1. Introduction
On 28 May 2025, after ten years of litigation, the Higher Court of Appeal in Hamm (Germany) decided the case Lliuya v. RWE.Footnote 1 The Court did not permit the appeal to the Federal Court of Justice and, as the claimant decided not to file a complaint against the denial of leave to appeal, the judgment is final.Footnote 2 Winners and losers in any climate litigation are often difficult to discern from a legal perspective – if less so from an ethical standpoint – and both sides in these proceedings have claimed some sort of victory.Footnote 3 Climate adjudication is ‘legally disruptive’Footnote 4 and, even though the case was not decided in favour of the claimants, it shows a growing judicial awareness of the duties owed by large emitters in responding to the threat of climate change. The case illustrates the transnational scope of German private nuisance law; it demonstrates that it is scientifically and legally possible to find a causal link between the greenhouse gas (GHG) emissions of a large emitter and the impacts of climate change, and it draws further attention to the importance of concretizing the meaning of ‘best available science’ for the application of normative thresholds and legal concepts.
More specifically, the judgment shows how nuisance law under section 1004 of the German Civil CodeFootnote 5 can be situated in the response to the global threat of climate change. The purpose of this article is to use a novel methodology that employs a combined law and science analysis of the judgment to develop an argument that goes beyond the specific case and is more widely applicable: case-specific expert evidence must remain firmly connected with and anchored in the existing body of climate science where high confidence levels have been established, to support robust legal reasoning. Using the concept of climate causality, the article explains how a baseline for best available scientific evidence comprises the universal laws of physics, the reports of the Intergovernmental Panel on Climate Change (IPCC), and relevant evidence that – even if not yet reviewed by the IPCC – is aligned with the laws of physics and the IPCC reports.Footnote 6
Central to the argument is the legal treatment of uncertainties in relation to climate causality. While causation in the context of climate change has attracted scholarly discussion,Footnote 7 this article closes a gap in clearly distinguishing between qualitative and quantitative causation in climate science, a distinction that will change the legal treatment of the uncertainties affecting changes in risks over certain periods of time, in cases such as Lliuya. It flows from our argument that for any deviation from the baseline of best available science (that is, evidence that directly or indirectly challenges, for example, existing IPCC assessments), not only is an additional scientific explanation warranted but also that the deviation will affect the confidence levels attached to the reported evidence. These confidence levels will inevitably change significantly as a result of the deviation. It resonates from the analysis that the presentation of evidence, and its subsumption within the concept of best available science, in a legal setting must evolve further to reposition and stabilize legal concepts and normative thresholds amidst the challenges of climate change.
This article proceeds in five sections. Section 2 turns to the discussion of German nuisance law and civil procedure law, and situates the case in the wider context of key developments pertaining to corporate duties in recent climate litigation. Section 3 analyzes significant legal developments derived from the judgment, and it critically evaluates the reasoning, including in respect of causation and the significance of RWE’s contributions to climate change. Section 4 examines the normative threshold of ‘imminent first impairment’ from an integrated law/science perspective, and discusses the importance of using IPCC reports as a baseline for best available science in climate litigation. On that basis, Section 5 uses the concept of climate causality to delve deeper into the requirement to differentiate, in the legal assessment, between qualitative and quantitative statements about the causal quality of a factor, in the context of a specific impact of climate change.
The overarching objective is to dispel some persistent misunderstandings afflicting legal readings of climate science, such as the role of the past in informing future risks in a rapidly changing climate, with a view to prompting more interdisciplinary research to advance the engagement with scientific evidence for legal concepts such as causality, probability, and imminence. These concepts will have not only different peculiarities but also shared characteristics across jurisdictions. Ultimately, such research can forge the knowledge to pre-emptively design legal frameworks that address the human contribution to extreme weather events. Section 6 synthesizes the main findings and concludes.
2. Employing Private Nuisance Law to Address the Risk of Receding Glaciers
Historically, private nuisance law has been concerned with the interference with another person’s use or enjoyment of their land, often in the nearby neighbourhood.Footnote 8 The Hamm Court has expanded on this traditional reading of nuisance law in a transnational context. It found that it can be possible to causally link the GHG emissionsFootnote 9 of one large emitter to a concrete risk of property damage in a country that is more than 5,000 miles away. The case before the Court involved a claimant from Peru and the energy provider RWE, a large company headquartered in Germany and responsible for large quantities especially of carbon dioxide (CO2) as a result of extracting and selling fossil fuels, including coal, for the purpose of energy generation.Footnote 10 The Court advanced the scope of private nuisance law in Germany’s civil law system, a legal framework that differs from strict liability rules applicable in common law private nuisance claims, but also shares similarities in respect of the requirement of causality and thresholds such as significance of the causal contribution.Footnote 11
Under German nuisance law, the owner of a property can request the removal of a disturbance in line with section 1004(1) of the Civil Code. In addition, if there is an imminent threat of a repeated impairment (‘further interferences’), the owner has the legal right to request the defendant (the ‘disturber’) to take measures to prevent such repeated disturbance. This can go beyond the claim to abstain from the activity that causes the disturbance and include the disturber being required to take positive action to prevent the imminent impairment by performing a so-called ‘actus contrarius’,Footnote 12 a point that is discussed at a later stage in this analysis.Footnote 13 Any right under section 1004 can exist only if there is no legal obligation for the claimant to tolerate the disturbance under either public or private law.Footnote 14
This statutory rule has been further specified in the case law of the Federal Court of Justice.Footnote 15 While the provision in question speaks of ‘further interferences’, an application for injunctive relief can also be filed where no interference with the property has occurred (yet), to address the risk of a ‘first’ interference.Footnote 16 Thus, according to the case law, the property owner can seek to prevent a first impairment where there is a risk but no actual damage to the property.Footnote 17 However, the claimant must prove that the risk of damage is imminent.
Building upon this body of law, the Court of Appeal in Hamm was asked to provide injunctive relief for an imminent first impairment. To succeed, the claimant had to prove that the GHG emissions of RWE, in particular CO2, constituted not only an increased risk of a glacial lake outburst flood (GLOF) for the claimant’s property but reached the normative threshold of an imminent first impairment. The case law has established benchmarks to determine the required threshold for ‘imminence’, and the Court found that the risk of flooding – identified by Court-appointed experts, with a likelihood of around 1% over a 30-year period – was insufficient to meet the threshold.Footnote 18
2.1. Proven or Not? Reasons that Inform the Conviction of the Judges
At first glance, it may be difficult to discern where the Court’s reasoning is based on the facts as argued and those that were sufficiently substantiated. According to the Code of Civil Procedure, it is the role of the Court to evaluate the evidence before it:
to decide, at its discretion and conviction, and taking account of the entire content of the hearings and the results obtained by evidence being taken, if any, whether an allegation as to fact is to be deemed true or untrue. The judgment is to set out the reasons informing the conviction of the judges.Footnote 19
The Court based its reasoning on the assumption that already underlay the decision to begin the evidentiary stage of the proceedings in November 2017 – that the case was ‘conclusively argued’.Footnote 20 ‘Conclusively argued’ means that the facts as stated in the claimant’s submission, if proven, would justify the claim and bring the relief sought. If the asserted facts of the claim are disputed by the defendant, the Court may order that evidence be taken, and the claimant must substantiate the contested facts upon which the claim is based, while the defendant can provide evidence for the reasons for disputing the facts.Footnote 21
While the judgment confirmed that the claimant made a valid legal argument, the Court found that the claimant had not succeeded in proving that a first impairment was imminent.Footnote 22 However, the Court reasoned that there were facts that constituted common knowledge where no further evidence was necessary.Footnote 23 Based on this common knowledge, the Court stated, for example, that it was the ‘conviction of the Senate’ (three judges are deciding) that ‘it had been foreseeable for the optimal observer in the role of an energy producer, since the mid-1960s, that anthropogenic GHG emissions would lead to global warming and the resulting harm’.Footnote 24 The Court also found that RWE’s GHG emissions made a traceable and material contribution to global GHG emissions,Footnote 25 just not one that resulted in an imminent threat of flooding in the concrete case.
2.2. Corporate Duties and State Obligations: Deep, Rapid, and Sustained Reductions in GHG Emissions
The case complements a growing field of litigation against corporations and between private actors that are transacting with each other in the shadow of legal concepts and regimes that have been altered or produced through climate law.Footnote 26 For example, in Milieudefensie et al. v. Royal Dutch Shell Plc, despite overturning the decision of the court of first instance,Footnote 27 the Dutch Court of Appeal confirmed that:
[companies] which contribute significantly to the climate problem and have it within their power to contribute to combating it, have an obligation to limit CO2 emissions in order to counter dangerous climate change, even if this obligation is not explicitly laid down in (public law) regulations of the countries in which the company operates.Footnote 28
Milieudefensie appealed against the decision to the Supreme Court of the Netherlands in February 2025.Footnote 29 It is interesting to note that a new narrative emerges from the decision of the Dutch Court of Appeal, which also permeates the decision in Lliuya v. RWE. Even though the appellate court fell short of confirming the finding of the court of first instance that Shell was obliged to reduce its CO2 emissions by net 45% at the end of 2030 relative to 2019,Footnote 30 the appeal ruling weaves in a social standard of care applicable to companies that is influenced by the horizontal effect of human rights and goes beyond public law rules (‘even if this obligation is not explicitly laid down in (public law) regulations’ – see quote above).Footnote 31 The Court stated:
[drastically reducing emissions by 2030 would require] not only taking measures to reduce demand for fossil fuels, but also limiting the supply of fossil fuels. The social standard of care, interpreted on the basis of Articles 2 and 8 [of the European Convention on Human Rights (ECHR)Footnote 32 ] and soft law such as the [United Nations Guiding Principles on Business and Human RightsFootnote 33 ] and [Organisation for Economic Co-operation and Development (OECD) Guidelines for Multinational Enterprises on Responsible Business ConductFootnote 34 ], requires producers of fossil fuels to take their responsibility in this respect.Footnote 35
Furthermore, the Court of Appeal reasoned that under the European Union (EU) Corporate Sustainability Due Diligence Directive (CSDDD) of 13 June 2024,Footnote 36 EU Member States are required to ensure that companies falling within the scope of the CSDDD, such as Shell:
adopt and put into effect a transition plan for climate change mitigation which aims to ensure, through best efforts, that the business model and strategy of the company are compatible with the transition to a sustainable economy and with the limiting of global warming to 1.5C in line with the Paris Agreement and the objective of achieving climate neutrality as established in Regulation (EU) 2021/1119, including its intermediate and 2050 climate neutrality targets, and where relevant, the exposure of the company to coal-, oil- and gas-related activities.Footnote 37
This climate transition plan must also set out time-bound targets in five-year increments from 2030 to 2050.Footnote 38 These new obligations, even if they leave some discretion, are nevertheless illustrative of the judicial acknowledgement of a responsibility to combat climate change that does not lie exclusively with states.Footnote 39
However, this shared responsibility is rooted in binding obligations of states under international law to regulate the activities of major emitters. The International Court of Justice (ICJ), in its advisory opinion on climate change of 23 July 2025,Footnote 40 seized the opportunity to confirm that states must adopt, both under the Paris AgreementFootnote 41 and under customary international law, ‘regulatory mitigation mechanisms designed to achieve deep, rapid and sustained reductions of GHG emissions’.Footnote 42 In particular, with respect to the mitigation obligation under the Paris Agreement, the ICJ specified that:
a party’s compliance with its obligations to pursue domestic mitigation measures under Article 4, paragraph 2, is to be assessed on the basis of whether the party exercised due diligence in its efforts and in deploying appropriate means to take domestic mitigation measures, including in relation to activities carried out by private actors.Footnote 43
Furthermore, the ICJ confirmed the observation of the International Tribunal for the Law of the Sea (ITLOS) that the ‘obligation of due diligence is particularly relevant in a situation in which the activities in question are mostly carried out by private persons or entities’.Footnote 44 Failure to ‘exercise due diligence by not taking the necessary regulatory and legislative measures to limit the quantity of emissions caused by private actors under its jurisdiction’ may constitute an internationally wrongful act.Footnote 45
Read in this context of the regulatory due diligence obligations of states, the judgment of the Hamm Court provides further affirmation that climate targets can be achieved only if states and private actors heed the clear imperative of making deep, rapid, and sustained reductions in GHG emissions. Regulating the activities of large emitters must therefore be embedded in a national legal framework that enables states and private actors to transition away from fossil fuels.
3. The Court’s Advancements of Private Nuisance Law
3.1. Facts and History of the Case
The legal dispute arose in respect of an alleged impairment of the property of Saúl Ananías Luciano Lliuya, a Peruvian farmer who lives in Huaraz (Peru). Huaraz is situated below the largest and most northern glaciated mountain region of the tropical Andes, the Cordillera Blanca. Underneath the glacier Palcaraju and the mountains Nevado Palcaraju and Nevado Pucaranra is the glacial lagoon Palcacocha. The front end of the lagoon (facing the glen) is held with a natural moraine dam, fortified by built structures.Footnote 46
Mr Lliuya claimed that the defendant company, the largest German energy provider RWE AG, had caused climate change impacts that affected the glacial lagoon Palcacocha through the company’s release of large quantities of GHGs, in particular CO2, arising especially from past and present coal-fired power generation. The danger for his property was a result of the reduced stability of the glacier and the increased water level in the lagoon located above the property.Footnote 47 A flood-triggering avalanche or a landslide were possible at any time.Footnote 48 A GLOF that could trigger a massive flood-wave, and take with it mud and debris, could be induced by a moraine-overtopping wave or through a breach of the dams.Footnote 49 Despite existing measures, the water level, as a result of increasing meltwater from the glacier, had again reached a dangerous level, making a GLOF very likely. A GLOF or a landslide could damage or destroy the claimant’s property.
The defendant, RWE, argued that there was no legal basis for the liability of an individual emitter for the alleged consequences of global climate change. By contrast, it was the responsibility of the legislator to regulate cumulative, distant, and long-term damage through means other than litigation.Footnote 50 The defendant disputed that there was adequate causation because CO2 emissions came from many sources and mix in the atmosphere. Natural processes in the atmosphere would lead to the unidentifiable mixing between GHGs, in addition to changes resulting from natural drivers, and the absorption of carbon through sinks.Footnote 51 It was undisputed between the parties that the GHG emissions were not prohibited by law and had been regulated – and permitted – in accordance with the Greenhouse Gas Emissions Trading Act since 2011.Footnote 52
The argument concerning lack of causation had convinced the court of first instance. The Regional Court in Essen dismissed the claim in its judgment of 15 December 2016.Footnote 53 It held that whether an impairment of the claimant’s property existed in the form of an acute flood risk could be left undecided as, in any case, there was no equivalent causation of the impairment. It found that the defendant’s GHG emissions would not constitute a causal factor for the purported flood risk, as the contributions of all emitters were indistinguishably mixed in the atmosphere.Footnote 54 The Higher Court of Appeal disagreed with this view and, following the scientific evidence, acknowledged the causal quality of RWE’s GHG emissions, as is explained in the following.
On appeal to the Higher Court of Appeal in Hamm, the claimant argued that RWE’s CO2 emissions were partially causal for the present danger to his property and that this form of contributory causation was sufficient to meet the threshold of section 1004(1) of the German Civil Code. Critically, without the defendant’s contribution, the risk of flooding would be lower, and the pro rata share of the impairment was estimated to be proportional to RWE’s share of emissions at 0.38% or, alternatively, should be determined by the Court.Footnote 55
RWE argued that the claim was inadmissible and unfounded. The company submitted that there was no immediate risk of a GLOF, and no individualised causal relationship could be established between its emissions and climate change or the alleged flood risk. There was no possibility to determine linear causality between its own emissions and climate change because of the incomprehensible number of emissions and the natural and chemical processes in the atmosphere.Footnote 56 Furthermore, the company disputed the alleged historic contribution of 0.47% to the global GHG emissions.Footnote 57 Fundamentally, RWE denied that climate models could prove causality in the legal sense.Footnote 58 RWE also stated that the claimant had settled below the lagoon with knowledge, or at least with negligent lack of knowledge, of an existing source of pollution and had built without planning permission, which meant he now had to tolerate the situation.Footnote 59
None of these arguments persuaded the Senate.
3.2. The Substantive Legal Assessment of the Appeal
The Higher Court of Appeal found that the claim was admissible in so far as the claimant sought a declaratory judgment regarding the defendant’s obligation to bear the costs pro rata for the asserted first imminent impairment of his property.Footnote 60
Jurisdiction
The Court stated that its international jurisdiction was founded on the EU Brussels I Recast Regulation on jurisdiction,Footnote 61 and based also on the defendant’s agreement to participate in the oral proceedings in the first instance, according to German civil procedure (section 39 of the Code of Civil Procedure).Footnote 62 The claimant had the necessary legal interest in a declaratory judgment, the claim was sufficiently precise, and German law was applicable, meeting the requirements of the EU Rome II Regulation on applicable law.Footnote 63 Furthermore, both parties had also agreed in November 2017 that German law was applicable.Footnote 64
The fact that Mr Lliuya lived in Peru was thus no hindrance to bringing a claim under German civil law. The Court of Appeal referred to the jurisprudence of the Federal Court of Justice according to which the distance between the source of the disturbance and the affected property is irrelevant; proximity is not a prerequisite for a successful claim, both according to the wording and the meaning and purpose of the provision.Footnote 65 The application of section 1004 of the Civil Code was also not excluded based on the consideration that protective measures could be required either for the lagoon or for the property of the claimant.Footnote 66
Causation
The meaning of ‘disturber’ (the party interfering with the claimant’s property) requires an actual or imminent impairment of the property through action or omission.Footnote 67 The question, therefore, was whether RWE had caused a disturbance by emitting large quantities of CO2, a question that the Court answered positively. It is important to note the nuances here. The Court identified a causal relationship between RWE’s emissions and the disturbance – that is, the impacts of climate change on the increase of the risk for the claimant’s property – but did not find that the normative threshold of a sufficiently imminent impairment had been reached.
In its elaboration to determine causation, the Court started by stressing that legal causality – and not causality as understood in the natural sciences – was relevant. This clarification dispelled the defendant’s argument that challenged the suitability of climate models to evidence causation, as these models were not necessary to determine causation in the case at hand.Footnote 68 The Court then confirmed that the GHG emissions of RWE, including those of its subsidiaries, were causal in the legal sense, based on the following reasoning.
As a first step, legal causation requires the application of the equivalence theory and, in a second step, must be determined according to the theory of adequate causation. The theory of equivalence uses the ‘conditio sine qua non’ test, which is similar to the ‘but for’ enquiry in common law.Footnote 69 Both tests require a counterfactual analysis to qualify a factor as a cause if it can be determined that the event would not have happened without the factor in question. Footnote 70 The theory of adequate causation is then used to eliminate unlikely factors from the causal chain.Footnote 71 The test reduces the factors that are legally relevant to those that fall within the scope of foreseeability for the optimal observer. Accordingly, a factor is adequately causal only ‘if it is capable of bringing about a result of the kind that has occurred, and not only under particularly peculiar, completely improbable circumstances that must be disregarded in the normal course of events’.Footnote 72
The Court first applied the conditio sine qua non test and confirmed that, based on the claimant’s submission, the defendant was a direct disturber, and that there was a direct causal relationship between the CO2 emissions, the increase in CO2 concentrations in the atmosphere, and the increase in risk for the claimant’s property.Footnote 73 In the second step, the theory of adequate causation was used,Footnote 74 taking into account all circumstances that would have been recognizable to an optimal observer at the time of the occurrence of the damaging event, or the act/omission causing the damage.Footnote 75 The Court clarified that in those cases where the point in time at which the damage occurs is significantly later than the damaging action or omission, the decisive point in time to determine foreseeability is the moment of the activity or omission in question.Footnote 76
Based on generally known facts, for an optimal observer in the role of an energy producer, it was already foreseeable in the mid-1960s that anthropogenic GHG emissions would lead to global warming and the associated adverse consequences.Footnote 77
Citing the scientific measurements and evaluations of the climate researcher Charles D. Keeling, the Court found that evidence for the warming associated with a steadily increasing CO2 concentration in the atmosphere was already available in 1958.Footnote 78 Keeling established that the increase in CO2 expected from the burning of fossil fuels was measurable and significant, and would cause global warming with consequences such as the melting of the ice caps, a rise in sea level, and a warming of seawater.Footnote 79
In contrast to the view of RWE, and based on the available scientific knowledge and, in particular, the public debate of the scientific evidence, the company could have recognized that the CO2 emissions generated as a ‘waste product’ of its coal-fired power generation were capable of contributing to the melting of glaciers as a result of the atmospheric greenhouse effect.Footnote 80 This would also be as a result of ordinary physical processes linked to the emissions and not a result of entirely unusual, unforeseeable circumstances that needed to be exempt from the causal chain. Considering this scientific knowledge would not have required the defendant to acquire special scientific expertise.Footnote 81
Furthermore, neither the lack of specific knowledge nor the complexity of natural science processes exclude foreseeability. Especially energy-producing companies of the size of the defendant could be expected to follow the results of scientific congresses and specialist events as well as evaluate the international specialist literature.Footnote 82 It is important to note that the Court at this point was convinced, based on the facts that it considered to be common knowledge, that a causal link between RWE’s GHG emissions and its contribution to climate change and the asserted increase in risk existed.Footnote 83 While the imminence of impairment could not be substantiated, the causal quality of RWE’s GHG emissions had been proven.
Significance of contributions to global emissions
The Court found the CO2 emissions of RWE to be significant. It disagreed with the defendant’s view that the case was symptomatic of an ‘instrumentalization of the judiciary’Footnote 84 and that the case would lead to claims by ‘everyone against everyone’ or ‘all against all’, given that there was a difference in the size of causal contributions where some of these significantly increased the risk.Footnote 85 Noting that the defendant’s argument in that respect was more of a political nature, the Court nevertheless explained that even if the defendant was one among many disturbers, this would not prevent the claimant from choosing to file a claim against RWE as a specific defendant making a measurable contribution to climate change.Footnote 86 In that respect, the Court explained that neither the defendant’s large share of 0.38% of industrial CO2 emissions nor its share of just under 0.24% of worldwide CO2 emissions would appear negligible.Footnote 87 Moreover, in the list of the world’s 81 largest CO2 emitters, the defendant ranked 23rd.Footnote 88 The Court concluded, therefore, that RWE’s share of industrial CO2 emissions worldwide qualified as a significant contributing factor of global warming and its consequences.
Unlawful situation
The Court disagreed with the argument that to establish liability the claimant had to substantiate a breach of a duty, noting that this reading of the law (as suggested by RWE) would correspond neither to the wording nor to the purpose of section 1004 of the Civil Code. According to the Court, the term ‘breach of duty’ belonged to tort law (sections 823 and following of the Civil Code), which was systematically different from nuisance law, where a completely lawful activity can still lead to an unlawful situation that the property owner does not have to tolerate.Footnote 89 In short, the impairment of the property must be unlawful, but not the activity leading to it.
The defendant’s argument that the claimant was under a legal obligation to tolerate the emissions because they were covered by German emissions regulations and, thus, not against the law, were therefore not accepted. Having a public law permit to emit under public law does not exempt a company from its obligations under civil law.Footnote 90 Furthermore, no duty to tolerate could be derived from the claimant’s choice to live below the glacial lagoon. In that respect the Court pointed out that it had taken the opportunity to visit the area and take evidence by eyesight in Peru. On that basis, there was no reason for the Court to believe that the local authorities were unaware of the settlement, so that at the very least there was a form of tacit approval.Footnote 91
4. The Challenges Posed by the Issue of Imminent First Impairment
Two critical numbers defined the Court’s analysis of ‘imminence’. Firstly, the period of 30 years; secondly, the factor that measured the probability of the potential impairment with 1%. In this section, we identify five points where the Court’s legal engagement with the scientific evidence is incomplete and could be strengthened by fully recognizing the scientific probabilities and levels of certainty.
4.1. The Normative Threshold of ‘Imminence’
According to the Court, the claimant, who incurred the burden of proof, failed in proving that a first imminent impairment of his property existed that would meet the threshold under section 1004 of the Civil Code. As mentioned earlier, to limit the scope of possible property interferencesFootnote 92 not any potential or abstract risk will qualify, only one that amounts to a concrete, immediately to be expected, imminent impairment.Footnote 93 The Civil Code is silent in terms of measuring the time span that suggests imminence. The Federal Court of Justice has established that a first impairment is imminent if there is a serious threat of a future, soon to be expected, or tangible event that makes a legal assessment possible.Footnote 94 Therefore, there are two intertwined factors that determine the legal assessment of imminence: (i) proximity in time and (ii) seriousness of the threat. The literature uses terminology such as ‘as soon as’, ‘in the foreseeable future’, ‘to be seriously and tangibly feared’.Footnote 95 In line with the case law, the Hamm Court of Appeal confirmed that the severity of the risk and the protected interests at stake had to be considered for the overall analysis of imminence.Footnote 96
Nevertheless, based on the evidence, the Court found that damage through flooding from the glacial lagoon Palcacocha and/or a mudslide – as a result of an increase in the volume of water in the lake, the release of an ice avalanche, a glacier collapse, a rock slide, or a combination of these circumstances – could not be expected with the probability required.Footnote 97 The Court-appointed experts specified that the risk of a large rockfall (Millionensturz) based on a 30-year observation period considered by the expert was around 0.5%, with a maximum probability of 4.7%, after applying a ‘climate factor’ to account for the changing risk of a GLOF-triggering avalanche as a result of climate change, as suggested by the claimant’s experts.Footnote 98
There was disagreement between the experts over the exact value of this climate factor.Footnote 99 For the Court, however, it was decisive that the experts also asserted that the concrete risk for the claimant’s property would be even lower, because of its location in relation to the lagoon.Footnote 100 In assessing the concrete risk, the Court also followed the views of its own experts, which disputed the evidence provided by the claimant’s expert concerning the changes in the distribution of permafrost and the resulting destabilization of the mountain side.Footnote 101 According to the Court’s experts, there was no relevant permafrost thawing below 5,000 metres, and no changes to the permafrost distribution had occurred during an earlier landslide in 2003 (which, according to the facts of the case, included significant changes through breaking of glacier ice, a flood wave, and increase in water volume in the lagoon).Footnote 102 The Court was also convinced by its own experts that there was no possibility to seriously consider the developments of the lagoon beyond a time frame of 30 years.Footnote 103 Therefore, while the Court had found earlier that there was an increased risk, this was not sufficient to meet the normative threshold of imminence.
4.2. Methodological Choices that Determine Time Frame and Risk Factor
The following five points relate to determination of the time frame and the risk factor based on the Court’s appointed expert evidence. These points not only merit further consideration, they also reveal why agreement on what constitutes ‘best available science’ must involve IPCC reports, as we explain in the following section.
Firstly, in accepting the analysis of the Court’s appointed experts that 30 years define the relevant time span for a legal assessment, the Court cited reasons such as the impossibility to seriously estimate the developments affecting the lagoon, the unpredictability of changes that occurred in the past, and the contribution of factors such as the economic situation in Peru and the growth of the population in Huaraz.Footnote 104 While this may explain the Court’s view to some extent, these reasons are not compelling for the determination of the cut-off date. Clearly distinguishing between qualitative and quantitative causality at this point may have prevented the Court from assigning a quantitative accuracy to the expected 1% risk of flooding over the next 30 years. The assumed certainty of a link between the increase in GHG emissions and the GLOF risk can be given only for the qualitative assessment – that is, the causal quality of RWE’s CO2 emissions for the change in the risk – but no equivalent quantitative certainty of the concrete contribution and the expected changes in the risk for the next 30 years can be established in a rapidly changing climate for a mountainous region with little observational data.Footnote 105 Therefore, the Court’s legal assessment of the scientific data at this point should have considered the differences between the qualitative and the quantitative assessment of the causal link, and not doing this almost inevitably leads to an assumption of certainty for the risk factor and its continuity over time that is not supported by scientific evidence.
Secondly, and related to the point above, the judgment leaves open a question that sits uneasily with the Court’s own rationale of using the best available scientific evidence. Even if one were to accept that the Court rightly concluded that a detailed legal and scientific assessment of the situation was possible for 30 years only, it would still be necessary to explain why there is an assumption that the risk factor remains constant during that time. By contrast, it is conceivable that there is decreasing scientific certainty as time passes. Considering the cumulative effects of climate change and current CO2 emissions trends, the risk factor could increase, already in the near future, and well before the end of this decade. This would not necessarily entail a change in the RWE pro rata contributions but would affect the actual risk of significant damage and the ensuing evaluation of imminence, in particular, given the fact that changes in the cryosphere have a considerable lag in time; in other words, the impacts of RWE’s emissions and all other past emissions will show up in the observational data in the decades to come.Footnote 106
Thirdly, the Court considered that even if, as argued by the claimant, a ‘climate factor of 2-4’ would be applied, the probability of an event that would threaten the claimant’s property would still be less than 5%.Footnote 107 This increase in probability was evaluated as insufficient to assume an imminent impairment. The different numbers of percentages and factors that were discussed by the experts and the Court poignantly illustrate the challenges faced by the judiciary in navigating conflicting scientific evidence in legal reasoning.Footnote 108
Yet, the very fact that the Court faces such a difficult task in identifying a normative threshold for imminence calls for judicial caution. It reveals not only a significant legal challenge but also a fundamental disagreement between the experts in making and presenting their scientific assessments. This lack of scientific coherence affects the robustness and the resulting confidence levels of the evidence.Footnote 109 The reduced scientific certainty makes any legal appraisal not only more complicated, but it also increases the risk that a legal evaluation inevitably proceeds with a presumption of accuracy that does not exist.
This means that the legal argument and conclusion set aside the importance of confidence levels, a point that we discuss further below. In the present case, the disputed science nevertheless had a twofold effect: the uncertainties related to the numerical risk and how it may or may not change over time, but, importantly, they also affected the identification of the legally protected right – that is, the assumption that a risk existed only in respect of the claimant’s property and not in respect of the right to health and the right to life. If these higher-ranking legally protected rights had been found to be involved, established case law would have suggested that a 1% risk over 30 years would be sufficient for the threshold of imminence.Footnote 110
Fourthly, the Court follows the methodology applied by its own experts,Footnote 111 while the methodology of the claimant’s expert was not accepted.Footnote 112 It has already been pointed out that the Court is required to present persuasive and reasonable arguments as to why it follows the views of one particular expert.Footnote 113 It is, of course, for the Court to assess the facts in the light of the evidence and there can be different evaluations of what constitutes a persuasive argument when accepting expert evidence. Indeed, the Court’s choices in this regard have been criticized for ‘downplaying the influence of climate change’.Footnote 114 Yet, these critiques equally replace one normative evaluation of the influence of climate change with another.
Fifthly, the important point that underlies this critique is that courts face a particular challenge in applying the normative thresholds and criteria of the law to the scientific evidence, especially in situations where they are confronted with different and complex scientific methodologies and, on that basis, decide which expert evidence they follow. There is a fine line between judicial engagement with the scientific evidence and acknowledging different assessments, confidence levels, probabilities, and data gaps, thus respecting the rationality of scientific evidence and legal reasoning that applies the epistemological rules of law to science. To illustrate this point, the Court stated that scientists decided to apply a safety buffer ‘to be on the safe side’,Footnote 115 and it then proceeded to correct these assumptions given that they favoured the claimant in ‘multiple ways’.Footnote 116 The Court thereby identified an overestimation of the risk based on certain assumptions about the height of the dam and the size and the direction of a potential flood wave. This then led the Court to conclude that the real risk was, by applying a ‘realistic consideration’, significantly lower, which would speak against the risk of an imminent impairment.Footnote 117 In so doing, the Court resorted to a legal rationality that risks replacing epistemological rules of science.Footnote 118 A similar risk emerges where the Court assigns priority to one specific scientific methodology over another, without explaining its choice based on the consideration of varying confidence levels.
4.3. The Baseline for ‘Best Available Science’
The Court chose not to mention, let alone assess, the IPCC evidence on the effects of climate change on glaciers and permafrost, including the IPCC findings that are clearly formulated to apply to glaciers and permafrost regardless of their specific location.Footnote 119 Some of the IPCC evidence has become common knowledge and must therefore be included in the legal assessment, according to section 291 of the Code of Civil Procedure.Footnote 120 As explained above, for facts that are common knowledge no further evidence must be provided.Footnote 121
Yet, even if one were not to follow the view that IPCC reports have become common knowledge, the very existence of such reports is a unique characteristic of the field of climate science, which not only sets it apart from the use of other forms of scientific evidence in legal proceedings but also substantiates the concept of best available science. The IPCC was established by the World Meteorological Organization and the United Nations Environment Programme to ‘provide governments at all levels with scientific information that they can use to develop climate policies’.Footnote 122 It currently has 195 member governments, and ‘thousands of experts from all over the world’ contribute to its reports.Footnote 123
The ICJ echoed the ‘view of participants’ that the reports ‘constitute the best available science on the causes, nature and consequences of climate change’, in its advisory opinion on state obligations in respect of climate change.Footnote 124 The IPCC assessment reports provide a regular update on the state of knowledge following expert assessments, and using IPCC reports in judicial proceedings differs fundamentally from presenting scientific evidence solely through the lens of a particular expert. Instead, the assessments are anchored in a comprehensive scientific review exercise, using expertise from all member countries and regions. In fact, it has been recognized that the ‘authors are required to reach agreement on their assessment of the state of the scientific research, specifying their level of confidence and clearly presenting any contradictory views, gaps in knowledge and uncertainties’.Footnote 125 Therefore, the IPCC’s high-confidence statements are well established and highly robust scientific evidence, representing a baseline for best available science. Newly published papers that are not yet assessed by the IPCC can shed further light on latest scientific developments and can also be relevant in legal assessments. However, deviating from IPCC high-confidence statements will require an equally robust scientific assessment to set aside IPCC baseline evidence in a court of law.
Given the procedures that determine the IPCC assessment cycles, the final reports provide the most authoritative scientific assessments on the causes, nature, and consequences of climate change, and have been used in nationalFootnote 126 and international courts to describe the scientific evidence and context of cases.Footnote 127 Using IPCC reports in courts provides the factual background for the legal reasoning. Any case-specific evidence that is not aligned with, or even contradicts, the findings of the IPCC reports, would require an additional, convincing scientific explanation. Absent any acknowledgement of the IPCC reports, let alone explaining why these reports fail to serve as a factual backdrop for the case-specific evidence in Lliuya, the Court’s legal reasoning inevitably fails to acknowledge the best available science and the established scientific baseline that constitutes the point of departure for the case-specific risk assessment.Footnote 128
The melting of permafrost has been identified by the IPCC as a particularly high risk for humans and ecosystems in the near term, and any additional warming will further amplify permafrost thawing (high confidence).Footnote 129 The IPCC found that even if a 1.5°C temperature limitation is achieved, many low-elevation and small glaciers around the world will lose most of their mass or disappear within decades to centuries (high confidence).Footnote 130 Therefore, and even if the Palcaraju glacier is not a low-elevation glacier, the IPCC evidence that relates to glaciers in a warming world provides the relevant scientific information and also the context for any additional evidence. The expected glacial retreat has direct consequences for the expansion of glacial lakes.Footnote 131
Moreover, a study specifically addressing the Palcaraju glacier, which builds on the IPCC baseline evidence, found that the greatest influence on the increase of a GLOF with regard to the role of the glacier are ‘the steepness of the glacier snout, the growing lake area and the glacier’s retreat’.Footnote 132 It provided specific evidence that the ‘retreat of the Palcaraju glacier, and the resultant expansion of Lake Palcacocha, have increased both the probability and potential magnitude of a GLOF’, placing it in the highest hazard category.Footnote 133 This retreat of the Palcaraju glacier, according to this attribution study, is ‘entirely the result of [anthropogenic] warming’.Footnote 134 Thus, while the concept of best available science can be vague in a non-climate science context, where courts must rely on various experts, this is not the case in climate science where IPCC reports provide an approved baseline that rests on robust assessments of the scientific literature. If a specific risk, impact or region is not covered, experts can provide more case-specific evidence, but they must use the IPCC reports as a baseline to frame specific studies and to calculate risk factors.
4.4. The ‘Actus Contrarius’ Requirement
In cases where there is an imminent impairment, the disturber is under an obligation to ensure that the imminent impairment does not materialize.Footnote 135 This can go beyond the obligation to abstain from further contributions that could increase the risk; it demands an ‘actus contrarius’, a contravening or preventive action.Footnote 136 Ruling on the facts that there was no imminent impairment not only means that the causally relevant GHG emissions can continue, but it also absolved the Court from investigating or explaining the nature of the contravening action that could be conceivable, if the risk became imminent.
Would this require only protective measures, either for the lagoon or the claimant’s property, or could it entail the reduction of GHG emissions and even the removal of past GHG emissions? Considering the strong emphasis placed on the significance of RWE’s contributions, it cannot be excluded that an actus contrarius that actively prevents the property impairment could be required in the future – depending on the evidence provided. It would go beyond the scope of this article to explore the nature of the required measures in depth. Acknowledging this legal twist of nuisance law could either be perceived as a further litigation risk for large emitting companies, or indeed as an incentive to drive a just transition away from fossil fuels.
5. Best Available Science and Causality
The importance of distinguishing between qualitative and quantitative causation resonates from the legal analysis: while the causal quality of the GHG emissions of RWE for the increased risk of property damage was established (qualitative), the exact size of the contribution (that is, the factor by which the risk increased (quantitative)) was difficult to determine. Even scientific studies at times lack the necessary clarity required for this legal consideration. Scientific findings of the IPCC, including on causation and impact attribution, have been used in the academic literature specifically to support the development of the causal analysis at the intersection of science and law;Footnote 137 and nationalFootnote 138 and international courts have used IPCC science and the emerging structural arguments on climate causality.Footnote 139
An important legal finding that emerges from Lliuya is that the Court was convinced that RWE’s emissions were a cause of climate change and that a causal link between RWE’s GHG emissions and the asserted increase in a GLOF risk existed. At the same time, the Court used scientific studies from around 60 years ago instead of IPCC science to establish causation. It was also not in a position – after hearing the expert evidence – to acknowledge and articulate in its reasoning the quantitative uncertainties that surround the assumed risk factor of 1% and its anticipated stability over 30 years.
The discussion of causation in science and law in this section firstly demonstrates how the law must rely on best available scientific evidence to establish a causal link and to attribute changes in risks. It then addresses the limitations of making causal inferences from past scenarios – and even from the present – to inform future risks in a rapidly changing climate.
5.1. Establishing Causality in Climate Science: Multiple Lines of Evidence
Causality in science, including in climate science, is established by using different independent lines of evidence. All of these various lines of evidence provide an answer to the question of whether a factor qualifies as a cause. As every method and data set has limitations, using different lines of evidence is the main method to establish higher levels of scientific certainty in causal inquiries.Footnote 140 Every study that adheres to the epistemological rules of scientific investigation includes an assessment of the level of confidence or certainty in the obtained results, but only taken together, in the context of multiple studies, can high confidence levels and, as such, scientific facts be established.Footnote 141
There is experimental evidence for the warming caused by GHGs that provides the earliest measurable evidence, even before observations started or climate models existed.Footnote 142 In addition, there is a strong physical foundation that also informed these first experiments, that even before warming was detectable, it was already foreseeable from the combustion of fossil fuels.Footnote 143 Indeed, industry documents confirm this awareness long before major impacts became visible.Footnote 144 Different lines of evidence support this finding and, in contrast to RWE’s argument, these lines of evidence, used to build causal chains in the climate context, are not based on climate models, but climate models only form one, often the least important line, of evidence. By contrast, in addition to the fundamental laws of physics (conservation of energy, mass, and momentum) that govern the climate system, and the experimental evidence mentioned above, the most important and constantly improved line of evidence for the causal relationship between GHG emissions and global warming is based on weather observations, including those upon which the calculation of global mean temperature is based.Footnote 145 These observations are analyzed using regression-based statistical methods, also called ‘fingerprinting’ methods, firstly to detect and then to attribute climate change to various external factors.Footnote 146 These methods use both, observations that provide empirical evidence and climate model simulations that provide a combination of theoretical and statistical evidence.Footnote 147 For global mean temperature rise, many different statistical approaches have been used, and they independently confirm that it is impossible to explain the observed warming without including the additional GHG from burning fossil fuels in the statistical models.Footnote 148 Variations of linear regression models address structural model uncertainty and uncertainty arising from internal variability and the influence of chaotic changes in day-to-day weather and other phenomena, such as the El Niño Southern Oscillation, that are inherent in the climate system.Footnote 149
No climate model needs to be employed to explain the strong causal relationship between the burning of fossil fuels and global warming. However, climate models do provide an additional line of evidence, as they are specific mathematical representations of physical laws.Footnote 150 Simple models simulate the planetary energy balance based on the conservation laws of energy and mass, while complex models include the fundamental physical principle of the conservation of momentum and simulate weather directly.Footnote 151 These physics-based models, independent of observations, reproduce the observed warming when using real GHG data. However, because of computational limits, models simplify aspects of the climate system, introducing uncertainties, rendering quantitative assessments – that is, measuring the exact contribution of global warming to a concrete event – on local scales more difficult to accomplish than qualitative assessments and those that relate to a global scale.Footnote 152
A similar physical foundation as for global warming in general exists with regard to the influence of climate change on permafrost in the Peruvian Andes. Even if – as stated by the Court’s experts and followed by the judges – no significant changes in the distribution of permafrost have been observed in the past for the mountain region surrounding the lake Palcacocha, it would be erroneous to conclude on that basis that there will be no or even a reduced likelihood of melting in the future. In fact, the best available science (see Section 4.3) of the IPCC implicates that melting of permafrost has already been attributed to continued and increasing global warming, and that future warming scenarios must be taken into account when considering the trajectory of permafrost thawing, including at the local level.Footnote 153 Assuming a low risk, or describing the risk as abstract-theoretical in the absence of certainty as to the future distribution of local permafrost and the potential danger of large rock falls, limits the ability of the law to trace evidentiary uncertainties comprehensively.
The absence of scientific evidence for a specific type of climate change harm does not entail that the already well-established causal evidence (that indicates a certain trend and demonstrates that the respective consequences will materialize) can be ignored in the risk assessment. This would set aside the best available science that integrates multiple lines of evidence, including the physical laws that underlie the reaction of frost and ice to rising temperatures. Concluding that the risk of harm is low, and will not change over a period of 30 years during which further warming will occur, fails to appraise established IPCC science and the underpinning physical laws.
Presenting scientific studies in any legal setting can be a form of best available science only if the evidence clearly distinguishes between physical laws and the case-specific data sets, using the IPCC assessments as the baseline. Deviations from the baseline of best available science (i.e., case-specific causal claims about a climate-related impact that are not fully aligned with or even contradict the established baseline) must be very carefully evaluated in law. It means that a specific case will be assessed against a causal field of evidence on the impacts of climate change on permafrost. This causal field that describes the impact of global warming on permafrost deserves more legal attention. Even in the absence of case-specific data, the observation of permafrost changes over the last 10,000 years, as mentioned by the Court, cannot be used to predict future permafrost distribution. Because of the delayed response of the cryosphere to warming, the implications of the warming will be measurable only in the coming decades. The case-specific evidence would need to establish why permafrost in this specific situation would react differently.
5.2. Qualitative and Quantitative Causality
So far, it has not been clearly articulated in the legal literature that the law should, as does the science, distinguish between a qualitative and a quantitative finding of causality, in the context of different trajectories of future warming. The well-established evidence that causally links increasing global warming to heightened flood risks in high mountainous regions constitutes a qualitative statement that can be made with high confidence.Footnote 154 By contrast, a quantitative statement would be the determination of the size of the contribution of a specific level of warming to the general or even an individual flood risk – that is, the role of sustenance that can be ascribed to a factor in a set of conditions that may influence the event in combination with other concurrent causes.Footnote 155 This requires additional scientific evidence. The confidence in the evidence of a causal link in most cases will depend on whether the aim is to establish qualitatively whether there is a causal link or whether the size of the contribution needs to be quantitatively determined.Footnote 156 Generally, certainty is higher in qualitative statements and in cases where physical laws underpin the evidence.Footnote 157
To determine the flood risk in a specific location requires scientists to establish evidence by estimating which drivers are important, followed by a qualitative assessment of changes in those drivers.Footnote 158 On that basis, a quantitative assessment can be undertaken. This could entail, for example, analyzing mean changes in temperatures across smaller regions (for instance, countries such as Peru) and climatic regions such as the Andes, in the wider context of examining changes in permafrost and glaciers globally. The starting point for the qualitative and the quantitative assessments would be the IPCC Special Report on the Ocean and Cryosphere in a Changing Climate, which found that permafrost temperatures have increased globally (very high confidence) and that glaciers worldwide have lost mass (very high confidence).Footnote 159 The Report states:
In many high mountain areas, glacier retreat and permafrost thaw are projected to further decrease the stability of slopes, and the number and area of glacier lakes will continue to increase (high confidence). Floods due to glacier lake outburst or rain-on-snow, landslides and snow avalanches, are projected to occur also in new locations or different seasons (high confidence).Footnote 160
These assessments are not only based on models but they also rely on physical laws for both the qualification and the quantification of the causal links, such as that when temperatures increase globally, the likelihood also increases that local temperatures reach levels where ice and snow melting begins.Footnote 161 These physical laws are the starting point for any quantification of the risk, and they must equally be considered in assessing the changes in risks and the determination of normative thresholds such as imminence. There would indeed need to be an additional source of cooling to reverse the physical laws that underpin the causal quality of GHG emissions for glacier and permafrost melting.
Without further assessments, the scientifically valid conclusion in Lliuya would be that the risk can be quantified only within very large margins of uncertainty. Assuming a stable risk factor in the legal reading of the science disregards the physical laws and risks ignoring the best available science. While the legal assessment of confidence levels will depend on the science presented in any concrete case, a low confidence level resulting from expert evidence that differs from the baseline of IPCC reports and the physical laws that underpin climate change will always require very cautious legal treatment.
Moreover, the normative evaluation of the role of permafrost and the Court’s assessment of past changes to determine the relevant time frame for the GLOF risk assessment point to an underlying assumption that the past can be a reliable guide for the future development in any case. However, available observations show that even in the past, the risk factor was not stable and the flood risk has increased further as a result of the increase in the size and water volume of the glacial lake, caused by the melting of the glacier.Footnote 162 It is certain that the melting will continue, as a result of the lagged effect of glacier melt with regard to global mean temperature increase, the expected direct effects on the thawing of permafrost, summer heatwaves destabilising slopes and increasing the risk of rockslides, in combination with other drivers of flood risk from glacier lakes as discussed by the IPCC.Footnote 163 All of these factors are also changing rapidly, and an assessment of how these rapid changes in the short, mid, and long term influence the flood risk within the next 30 years would have required an assessment of changes in these factors. It is not possible to arrive at a conclusion for the quantification of risk without an evaluation of studies that pertain to changes in these drivers.Footnote 164
6. Conclusions
The pioneering judgment of the Hamm Court of Appeal has concluded a long lawsuit and clarified the applicable standards under section 1004 of the German Civil Code in the context of climate change. The proceedings have prompted significant scientific,Footnote 165 legal,Footnote 166 and interdisciplinary research along the way,Footnote 167 and the judgment paves the way for further research. It provides valuable insights into the role of private nuisance law in defining corporate obligations in respect of climate change and demonstrates how causation and normative thresholds, such as the significance of contributions and imminence of first impairment, can be applied under section 1004 of the Civil Code.
As the analysis has revealed, the judges faced significant challenges in navigating the scientific evidence and experts’ views. The decision reaffirms the imperative to identify, present, and evaluate the best available science in legal assessments. In climate litigation, the physical laws of climate change and the IPCC assessment reports form a baseline of best available science, and high-confidence statements can be treated as scientific facts in the courtroom. Scientific certainties, margins of uncertainties, and variations in confidence levels must be clearly articulated in science and assessed accordingly in law. Using the existing best available science as a starting point, the research community is tasked to pursue research that can generate additional case-specific scientific evidence.
An improved legal reading of scientific evidence can emerge from that co-production and improve the likelihood of resolving disputes in a court of law. Most importantly, it will forge the knowledge to pre-emptively design legal frameworks cognizant of both the human contribution to extreme weather events and slow onset events, and the responsibility to act upon that knowledge to avoid harm. A clearer delineation between qualitative and quantitative evidence in the causal reasoning, the appreciation of confidence levels and probabilities, a better understanding of normative and scientific thresholds and the role of physical laws, and improving the reliability of scientific evidence all contribute to this aim.
It should be noted that other courts in Germany will not be bound by the decision of the Court of Appeal. They will certainly be well advised to consider the judicial reasoning and the standards that emerge, but there is no rule of binding precedent. Writing about the emerging accountability of carbon majors, Heede suggested that, in science, progress is built on the work of others.Footnote 168 The same is true for law and legal reasoning. The evolution of legal concepts in the context of climate change requires interdisciplinary research to address climate change and its devastating impacts that are felt and feared especially by vulnerable communities.
Acknowledgements
Many thanks to C.T. Reid, Professor Emeritus, University of Dundee (Scotland), and M. Aselmann, Dr. iur., Judge of the Higher Court of Appeal, Oldenburg (Germany), for their helpful comments to earlier drafts.
Funding statement
Not applicable.
Competing interests
The authors declare none.
Open access
