Integrating Climate Impacts Models. Modeling the Health Risks of Climate Change 4 November 2014 Washington, DC Anthony C. Janetos Boston University

Transcription

1 Integrating Climate Impacts Models Modeling the Health Risks of Climate Change 4 November 2014 Washington, DC Anthony C. Janetos Boston University

2 Prediction is very difficult, especially about the future. Niels Bohr

3 Overview What do we mean by Integrating Climate Impacts? Thinking about health outcomes in a systems perspective Thinking about whether we are just interested in disease or other health burdens as well Different levels of integration What might the future hold?

4 What do we mean by Integrating Climate Impacts? There is a growing number of integrated reports of climate impacts, most of which include some discussion of health outcomes Brief survey of these all of them from a climate change perspective

5 Two Prior National Scientific Assessments

6 Third National Climate Assessment Climate Change Impacts in the United States

7 Human-induced climate change has moved firmly into the present. Dave Martin/AP/Corbis Date Name of Meeting 7

8 Americans are already feeling the effects of increases in some types of extreme weather and sea level rise. Stan Honda/AFP/Getty Images Date Name of Meeting 8

9 Impacts are apparent in every region and in important sectors including health, water, agriculture, energy, and more. Scott Olson/Getty Images Date Name of Meeting 9

10 There are many actions we can take to reduce future climate change and its impacts and to prepare for the impacts we can t avoid. Esperanza Stancioff, UMaine Extension and Maine Sea Grant Dennis Schroeder, NREL Date Name of Meeting 10

11 CLIMATE CHANGE 2014: IMPACTS, ADAPTATION, AND VULNERABILITY

12 IMPACTS CLIMATE Natural Variability Vulnerability SOCIOECONOMIC PROCESSES Socioeconomic Pathways Anthropogenic Climate Change Hazards RISK Adaptation and Mitigation Actions Exposure Governance EMISSIONS and Land-use Change

13 WIDESPREAD OBSERVED IMPACTS A CHANGING WORLD

14 WIDESPREAD OBSERVED IMPACTS A CHANGING WORLD

15

16

17 Very Low Risk-Level Med Very High Present POLAR REGIONS Risks for Ecosystems Risks for Health and Well-Being Unprecedented Challenges, Especially from Rate of Change Near Term ( Long Term 2 C ( C Risk Level with High Adaptation Potential for Additional Adaptation to Reduce Risk Risk Level with Current Adaptation NORTH AMERICA Increased Flood Losses and Impacts EUROPE Increased Risks from Wildfires Heat-Related Human Mortality Damages from River and Coastal Urban Floods Increased Water Restrictions Increased Losses and Impacts from Extreme Heat Events Increased Flood Damage to Infrastructure, Livelihoods, and Settlements Heat-Related Human Mortality ASIA Increased Drought- Related Water and Food Shortage THE OCEAN Reduced Fisheries Catch Potential at Low Latitudes CENTRAL AND r SOUTH AMERICA AFRICA Compounded Stress on Water Resources Increased Mass Coral Bleaching and Mortality Coastal Inundation and Habitat Loss Reduced Water Availability and Increased Flooding and Landslides Reduced Food Production and Quality Reduced Crop Productivity and Livelihood and Food Security Vector- and Water- Borne Diseases SMALL ISLANDS Loss of Livelihoods, Settlements, Infrastructure, Ecosystem Services, and Economic Stability Significant Change in Composition and Structure of Coral Reef Systems Increased Flood Damage to Infrastructure and Settlements AUSTRALASIA Increased Risks to Coastal Infrastructure and Low-Lying Ecosystems Vector-Borne Diseases Risks for Low-Lying Coastal Areas

18 What do we mean by Integrating Climate Impacts? But is this profusion of literature actually sufficient? All conclude that risks to health from multiple pathways will increase All are extremely cautious about actual health outcomes Whether adaptation is included or not is highly variable in the underlying literature

19 So, what do we mean by Integrating Climate Impacts? Suggest that one element of integration that we are mostly missing is whether we can consider health impacts in integrated modeling studies Distinguish three types of integrated modeling studies One way modeling including climate variables as additional drivers in both disease and non-disease pathways for health outcomes One way modeling ability to include adaptation and institutional response so that can come closer to health outcomes, rather than assuming today s health system and institutions Two way modeling including the feedbacks from health outcomes and adaptation on both economic outcomes and on subsequent consequences for both mitigation and future demographics

20 So, what do we mean by Integrating Climate Impacts? Category One way modeling including climate variables as additional drivers in both disease and non-disease pathways for health outcomes One way modeling ability to include adaptation and institutional response so that can come closer to health outcomes, rather than assuming today s health system and institutions Two way modeling including the feedbacks from health outcomes and adaptation on both economic outcomes and on subsequent consequences for both mitigation and future demographics Yes/No?

21 Health outcomes in a systems perspective The Simplest Systems Diagram I Could Think Of Climate Drivers Climate System Technology and Emissions Socioeconomic Drivers Direct Impacts Healthcare Institutions And Access Public Infrastructure Indirect Impacts Zoonoses Food security Water Availability Storms Demographics and Economic Costs/Consequences Outcomes

22 Different levels of integration Any one of those components could itself be expanded into a systems diagram So question is to what degree to we want to think about health impacts in a systems context? Part of an overall climate-demography system? As an impact sector with forcings, responses, human interventions? We have already heard both perspectives

23 Different goals of integration Why are we concerned about integrating impacts information? Better understanding of the impacts themselves, especially when they are the result of multiple factors Ability to generalize from one situation to another, or recognize when this is difficult to do Ability to target interventions to reduce vulnerability or incidence of harm These are not the same thing Designing interventions may or may not demand the same degree of understanding of processes that the other goals do

24 Scaling There are at least three aspects of scaling issues that we should distinguish clearly Spatial scales is it always the case that local information is required? Maybe for some interventions, but clearly not for others Temporal scales dealing with the aftermath of events, spikes in frequency of disease or understanding longer-term trends in both forcings and consequences A well-known challenge for environmental health issues, especially when mechanisms poorly understood Process resolution how much process-level, mechanistic information is desired vs how much can be parameterized in a systems context Always an issue for systems modeling and understanding

25 Do models span this broad Mostly not perspective? Models that come the closest are integrated assessment models

26 Integrated Assessment Models (IAMs) IAMs integrate human and natural Earth system climate science. IAMs capture interactions between complex and highly nonlinear systems. IAMs provide insights that would be otherwise unavailable from disciplinary research. IAMs provide physical science researchers with information about human systems such as GHG emissions, land use and land cover. IAMs provide important, science-based decision support tools. IAMs support national, international, regional, and private-sector decisions. 26

27 Different types of IA models Some integrated assessment models (e.g., DICE, FUND) have focused on cost-benefit analysis. That is, weighing the costs of mitigation against the costs of inaction. These models have very simple representations of the economy, but incorporate all potential feedbacks from the climate to the human system. 27

28 Different types of IA models Other integrated assessment models (e.g., IGSM, GCAM, MESSAGE) have focused on cost-effectiveness analysis. That is, quantifying the transition pathways and costs associated with stabilizing climate at a pre-defined level. These models have more complex representations of different components of the economy (e.g., energy, agriculture), but have largely excluded feedbacks from the climate to the human system. 28

29 Where could we make progress? Integrated modeling community has been doing a lot of experiments with one-way modeling of impacts mostly for physical/biological impacts Impacts are rarely taken all the way through to human health or economic consequences Right now no integrated assessment modeling framework has a feedback loop from climate impacts to demographic change Could be developed at a population level, but will take some thinking and work to do in a sensible way

30 Where could we make progress? A one-way modeling approach, or a more complete systems approach must deal with the problem of what will they use for climate information about the future You can get high-resolution climate models, but not for long simulations Long simulations are inevitably at coarser resolution than many impact scientists desire We have either no information, uncertain information, or in some cases negative information about model skill and reliability The fact that we can make the computers generate numbers is not the same as saying we should therefore believe them There are both statistical and dynamical methods for downscaling climate information, but tracking uncertainty quantitatively can become difficult

31 Where could we make progress? Modeling to understand systems at any level is essentially the same as asking whether we understand the systems well enough to make projections, predictions, or forecasts But we do not need to do this forward modeling to understand how best to target interventions We might do well to turn the problem around, and ask about the degree to which we understand the parts of the system that determine vulnerability and sensitivity Then model the degree to which interventions could have an effect And only then ask the question about whether change in climate makes those interventions easier or harder

32 What might the future hold? Could quite easily see a future in which several forms of integrated science are pursued Integrated systems modeling at a high level for understanding long-term trends and evolution of climate-health linkages over several decades, and in the developing world Systems modeling at more detailed levels to understand particular cases, perhaps to assess the value of particular interventions Empirical research and modeling to understand where sensitivities arise and best methods for reducing vulnerability

33 A Risk Management Problem in Addition to a Scientific Problem Acknowledge uncertainties in science, but manage the risks Focus on what s really important

34 Analysis for a better tomorrow, today. 67 Bay State Road Boston, Mass (617) flickr.com/bupardeecenter