Climate Engineering Liability and Reliability: An Integrated Treatment (CELARIT)

鉴于气候工程（CE）的潜在灾难性副作用，国际责任制度的存在通常被认为是国际上接受的部署的先决条件。然而，有两个基本问题出现：第一，责任制度作为一种手段，分配和分担可能的损害的风险是可取的和可行的;第二，如何处理的事实，即估计的气候效应的CE只能依赖于数值气候模拟，而不是经验数据。虽然CE引起的损害赔偿责任的话题最近引起了越来越多的关注，这些问题还没有得到全面的方式在文献中。此外，如何能够对实施行政长官的影响的竞争模型的稳健性和可靠性做出判断，特别是当这些模型可以在法庭诉讼中作为证据时，这一点尚未得到回答。在此背景下，CELARIT的核心目标是制定一个综合处理国际责任问题和模型可靠性和鲁棒性问题的方法。这种处理将建立在三个支柱上：（1）在前体项目CEIBRAL下进行的气候建模，经济学，哲学和法律方面的主要学科研究的成果;（2）CEIBRAL期间这些观点的跨学科整合的成果;以及（3）气候和CE研究以及组成学科的最新发展。这些支柱共同支持和通报了该项目的特定目标，即跨学科解决在CE责任背景下使用模型作为证据的问题。因此，本项目的工作将：（1）解决法院如何评估和比较竞争模型的问题，以及增加稳健性和可靠性的代价是什么;（2）评估在什么条件下气候模型可以被认为在CE责任审判中提供可接受的和充分的证据;（3）研究如何在计算损害的反事实（一个没有CE，也许没有气候变化的世界）无法观察到，但本身是数值模型的产物的情况下确定损害。最后（4），通过分析模型如何在可靠性有限的情况下用于指导与CE相关的行动，以及科学的政策建议如何充分应对不确定性和无知，将在更广泛的背景下审视这一问题。该项目使CEIBRAL的参与者重新团结起来。然而，CELARIT在CEIBRAL之上迈出了重要的一步，因为它对跨学科的综合治疗进行了方法学上的重新定位：它包括对模型不确定性的局限性和潜力的系统调查，以社会负责的方式管理CE实施。

In light of the potentially catastrophic side-effects of Climate Engineering (CE), the existence of an international liability regime is generally considered to be a prerequisite for its internationally accepted deployment. However, two fundamental questions arise: First, whether a liability regime as a means of allocating and sharing the risks of likely damages is desirable and feasible; and secondly, how to deal with the fact that estimates of the climate effects of CE could only rely on numerical climate modeling, but not on experiential data. Although the topic of liability for CE-induced damages has recently attracted increasing attention, these questions have not yet been addressed in a comprehensive manner in literature. Also, it has not yet been answered how it is possible to arrive at judgments on robustness and reliability of competing models of the impacts of CE implementation, in particular when these may serve as evidence in court proceedings. Against this background, the core objective of CELARIT is the development of an integrated treatment of international liability issues and problems of model reliability and robustness. This treatment will build on three pillars: (1) the outputs of the predominantly disciplinary research in climate modelling, economics, philosophy, and law carried out under the precursor project CEIBRAL; (2) the outputs of the interdisciplinary integration of these perspectives during CEIBRAL; and (3) the most recent developments in the state of the art in climate and CE research and in the constituent disciplines. Jointly, these pillars support and inform the particular objective of the project, namely to tackle, across disciplines, the problem of using models as evidence in the context of liability for CE. The work conduced in this project will thus: (1) address the question how competing models can be assessed and compared by a court, and at what cost additional robustness and reliability come; (2) assess under what conditions a climate model may be considered as providing admissible and sufficient evidence in a CE liability trial; (3) examine how to determine damage in a setting where the counterfactual for computing harm (a world without CE and, perhaps, without climate change) cannot be observed, but is itself a product of a numerical model. Finally (4), the problem will be looked at within a broader context by analyzing how models can be used for guiding CE-related action in spite of their limited reliability, and how scientific policy advice can adequately deal with uncertainty and ignorance. The project reunites the participants of CEIBRAL. However, CELARIT takes an important step above and beyond CEIBRAL in that it undertakes a methodological reorientation towards an integrated treatment across disciplines: It includes a systematic investigation of the limitations and potentials that model uncertainty offers for governing CE implementation in a societally responsible fashion.