Intervenants invités :
Pascale Braconnot, LSCE
Credibility of climate projections to explore the climate and anticipate its evolution
Results of climate projections run with state of the art general circulation models are used for a wide range of applications, including climate change impact, mitigation or adaptation strategies. This presentation will review how the climate modelers assess the credibility of these simulations considering the way models are built, natural climate variability, and the representation of the response of the simulated climate to different external drivers that can be natural or induced by human activity. In addition, using recent examples of interactions with the industrial sectors of energy and water, I’ll also discuss the need to properly formulate the questions and vulnerability assessments to make the most appropriate use of these simulations and of their uncertainties. This opens the question of the level of expertise that should be included in the emerging climate services.
Mathias Frisch, University of Maryland
From Climate Models to Public Policy Advice
How do we bridge the gap between climate models and their uncertain predictions and climate policy advice? In this talk I examine the impact that deep uncertainties infecting detailed quantitative predictions of future climate have on different strategies for deciding on a climate policy. In particular I compare and contrast quantitative cost-benefit analyses with more qualitative strategies, as embodied in the 2 degree Celsius goal adopted in the Cancun accord.
Johannes Lenhard, Universität Bielefeld
Uncertainty in Early and Contemporary Modeling
My strategy will be to approach the topic of the conference from two different directions. The first starts back in history with the allegedly first atmospheric scientist and decision theorist, namely Pascal. He introduced not only the issue of risk and framed the first probabilistic argument for dealing with it. Moreover, I will argue, one can read Pascal’s wager as a robustness argument that distinguishes between risk and a deeper sort of uncertainty. Pascal was arguably well aware that arguments about risk as well as about uncertainty suffer from problematic applicability.
The second part of my talk will discuss the methodology of simulation modeling and will focus on tuning and its impact on uncertainty. I will argue that tuning is inevitable when building complex simulation models, and furthermore, that tuning restricts the basis for uncertainty estimation to considering the global model behavior.
From these two lines of argumentation, I will discuss option for dealing with uncertainty.
Hervé Le Treut, IPSL
Title
Abstract
Katie Steele, London School of Economics
International Paretianism: A palatable response to climate change?
International Paretianism (IP) is billed as the positive and feasible response to international problems such as climate change (see, especially, Posner and Weisbach 2010). In the tradition of public goods economics, climate change is depicted as a collective-action problem for nation states, where the current ‘business-as-usual’ arrangement is Pareto inferior to some alternative involving climate-change mitigation. According to IP, a global climate deal should be framed to achieve such a Pareto improvement (making no state worse off). Critics argue, however, that such a deal would not be adequately just, and moreover, to the extent that it supports substantial mitigation, IP involves inconsistent claims about feasibility. This paper contributes to the debate by initially clarifying what would be achieved by an IP response to climate change, under various assumptions about the game theoretic structure of the problem and two possible stances on feasibility. The further consideration is whether an IP deal plausibly has some merit, however minimal, in terms of advancing justice.
Contributions :
Marina Baldissera, University of Pittsburgh
In what sense is uncertainty intrinsic to climate science?
I explore the extent to which claims about structural model uncertainty are intrinsic to climate science. Philosopher’s focus on climate science seems to rest on the assumption that climate science presents unique challenges in dealing with uncertainty. I argue that such an assumption is misguided. I start by analyzing the state of the science and explore the extent to which this state influences claims about epistemic and ontological uncertainty. I continue by drawing analogies in modeling of biological systems and climate science, and argue that there might be a quantitative but not qualitative difference in uncertainty between the two sciences.
Seamus Bradley, Munich Centre for Mathematical Philosophy
Confidence from Robustness: A Cautionary Tale
The standard sources of confirmation for a theory are often unavailable in climate science. Climatologists have therefore turned to alternative sources of confidence, including robustness.
I give an example of a model result that was robust enough for it to be commented on as a source of confidence in the IPCC's Assessment Report 4. Since then, further modelling work has shown that the result is not as robust as previously thought.
I discuss the possible reasons for this spurious convergence of model results and conclude that appeals to robustness in climate models should be viewed with suspicion.
Jay Geyer, University of Colorado at Boulder
Toward a Better Precautionary Principle
Risk-averse policy approaches to climate change under the heading of ‘the precautionary principle’ have come under heavy criticism, though more recent approaches oriented around the Maximin decision procedure fare better than their predecessors. I argue that Maximin-style articulations of the precautionary principle are hopelessly flawed because they fail to recognize that decision-makers facing the challenge of climate change are not in a state of complete ignorance; they have some information about the probabilities of various states obtaining. Once this is acknowledged, counter-examples to Maximin-style approaches abound. I then articulate a risk-averse Maximize Expected Utility procedure as an alternative precautionary approach.
Casey Helgeson, London School of Economics
Richard Bradley, London School of Economics
Brian Hill, HEC Paris
Decision and Climate Change Assessments
The Intergovernmental Panel on Climate Change (IPCC) has adopted official guidelines for communicating uncertainties attached to the scientific findings that appear in their assessment reports. The novel feature of these guidelines is their use of ‘confidence’, a second-order qualification that indicates the level of scientific understanding behind a claim. So far, it is unclear how findings communicated in this format can be used to inform rational decisions. We explore how the IPCC guidelines fit together with a new, confidence-based normative model of decision making (Hill, 2013, ‘Confidence and Decision’, Games and Economic Behavior).
Carlo Martini, University of Helsinki
Measuring risk by subjective indicators
"How can risk and uncertainties be taken into account in decisions about environmental and political issues? Is their ascertainment by numerical methods always helpful?" In this paper I argue that subjective indicators are better suited at measuring risk (and, by extension, uncertainties) in at least those sectors of the climate sciences that are concerned with economic and social aspects of climate change. This is not only because risk is a moving target, and therefore hard to assess with purely rule-based and mechanical methods, but also because in the economic sciences the measurement of risk is at the same time creation of opportunity, and by extension, a potential source of new risk. The reflexivity of risk measurement in economics bears on important sectors of the climate sciences, and it suggests that subjective indicators are better suited at understanding, measuring and controlling risk.
Pascal Maugis, LSCE
Lionel Scotto D'Apollonia, GDR PARCS
Participative workshop: Towards a global citizen expertise integrating complex treatment of uncertainty?
This session proposes an original participatory workshop where you are invited to discuss for providing answer to one of questions on this conference: “How can uncertainties be taken into account in decisions about environmental and political issues?” In a 30’ first part we present heuristic approach of "post-normal science", especially the ways a citizenship climate expertise may emerge that would account for the complexity of uncertainties in comparaison to the modern model of IPCC : “science speaks truth to power”. Then in a 30’ second part, based on techniques of participatory action research (PAR), you are invited to an open debate on these questions, aiming at deepening the communication norms in a context of decision-making.
Kian Mintz-Woo, University of Graz
Moral Uncertainty in Climate Economics
Integrated Assessment Models require addressing uncertainty in moral parameters regarding economic intergenerational distribution. Descriptivists like Hutt (1940); Manne (1995); Nordhaus (1994, 2008) think that, by appealing to current market data, they can avoid making their own value judgments. This view contrasts with prescriptivists (e.g. Arrow et al. 1996; Broome 1994, 2012; Dasgupta 2008; Stern 2007), who have argued that explicitly weighing the moral import of parameter assignments is unavoidable.
I provide alternative nomenclature for the positions. In the case of the Ramsey formula, I introduce new worries for both positions. These come from recent behavioral psychological theories: prospect theory and heuristic theory.
Samsul Mujiharto, Gadjah Mada University
Scientific-Political Corridor for Climate Policy
Climate policy is inseparable from discussion about science and policy interaction. The discussion provides us insights that science and policy are different realms in terms of ends and procedures, but they have to collectively work for establishing climate policy. This collaboration appears when climate policy is understood as much as a matter of working on scientific certainties (seeking for the truth) but also on modeling a range of probabilities and possible options (political language). However, the scientific-political collaboration remains two possible consequences. On the one hand it enables us to have more accurate climate-related policy options. On another hand, the collaboration also has possibility to degrade both scientific credibility and political legitimacy. This paper aims to fix the problem of potential degradation by maintaining climate policy in an intersection issue between science and policy. However, to manage such an intersection area the tiny border between science and policy should be clearly identified. The identification can help decisionmakers to provide arrangement of minimal threshold for scientific credibility and political legitimacy. If done well, they can ensure that the dicision is in-between corridors of scientific credibility and political legitimacy.
Lionel Scotto D'Apollonia, GDR PARCS
The “communicative dissonance” of climate “alarm carriers”
The originality of this communication is to address uncertainties in climate science and its impact on decision-making by the way of science studies. Specifically, analysis of climatologists’ discourses in differentcommunicative situations demonstrates that modern structures of IPCC forcing climate “Alarm carriers” is characterized by a form of "communicative dissonance". This language reveals a tension between ethic of "conviction" (talking about science and uncertainties) and ethic of "responsibility" (talking about the risk they perceive). So this communication opens debate about the evolution of the relationship between science and societies.
Mariam Thalos, University of Utah
Policy decisions in the context of potential future “catastrophe”
Uncertainty divides into two types. Moreover, it has numerous sources. Understanding the types and sources of uncertainty correctly is indispensable in contexts of decision, especially in cases where the stakes are very high and the consequences of decisional error are irreversible. This paper studies the types and sources of error, focusing on situations where we are contemplating the kinds of events known to dynamicists as “catastrophes”—which are essentially unpredictable. It argues that employment of workaday decision formulae like Expected Utility theory are inappropriate to such a context.
Jacques Treiner, Sciences Po Paris
Playing around with climate numbers: a carbon budget approach
The international conference (COP21) which will take place in Paris in december 2015 should actualize the objectives of reduction of green house gas (GHG) emissions. The target is often expressed as :maintaining the increase of the average temperature of the Earth below 2°C with respect to pre-industrial era. This can be achieved provided that the GHG emissions are soon stabilized, then decrease in a way that the total future emissions do not exceed 1000 GtCO2. It is easy to see that both requirements cannot be met, given the present trajectory of fossil fuels consumption and the technical and economical inertia of energy systems. Indeed, mankind is readily engaged on a 3-5°C temperature increase trajectory. It appears also that, if nothing is done to reduce our emissions, one cannot count of the exhaustion of fossil fuels to maintain climate change within manageable limits. However the most emitting scenarios considered in the AR5 report of the IPCC do not seem to match the known reserves of fossil fuels.
