气候变化综合评估模型的损失函数研究进展

该研究从综合评估模型(IAM)的模型耦合视角出发,介绍了当前损失函数的研究进展,主要从损失函数的构建方法、损失函数与IAM气候模块和经济模块的耦合以及IAM与气候模式的耦合角度分析了损失函数的耦合功能及其存在的科学问题,探讨了损失函数的改进方向。通过文献梳理发现,损失函数的构建方法上,主要采用专家判断法、元分析法和统计学方法,但各有优缺点;与气候模式的耦合功能上,损失函数多以温升为气候变化因素,降水等气候变化信息无法表达,且由全球尺度的年平均值进行标定,不能体现区域的差异和季节的变化,无法直接描述极端气候事件造成的巨大损失;与经济模块的耦合功能上,基于生产部门的损失函数缺乏间接损失评估功能,缺乏对经济增长的动态影响机制。针对上述IAM中气候变化对经济影响的反馈机制的不足,需重点从细化区域气候变化因素影响和细分经济产业部门两个方向重构损失函数,紧密连接气候模式与经济模块,全面评估气候变化经济损失,并需要从技术上解决损失函数在耦合经济模块与气候模式时出现的时空尺度不匹配问题,最终为IAM与气候模式甚至地球系统模式的耦合提供重要的解决方案。     

Learning from integrated assessment of climate change

The objective of integrated assessment of climate change is to put available knowledge together in order to evaluate what has been learned, policy implications, and research needs. This paper summarizes insights gained from five years of integrated assessment activity at Carnegie Mellon. After an introduction, in Section 2 we ask: who are the climate decision makers? We conclude that they are a diffuse and often divergent group spread all over the world whose decisions are primarily driven by local non-climate considerations. Insights are illustrated with results from the ICAM-2 model. In Section 3 we ask: what is the climate problem? In addition to the conventional answer, we note that in a democracy the problem is whatever voters and their elected representatives think it is. Results from studies of public understanding are reported. Several other specific issues that define the problem, including the treatment of aerosols and alternative indices for comparing greenhouse gases, are discussed. In Section 4 we discuss studies of climate impacts, focusing on coastal zones, the terrestrial biosphere and human health. Particular attention is placed on the roles of adaptation, value change, and technological innovation. In Section 5 selected policy issues are discussed. We conclude by noting that equity has received too little attention in past work. We argue that many conventional tools for policy analysis are not adequate to deal with climate problems. Values that change, and mixed levels of uncertainty, pose particularly important challenges for the future.     

Three dilemmas in the integrated assessment of climate change

Conclusions These three dilemmas embody the hardest, most important, and most enduring problems of doing assessment well. None admits simple, obvious solutions. Each can be managed better or worse for any particular assessment endeavor, but doing better requires clear understanding of the purpose of the endeavor. What ways of combining different pieces of disciplinary knowledge, of making projections, and of pursuing policy relevance are more or less appropriate will differ, depending on whether a project seeks to characterize uncertainties and gaps in knowledge; to advise a particular policy choice; to support dialog among policy actors; or to facilitate inquiry into relevant values or goals. Evaluation of the relative emphasis, the methods, and the process of an assessment can only be done relative to some such purpose.Of course, some pitfalls may be so serious as to thwart any purpose, as Risbey et al.'s discussion of the global modeling movement reminds us. The global models' most obvious pitfalls — inadequate treatment of uncertainty, neglect of economic adjustment, excessive confidence in predictions — have largely been seen and avoided by the current assessment community (though there may be more to be learned even here). But on the subtler questions of how assessment or modeling can contribute most usefully to policy, little progress has been made since the 1970s. Consequently, though assessment has advanced in many ways since then, IA remains at risk of suffering the same fate as the global models: a cycle of early enthusiasm, followed by a reaction of frustration and excessive, undeserved rejection.Current endeavors in IA have made substantial contributions to identifying and prioritizing knowledge needs, less to informing specific policy choice. Further progress cannot be guided by a single canonical view of what assessment should be and do, but will proceed incrementally down multiple paths. Several paths currently appear promising: analytic approaches to better represent multiple actors, diverse preferences, and multiple valued outcomes; better representation and application of uncertainty, including diverse expert opinion; novel methods to link assessment with policy communities; and broader participation in assessment teams and explicit focus on negotiating and elaborating pragmatic, viable critical standards. Risbey et al.'s call to develop institutions for critical reflection, mutual learning, and self-improvement will be crucial in developing and evaluating the progress made down these paths.Morgan and Dowlatabadi's checklist for desiderata of IA is a good starting point for a conversation about assessment standards, to which I would propose a few extensions and elaborations. First, there should be not just multiple assessments, but multiple assessment projects using diverse collections of methods and approaches. Second, assessment projects should explore novel methods for connecting their work with the policy community. Third, the approach should be iterative not just within each project, but across assessment projects and between them and the policy community. Fourth, assessors should not be embarrassed by, or seek to disguise, results that are merely illustrative, non-authoritative, and suggestive; these should be acknowledged as such, and the vigorous questioning and critique that will come, including partisan critique, accepted. Do not seek to avoid criticism by mumbling. An important limit to this checklist approach is suggested, though, by the way various writers have groped to define assessment standards by analogy to other domains, revealing how limited is our understanding of how to evaluate assessment. Risbey et al. refer to connoisseurship, as if assessment is like fine wine; Clark and Majone (1985) refer to artistic criticism, as if assessment is like opera singing. If these analogies are appropriate, then pursuing a single set of critical standards for assessment is at least premature, possibly erroneous. Rather, there should be a diversity of approaches, perhaps so broad that no single set of criteria for excellence could be defined. The pragmatic middle way between the too-limiting application of a single set of standards, and an anarchic refusal to evaluate, will have to be negotiated, defined, and improved incrementally.     

The benefits of climate change mitigation in integrated assessment models: the role of the carbon cycle and climate component

Integrated Assessment Models (IAMs) are an important tool to compare the costs and benefits of different climate policies. Recently, attention has been given to the effect of different discounting methods and damage estimates on the results of IAMs. One aspect to which little attention has been paid is how the representation of the climate system may affect the estimated benefits of mitigation action. In that respect, we analyse several well-known IAMs, including the newest versions of FUND, DICE and PAGE. Given the role of IAMs in integrating information from different disciplines, they should ideally represent both best estimates and the ranges of anticipated climate system and carbon cycle behaviour (as e.g. synthesised in the IPCC Assessment reports). We show that in the longer term, beyond 2100, most IAM parameterisations of the carbon cycle imply lower CO₂ concentrations compared to a model that captures IPCC AR4 knowledge more closely, e.g. the carbon-cycle climate model MAGICC6. With regard to the climate component, some IAMs lead to much lower benefits of mitigation than MAGICC6. The most important reason for the underestimation of the benefits of mitigation is the failure in capturing climate dynamics correctly, which implies this could be a potential development area to focus on.     

The bias of integrated assessment models that ignore climate catastrophes

Climate scientists currently predict there is a small but real possibility that climate change will lead to civilization threatening catastrophic events. Martin Weitzman has used this evidence along with his controversial “Dismal Theorem” to argue that integrated assessment models of climate change cannot be used to determine an optimal price for carbon dioxide. In this paper, I provide additional support for Weitzman’s conclusions by running numerical simulations to estimate risk premiums toward climate catastrophes. Compared to the assumptions found in most integrated assessment models, I incorporate into the model a more realistic range of uncertainty for both climate catastrophes and societal risk aversion. The resulting range of risk premiums indicates that the conclusions drawn from integrated assessment models that do not incorporate the potential for climate catastrophes are too imprecise to support any particular policy recommendation. The analysis of this paper is more straightforward and less technical than Weitzman’s, and therefore the conclusions should be accessible to a wider audience.     

Context and climate change: an integrated assessment for Barrow, Alaska

An intensive approach to Barrow, Alaska’s adaptations to climate change and variability during recent decades suggests reconsideration of the interconnected roles of science, policy, and decision-making structures. First, profound uncertainties are inherent in unique interactions among the many natural and human factors affecting Barrow’s vulnerability. Science cannot significantly reduce these uncertainties through extensive approaches, but intensive approaches can reconstruct and update local trends, clarify the underlying dynamics, and harvest experience for policy purposes. Second, sound policies to reduce Barrow’s vulnerability to coastal erosion and flooding must incorporate these profound uncertainties and the multiple values of the community. Minimizing vulnerability to climate change is only one of the community’s interests, and must compete with other interests for limited time, attention, funds and other resources. Third, the community itself is in the best position to understand its own context, to decide on sound policies, and to take responsibility for those decisions. In short, local context matters in science, policy, and decision-making structures for adaptation to climate change and variability. Overall, cognitive constraints may be the most important human dimension of climate change. Factoring the global problem into more tractable local problems would make the most of our cognitive capacity.     

Polar amplification of climate change in coupled models

The Northern Hemisphere polar amplification of climate change is documented in models taking part in the Coupled Model Intercomparison Project and in the new version of the Community Climate System Model. In particular, the magnitude, spatial distribution, and seasonality of the surface warming in the Arctic is examined and compared among the models. The range of simulated polar warming in the Arctic is from 1.5 to 4.5 times the global mean warming. While ice-albedo feedback is likely to account for much of the polar amplification, the strength of the feedback depends on numerous physical processes and parametrizations which differ considerably among the models. Nonetheless, the mean sea-ice state in the control (or present) climate is found to influence both the magnitude and spatial distribution of the high-latitude warming in the models. In particular, the latitude of the maximum warming is correlated inversely and significantly with sea-ice extent in the control climate. Additionally, models with relatively thin Arctic ice cover in the control climate tend to have higher polar amplification. An intercomparison of model results also shows that increases in poleward ocean heat transport at high latitudes and increases in polar cloud cover are significantly correlated to amplified Arctic warming. This suggests that these changes in the climate state may modify polar amplification. No significant correlation is found between polar amplification and the control climate continental ice and snow cover.     

Limitations of time-slice experiments for predicting regional climate change over South Asia

While time-slice simulations with atmospheric general circulation models (GCMs) have been used for many years to regionalize climate projections and/or assess their uncertainties, there is still no consensus about the method used to prescribe sea surface temperature (SST) in such experiments. In the present study, the response of the Indian summer monsoon to increasing amounts of greenhouse gases and sulfate aerosols is compared between a reference climate scenario and three sets of time-slice experiments, consisting of parallel integrations for present-day and future climates. Different monthly mean SST boundary conditions have been tested in the present-day integrations: raw climatological SST derived from the reference scenario, observed climatological SST, and observed SST with interannual variability. For future climate, the SST forcing has been obtained by superimposing climatological monthly mean SST anomalies derived from the reference scenario onto the present-day SST boundary conditions. None of these sets of time-slice experiments is able to capture accurately the response of the Indian summer monsoon simulated in the transient scenario. This finding suggests that the ocean–atmosphere coupling is a fundamental feature of the climate system. Neglecting the SST feedback and variability at the intraseasonal to interannual time scales has a significant impact on the projected monsoon response to global warming. Adding interannual variability in the prescribed SST boundary conditions does not mitigate the problem, but can on the contrary reinforce the discrepancies between the forced and coupled experiments. The monsoon response is also shown to depend on the simulated control climate, and can therefore be sensitive to the use of observed rather than model-derived SSTs to drive the present-day atmospheric simulation, as well as to any approximation in the prescribed radiative forcing. While such results do not challenge the use of time-slice experiments for assessing uncertainties and understanding mechanisms in transient scenarios, they emphasize the need for high-resolution coupled atmosphere-ocean GCMs for dynamical downscaling, or at least for high-resolution atmospheric GCMs coupled with a slab or a regional ocean model.     

On interpreting hydrological change from regional climate models

Although representation of hydrology is included in all regional climate models (RCMs), the utility of hydrological results from RCMs varies considerably from model to model. Studies to evaluate and compare the hydrological components of a suite of RCMs and their use in assessing hydrological impacts from future climate change were carried out over Europe. This included using different methods to transfer RCM runoff directly to river discharge and coupling different RCMs to offline hydrological models using different methods to transfer the climate change signal between models. The work focused on drainage areas to the Baltic Basin, the Bothnian Bay Basin and the Rhine Basin. A total of 20 anthropogenic climate change scenario simulations from 11 different RCMs were used. One conclusion is that choice of GCM (global climate model) has a larger impact on projected hydrological change than either selection of emissions scenario or RCM used for downscaling.     

On the added value of regional climate modeling in climate change assessment

Regional climate models represent a promising tool to assess the regional dimension of future climate change and are widely used in climate impact research. While the added value of regional climate models has been highlighted with respect to a better representation of land-surface interactions and atmospheric processes, it is still unclear whether radiative heating implies predictability down to the typical scale of a regional climate model. As a quantitative assessment, we apply an optimal statistical filter to compare the coherence between observed and simulated patterns of Mediterranean climate change from a global and a regional climate model. It is found that the regional climate model has indeed an added value in the detection of regional climate change, contrary to former assumptions. The optimal filter may also serve as a weighting factor in multi-model averaging.     

Operationalizing longitudinal approaches to climate change vulnerability assessment

The past decade has seen a proliferation of community-scale climate change vulnerability assessments globally. Much of this work has employed frameworks informed by scholarship in the vulnerability field, which draws upon interviews with community members to identify and characterize climatic risks and adaptive responses. This scholarship has developed a baseline understanding of vulnerability in specific places and industries at particular times. However, given the dynamic nature of vulnerability new methodologies are needed to generate insights on how climate change is experienced and responded to over time. Longitudinal approaches have long been used in sociology and the health sciences to capture the dynamism of human processes, but their penetration into vulnerability research has been limited. In this article, we describe the application of two longitudinal approaches, cohort and trend studies, in climate change vulnerability assessment by analyzing three case studies from the Arctic where the authors applied these approaches. These case studies highlight how longitudinal approaches can be operationalized to capture the dynamism of vulnerability by identifying climate anomalies and trends, and how adaptations develop over time, including insights on themes such as social learning and adaptive pathways.     

Understanding public complacency about climate change: adults’ mental models of climate change violate conservation of matter

Public attitudes about climate change reveal a contradiction. Surveys show most Americans believe climate change poses serious risks but also that reductions in greenhouse gas (GHG) emissions sufficient to stabilize atmospheric GHG concentrations can be deferred until there is greater evidence that climate change is harmful. US policymakers likewise argue it is prudent to wait and see whether climate change will cause substantial economic harm before undertaking policies to reduce emissions. Such wait-and-see policies erroneously presume climate change can be reversed quickly should harm become evident, underestimating substantial delays in the climate’s response to anthropogenic forcing. We report experiments with highly educated adults – graduate students at MIT – showing widespread misunderstanding of the fundamental stock and flow relationships, including mass balance principles, that lead to long response delays. GHG emissions are now about twice the rate of GHG removal from the atmosphere. GHG concentrations will therefore continue to rise even if emissions fall, stabilizing only when emissions equal removal. In contrast, most subjects believe atmospheric GHG concentrations can be stabilized while emissions into the atmosphere continuously exceed the removal of GHGs from it. These beliefs – analogous to arguing a bathtub filled faster than it drains will never overflow – support wait-and-see policies but violate conservation of matter. Low public support for mitigation policies may arise from misconceptions of climate dynamics rather than high discount rates or uncertainty about the impact of climate change. Implications for education and communication between scientists and nonscientists (the public and policymakers) are discussed.     

Role of energy efficiency in climate change mitigation policy for India: assessment of co-benefits and opportunities within an integrated assessment modeling framework

Addressing the challenges of global warming requires interventions on both the energy supply and demand side. With the supply side responses being thoroughly discussed in the literature, our paper focuses on analyzing the role of end use efficiency improvements for Indian climate change mitigation policy and the associated co-benefits, within the integrated assessment modeling framework of Global Change Assessment Model (GCAM). Six scenarios are analyzed here in total- one no climate policy and two climate policy cases, and within each of these one scenario with reference end use energy technology assumptions and another with advance end use energy technology assumptions has been analyzed. The paper has some important insights. Final energy demand and emissions in India are significantly reduced with energy efficiency improvements, and the role of this policy is important especially for the building and transportation sector under both reference and climate policy scenarios. Though energy efficiency policy should be an integral part of climate policy, by itself it is not sufficient for achieving mitigation targets, and a climate policy is necessary for achieving mitigation goals. There are significant co-benefits of energy efficiency improvements. Energy security for India is improved with reduced oil, coal and gas imports. Significant reduction in local pollutant gases is found which is important for local health concerns. Capital investment requirement for Indian electricity generation is reduced, more so for the climate policy scenarios, and finally there are significant savings in terms of reduced abatement cost for meeting climate change mitigation goals.     

A population health model for integrated assessment models

In this paper we present the initial results from a project to develop a population health model so we can extend the scenarios included in the IPCC’s Special Report on Emission Scenarios to include population health status. Our initial hypothesis was that some climatic variable, particularly temperature, would have a significant impact on health outcomes. After experiments – using the Global Burden of Disease (GBD) data on Years of Life Lost (YLL) and Years Lived with disability (YLD) both by WHO region and by five degree latitude band as outcome variables – failed, we settled on life expectancy (LE) as the best measure of health status. We discovered that there is a solid relationship between LE and the GBD data from our first experiments, allowing us to extend the results from the LE model. The LE model used cross section data on LE for 91 countries and included temperature, per capita income, access to clean water and sanitation, literacy, simple medical attention, nutrition, per capita medical expenditure, electricity use per capita, and automobiles per capita as independent variables. While all were individually associated with LE, our model of choice included literacy, access to clean water and sanitation, simple medical attention, an indictor variable for Sub-Saharan Africa and purchasing-power parity per capita income. Note that neither temperature nor calories enter into this model. The fit between life expectancy, as predicted by this model, and actual life expectancy was quite good (R ² =0.90), except for Rwanda, Uganda, and Madagascar; these countries accounted for one half of the unexplained variation in the model. The LE model was then used to develop trajectories of life expectancy in India for the four IPCC SRES storylines, where values for the independent variables were extrapolated based on the story line content. YLL and YLD estimates were created using the current cross relationship of these outcomes to LE. Given the lack of a general role for climate in our LE model, future work is planned to explore how to add detailed climate related impacts, to explore alternative nutritional variables, as well as extend the data set to allow a cross-section time-series approach.     

Uncertainties in the regional climate models simulations of South-Asian summer monsoon and climate change

The uncertainties in the regional climate models (RCMs) are evaluated by analyzing the driving global data of ERA40 reanalysis and ECHAM5 general circulation models, and the downscaled data of two RCMs (RegCM4 and PRECIS) over South-Asia for the present day simulation (1971–2000) of South-Asian summer monsoon. The differences between the observational datasets over South-Asia are also analyzed. The spatial and the quantitative analysis over the selected climatic regions of South-Asia for the mean climate and the inter-annual variability of temperature, precipitation and circulation show that the RCMs have systematic biases which are independent from different driving datasets and seems to come from the physics parameterization of the RCMs. The spatial gradients and topographically-induced structure of climate are generally captured and simulated values are within a few degrees of the observed values. The biases in the RCMs are not consistent with the biases in the driving fields and the models show similar spatial patterns after downscaling different global datasets. The annual cycle of temperature and rainfall is well simulated by the RCMs, however the RCMs are not able to capture the inter-annual variability. ECHAM5 is also downscaled for the future (2071–2100) climate under A1B emission scenario. The climate change signal is consistent between ECHAM5 and RCMs. There is warming over all the regions of South-Asia associated with increasing greenhouse gas concentrations and the increase in summer mean surface air temperature by the end of the century ranges from 2.5 to 5 °C, with maximum warming over north western parts of the domain and 30 % increase in rainfall over north eastern India, Bangladesh and Myanmar.