarily focus on whether States remain under a corollary obligation to act in accordance with the mitigation measures that are identified as the outcome of the due diligence process. The existence of such an obligation would benefit the applicant State, which would potentially be able to assert connected procedural obligations – and breaches of these – to sidestep trickier issues relating to scientific causation, proof, and evidence, and to focus instead on the inter-connected procedural obligations which are indispensable for the effective implementation of mitigation measures. Formal obligations of this kind are more familiar ground for judges, counsel, and States than the intricacies of scientific causation, and it may well be more attractive to an applicant to focus on arguing that the respondent is in breach of these – eminently “legal” – procedural obligations.

However, a respondent could plausibly argue that the obligation has been met differently. That is to say, it may seek to argue that States in general have identified the potential harms of climate change rising above a certain temperature target through the conclusion of the Framework Convention and the Paris Agreement; that they have consulted through the Paris process and elsewhere; and that they have determined appropriate responsive measures through submission of their Paris Agreement NDC s. The Court might well consider these general procedural processes to be sufficient to discharge the specific due diligence obligation. However, there are two further weaknesses with such a line of reasoning. Foremost, even if current NDC s would be implemented fully, such combined efforts are still not consistent with the objective of keeping the global average temperature increase below 2 or 1.5 degrees Celsius.¹⁰⁰ Moreover, as one of the present authors has contended elsewhere, such reasoning relies exclusively on multilateral processes to discharge what are, essentially, divisible obligations of each party.¹⁰¹ While obligations pertaining to mitigation measures and the prevention of climate change-related harms might arise in a collective context and oblige States to work towards a common goal, they are divisible insofar as each State party remains bound independently to do its share. A collective approach to performance, such as engaging in cooperation in multilateral fora, would arguably be necessary to satisfy such obligations, but reliance on multilateral processes is not in and of itself sufficient to fulfil one’s share of the obligation.

Even once procedural elements are satisfied, the content of the substantive obligation – to act so as to mitigate harm – remains to be identified. Though a question for the merits, such considerations are salient at the provisional measures stage, as an applicant State will need to establish plausibly that such an obligation both exists and is not being satisfied.

Conceptually, we believe that there remains some latitude in how to determine the precise content of such obligations, and the possible factors informing their interpretation, not least as no international judicial body or authoritative body has yet engaged with such an obligation. We take the view that such an obligation plausibly exists, both as a self-standing obligation under customary international law, and as flowing from the structure of the obligation of consultation (read in combination with good faith). The influential statement on customary international law in Trail Smelter makes the point straightforwardly:

[U]nder the principles of international law […] no State has the right to use or permit the use of its territory in such a manner as to cause injury by fumes in or to the territory of another or the properties or persons therein, when the case is of serious consequence and the injury is established by clear and convincing evidence.¹⁰²

That statement of principle has not subsequently been refuted, and tallies closely with the ICJ’s conclusions in Corfu Channel, Pulp Mills, and Certain Activities.¹⁰³ To recall the general proposition in Corfu Channel, each State is under an “obligation not to allow knowingly its territory to be used for acts contrary to the rights of other States”.¹⁰⁴ This, in turn, implies an obligation that is to some degree tied to a concrete result, albeit one modulated by due diligence: when faced with knowledge of a given risk, the State is required to act in good faith to take all necessary mitigation measures.

Such a construction of the no-harm rule is broadly in line with prevailing practice and a range of treaties, and there is good reason to believe such reasoning should prevail in judicial proceedings. However, some caution regarding the corollary obligation to exercise due diligence is required, given that the Court has not yet arrived at a finding of environmental harm engendered by a failure to exercise due diligence in meeting a legal obligation. But this is a question for the merits: a carefully-worded assertion of this claim ought easily to surmount the lower “possibility” threshold,¹⁰⁵ and in our view it could also satisfy the more exacting standard adopted by the Court in certain later cases.¹⁰⁶

A final element is required in order for the no-harm rule to be applicable, given the largely indirect nature of climate-related harms. It will be necessary to assert in principle – and to the standards demanded at the interim protection stage – that the no-harm rule can apply to harms of a kind caused by climate change. A degree of factual uncertainty, in the strictly legal and not scientific sense, will need to be overcome. For example, carbon dioxide emissions indisputably contribute to climate change, but it is not the case that specific releases of carbon dioxide by a State are – in and of themselves – directly harmful to its neighbour: carbon dioxide mixes evenly in the atmosphere and does not give rise to local effects around the emissions point. No individual instance of damage can thus be attributed to a particular source of carbon dioxide (that is to say, attributed to a particular State’s (in)action); rather, it is the cumulative effect of all present and historical carbon emissions which causes the harms. The Court’s reasoning in Pulp Mills, which took a highly individualised, reductionist approach to attribution of harms, thus requires adaptation to the variegated causes of climate change, and remains a challenge even if the no-harm rule applies.¹⁰⁷

For an applicant, it would be worth exploring the extent in which such challenges may be overcome by relying on recent advances in attribution science. Not only can attribution science establish factual links between global greenhouse gas emissions and specific harms, such as extreme weather events, sea level rise or ocean acidification, it has also developed methods that “enable quantifying individual emitters’ marginal contributions to extreme weather events and slow-onset changes”.¹⁰⁸ On this basis, it is possible to establish the extent in which one respondent’s conduct has increased the intensity or likelihood of concrete climate change-related harms.¹⁰⁹ By relying on such methods one could for instance argue that a State contributes to a real risk of harm and that an individual emitter “can, by taking preventative measures, decrease the likelihood or intensity of harm.”¹¹⁰ Seeing that such methods have not yet been developed for all types of climate impacts, a prospective applicant would need carefully to select the types of harms that would be the subject of a request for provisional measures.

Beyond this, a potential obstacle is the probability-based methodology of scientific attribution. A respondent State may assert that these methodologies do not establish the level of certainty required for legal attribution,¹¹¹ in particular in relation to weather-related harms.¹¹² Although increasingly events will materialise to which this proviso does not apply – including harms associated with sea level rise, glacial melting events, and changes in the habitable ranges for crops or animal species¹¹³ – scientific studies generally indicate the extent to which climate change has made a weather event more intense or more likely, but it is not immediately clear that this would satisfy legal tests for causation.¹¹⁴ Within the scope of this article, it is not possible to assess such questions in detail, given that any answer will be entirely dependent on the factual circumstances; but neither is it necessary to do so as the threshold of plausibility requires a different standard of evidence than at the merits.

A further and crucial hurdle to clear is that climate harms must be established unambiguously to fall within the category of “transboundary” harms. This question has two aspects, only the first of which is relevant to plausibility: first, whether a harm with global causes can plausibly fall within the ambit of an obligation to prevent “transboundary” harms; and secondly, whether in the light of that global cause, a provisional obligation to prevent can legitimately be applied to a single State.¹¹⁵

In relation to the former, it is certainly the case that climate harms bear little relation to the classic paradigms of transboundary environmental harm (escapes of gasses or chemicals, for example): climate harms have global and multivector causes. Rather, they often originate at the scale of weather systems or oceans. Nevertheless, we contend that climate harms fall within the scope of the obligation of a State “not to allow knowingly its territory to be used for acts contrary to the rights of other States.”¹¹⁶ Should this be so, the lack of a direct transfer of something harmful from one territory to another does not present an obstacle: the release of carbon dioxide from the territory of the respondent State into the global atmosphere has the direct consequence of increasing the risks of global extreme weather and other environmental harms, and thus a fortiori also on the territory of the applicant.

Whether that nexus is sufficiently strong is the core element of the culpability assessment, and as such its full determination should remain a question for the merits. On the basis of the foregoing, we believe the Court could declare as plausible a provisional measures request premised on a violation of the no-harm rule.

5.1 Defining the Threshold of Irreparable Prejudice

“Irreparable” prejudice was first discussed by the PCIJ in the Sino-Belgian Treaty case, where it determined that an order for interim protection required an impending breach of obligation that “could not be made good simply by the payment of an indemnity or by compensation or restitution in some other material form”.¹²³ This narrow approach was maintained in all cases of interim protection before the PCIJ,¹²⁴ and construes the concept of irreparability as purely a question of legal irreparability. The ICJ similarly linked irreparability to the impossibility of reparation in Aegean Sea, where it decided against the indication of interim measures because “the alleged breach by Turkey [to Greece’s claim of an exclusive right to acquire information concerning the natural resources of the continental shelf] is one that might be capable of reparation by appropriate means”.¹²⁵ When considering Argentina’s request for provisional measures in Pulp Mills, the Court suggested, with little further explanation, that it was “unconvinced” of the imminent risk of irreparable prejudice, though it did note that the pulp mills would not be operational for four years.¹²⁶

To fixate rigidly on irreparability in law would surely be problematic for a potential climate litigant; after all, one might argue that ultimately everything is susceptible to some degree of reparation. The practice of domestic and international courts shows that the economic value of a wide variety of harms that can be suffered, ranging from environmental harm¹²⁷ to the loss of human lives, has been deemed capable of quantification in legal proceedings.¹²⁸ If such an approach is maintained rigidly, a climate applicant would be required to establish irreparability in law, and thus the (insurmountable?) task of arguing that loss of biodiversity, human lives, or even of territory¹²⁹ resulting from climate change are not only impossible to restore to the status quo ante, but are also all incapable of financial valuation.

Fortunately, the Court’s recent practice suggests that the strict threshold requiring irreparability in law has been abandoned,¹³⁰ and it appears that the Court now adopts a margin of appreciation that allows examination of all circumstances of the case.¹³¹ Within that more contextual approach, the Court now seems to treat the factual impossibility of restoration of the status quo ante as the criterion for irreparability.¹³² Accordingly, in the case that the harm alleged would impede the provision of restitution (presuming a final judgment in the applicant’s favour), a request for interim measures will be granted “irrespective of the capacity of subsequent remediation or compensation.”¹³³ In Immunities and Criminal Proceedings, a risk of irreparable prejudice was held to exist in view of the fact that breaching the inviolability of diplomatic premises “may not be capable of remedy, since it might not be possible to restore the situation to the status quo ante”.¹³⁴ According to Oellers-Frahm, these cases “reflect the Court’s view that where the action causing a risk of irreparable prejudice has not yet occurred, the status quo ante should be maintained.”¹³⁵ Applied to the climate context, the most recent IPCC report warns that many changes have already been set in motion by greenhouse gas emissions which will be irreversible for centuries to millennia¹³⁶ and moreover that, without any scientific doubt, the full range of potential climate change-related harms also encompasses harm that is yet to materialise, but which will be “locked in” by failure to take immediate mitigation measures.¹³⁷ It would almost beggar belief, if the Court were to find that the status quo ante is recoverable in the face of climate change-related harms, and on this ground were to refuse an order to ensure this harm does not materialise before a judgment on the merits.

A similar cause for optimism may be found in recent cases, where commentators have noted that requests for provisional measures have been granted by the Court “when an obvious and flagrant violation of the rights claimed on the merits cannot be tolerated until the delivery of the final judgment”.¹³⁸ The 2022 Order in Allegations of Genocide under the Convention (Ukraine v. Russia) is illustrative of this point, with the Court highlighting the “extremely vulnerable” situation of the civilian population, in particular the deteriorating humanitarian situation, to conclude that there was a “real and imminent risk” of irreparable prejudice, and thus sufficient urgency to grant the Order.¹³⁹ In such cases the criterion would seem to be whether the prejudice that would be suffered by the applicant would be “unendurable”.¹⁴⁰ It is primarily in cases involving armed activities that such an approach has been implemented to date,¹⁴¹ but it is feasible to extend that reasoning, mutatis mutandis, to the situation of a population at risk from climate change-related harms which, after all, could cause suffering on similar or greater scales.

The above surely offers hope for a potential applicant in a contentious climate case: illustrating the unendurability of climate change-related harms, or the impossibility of restoration to the status quo ante, certainly appears to be within reach. In recent years, the ICJ has recognised certain forms of irreparable prejudice relevant here, specifically, risks to human life or health,¹⁴² and environmental damage.¹⁴³

5.2 Establishing the Urgency of Climate Change-Related Harms

As with irreparability, the settled practice of the Court established over time that provisional measures will not be indicated unless the risk of prejudice is “urgent”,¹⁴⁴ that is to say, where “action prejudicial to the rights of either party is likely to be taken before the Court has given its final decision”.¹⁴⁵ Urgency is habitually understood to be implicit in the notion of “provisional measures”,¹⁴⁶ in that the Court will not make definitive findings of fact, but to order a provisional measure will require evidence of both a temporal element – a “real and imminent” risk – as well as the degree of likelihood that the foreseen prejudice will in fact materialise:

The ICJ does not require that [a claim] be proved with absolute certainty, though at the same time it is not sufficient for the applicant to assert that prejudice is hypothetically possible and that relief should be granted accordingly. Rather, the Court’s recent jurisprudence requires that there be demonstrated a ‘real and imminent risk’ of irreparable prejudice, the precise threshold of which is to be assessed on a case-by-case basis by reference to the ongoing factual matrix.¹⁴⁷

Though a determination of urgency remains a factual appraisal, its assessment retains an element of subjectivity and discretion, and the Court’s discussions of urgency tend to be brief to the point of being sibylline. For example, in its Order in Application of CERD (Georgia v. Russia), the Russian Federation had submitted that Georgia’s submission on urgency was unsound, being “built almost exclusively on inference”.¹⁴⁸ Without analysing the submission in detail, the Court curtly stated that “there exists an imminent risk that the rights at issue in this case […] may suffer irreparable prejudice”.¹⁴⁹

Assessments of whether or not future harm will occur (and on what scale) have an unavoidable tendency to be speculative, and the scientific nature of the subject matter at hand in a climate change-related dispute raises its own challenges. On the one side, there exists a wealth of data, above all the reports from the IPCC, an intergovernmental expert body under the auspices of the World Meteorological Organisation and the United Nations Environment Programme, the transparency, methodology, and reliability of which is regarded as a gold standard in the scientific community.¹⁵⁰ Accordingly, a greater volume of scientific material and of a higher standard of reliability is available than in almost any previous case.¹⁵¹ That is a double-edged sword, however: the Court would also be faced with a set of highly complex technical and scientific issues, including that of scientific attribution (for instance, linking climate change to the occurrence of extreme weather events), which may be unintelligible to the international judge. A particular barrier is that scientific conclusions on climate change – whether with regard to specific instances of harm or long-term assessments of likely developments under specific emissions scenarios – are generally expressed in probabilistic terms, rather than as statements of fact.¹⁵² As noted above,¹⁵³ this invites the contention that the available scientific evidence does not satisfy exacting legal standards, including the requirement of “evidence that is fully conclusive” required at the merits phase.¹⁵⁴ This is not to say that such contention will necessarily be successful; further substantive efforts are ongoing to develop the tools to “translate” scientific findings into legal terms capable of satisfying the Court.¹⁵⁵

Provided that these challenges can be overcome, however, it seems highly likely that an applicant State will be able to establish an imminent risk on the basis of the available scientific evidence. The Court has already warned against the inherent limitations on the mechanism of reparation in situations of irreversible environmental damage.¹⁵⁶ The most recent IPCC Report warns emphatically that immediate mitigation measures are necessary in order to avoid a range of new, potentially disastrous climate change-related harms.¹⁵⁷ It is not the case that the full range of envisaged future harms, such as the above-mentioned risk to human life or health, loss of territory due to sea level rise or environmental damage, will unfold before a decision on the merits by the Court; but, crucially, the mitigation measures needed to avoid those harms from materialising would need to take place within that period. Failure to mitigate immediately risks “locking in” future rises in temperature and the concomitant harms, hence making it likely that “[in]action prejudicial to the rights of either party is […] taken before such final decision is given.”

Finally, the seriousness of the risk also satisfies the criterion of imminence where “irreparable damage resulted from the risk to human life on a large scale”,¹⁵⁸ as the Court has done repeatedly in cases concerning genocide, and other direct harms to individuals.¹⁵⁹ Moreover, such grave risks to human life and health have effects well beyond the parties to a contentious dispute. The Court is alert to such broader effects: in its 1973 Order on Provisional Measures in Nuclear Tests, it noted that:

[T]he information submitted to the Court […] does not exclude the possibility that damage to New Zealand might be shown to be caused by the deposit on New Zealand territory of radio-active fallout resulting from such tests and to be irreparable. […] The Court is satisfied that it should indicate interim measures of protection in order to preserve the right claimed by New Zealand in the present litigation in respect of the deposit of radio-active fall-out on the territory of New Zealand, the Cook Islands, Niue or the Tokelau Islands.¹⁶⁰

We would argue that the 1973 Nuclear Tests Order further informs the successful request by The Gambia for the Court to indicate provisional measures to Myanmar in respect of the “common interest” of all States in respect of the obligations under the Genocide Convention.¹⁶¹ Should an applicant choose to include these harms in its request for provisional measures, it need not show that this full range of harms will itself materialise before a final decision by the Court, but could simply make the case that loss of life and other direct human harms will be “locked in” in the absence of an interim measures order.

5.3 Making the Case for Imminent Irreparable Prejudice When Harm Is the Result of a Plurality of Causes

Our final point relates to the limited jurisdictional competence of the Court in a contentious case, which simply cannot be instituted against all States contributing to climate change, but only against one or a small group of States that have accepted the ICJ’s jurisdiction. In such situations, the question may well arise as to whether to take measures against a respondent State’s contribution to climate change alone, rather than the combined conduct of a large plurality of States, would suffice to fulfil the requirement of imminent irreparable prejudice. A respondent State could well argue that granting a request for provisional measures would make no difference in the bigger picture, since its conduct alone would neither be sufficient to cause nor to prevent the imminent irreparable harm that is likely to follow from climate change. Arguments in this spirit have been attempted elsewhere,¹⁶² albeit unsuccessfully and not (yet) in the context of interim measures of protection. In Urgenda v. the Netherlands, the Dutch government argued that imposing an individual duty on the Netherlands to reduce its global greenhouse gas emissions would be to no avail in view of the fact that the Netherlands’ own share of such emissions is small, and hence would make little to no difference on a global scale.¹⁶³

Yet as the Dutch Supreme Court noted, every reduction of greenhouse gas emissions has a positive effect on combatting dangerous climate change, as every reduction means that more headroom remains to forestall climate change. Simply put, no reduction is negligible.¹⁶⁴ A similar sentiment was expressed by the UN Committee on the Rights of the Child, when it considered that the collective nature of climate change obligations and causation in respect of climate damage has not absolved States parties to the Convention on the Rights of the Child of their individual responsibility that may derive from the harm that the emissions originating within its territory may cause to children, whatever their location.¹⁶⁵ Moreover, and although States have collectively undertaken to pursue emissions reductions, these remain divisible obligations to which each State is independently bound, as one of the present authors has argued elsewhere.¹⁶⁶ Our argument is that to make each State’s obligation to mitigate contingent on the fulfilment of every other State’s mitigation obligation would be effectively to nullify the Paris Agreement and UNFCCC processes, and thus to defeat the object and purpose of those treaties. More generally, if one were to accept that the contribution of others to a collectively caused harm such as climate change would somehow excuse an individual State’s contribution to harm, this would be akin to transforming international legal rules relating to common interests into empty shells.

Nevertheless, an applicant State is likely to face the objection that a provisional measures order, even if granted and fully complied with, would make little or no difference to the advance of climate change. In our view, this has the potential to be one of the most dangerous aspects of the interim process, both for the success of an applicant’s claim and in strategic litigation terms. But it is surmountable, through detailed reference to other courts’ rejections of severability-based arguments (such as in Urgenda, Sacchi, and elsewhere), careful treatment of divisible and indivisible collective obligations, and by presenting the Court with sufficient factual information on the basis of climate science concerning the effect of the specific respondent’s emissions. This too, however, is an area which will require careful preparation.