Executive Summary



Climate Change and MortalityBy the Resources and Environment Working Group the International Actuarial AssociationWorking Draft – 14 May 2016Table of ContentsTable of Contents TOC \o "1-3" \h \z \u HYPERLINK \l "_Toc451678945" Executive Summary PAGEREF _Toc451678945 \h 4 HYPERLINK \l "_Toc451678946" 1.Introduction PAGEREF _Toc451678946 \h 5 HYPERLINK \l "_Toc451678947" 1.1Objectives PAGEREF _Toc451678947 \h 5 HYPERLINK \l "_Toc451678948" 1.2Acknowledgments PAGEREF _Toc451678948 \h 5 HYPERLINK \l "_Toc451678949" 1.3Roadmap to the report PAGEREF _Toc451678949 \h 5 HYPERLINK \l "_Toc451678950" 2.Background PAGEREF _Toc451678950 \h 7 HYPERLINK \l "_Toc451678951" 2.1Climate change – types of effects PAGEREF _Toc451678951 \h 7 HYPERLINK \l "_Toc451678952" 2.2Systems thinking PAGEREF _Toc451678952 \h 9 HYPERLINK \l "_Toc451678953" 2.3Climate components PAGEREF _Toc451678953 \h 10 HYPERLINK \l "_Toc451678954" 2.4Regional effects PAGEREF _Toc451678954 \h 11 HYPERLINK \l "_Toc451678955" 2.5Age related effects (possibly later chapter?) PAGEREF _Toc451678955 \h 12 HYPERLINK \l "_Toc451678956" 2.6Relative wealth (possibly later chapter?) PAGEREF _Toc451678956 \h 12 HYPERLINK \l "_Toc451678957" 2.7Positive and negative impacts PAGEREF _Toc451678957 \h 13 HYPERLINK \l "_Toc451678958" 2.8Measuring the impact PAGEREF _Toc451678958 \h 13 HYPERLINK \l "_Toc451678959" 3.Adverse Mortality Consequences PAGEREF _Toc451678959 \h 16 HYPERLINK \l "_Toc451678960" 3.1Background PAGEREF _Toc451678960 \h 16 HYPERLINK \l "_Toc451678961" 3.2Increase in certain diseases PAGEREF _Toc451678961 \h 16 HYPERLINK \l "_Toc451678962" 3.3Drought and resulting famine PAGEREF _Toc451678962 \h 17 HYPERLINK \l "_Toc451678963" 3.4Increased and more frequent severe weather conditions PAGEREF _Toc451678963 \h 19 HYPERLINK \l "_Toc451678964" 3.5Increase in natural disasters PAGEREF _Toc451678964 \h 19 HYPERLINK \l "_Toc451678965" 3.6Coastal flooding and rising sea levels PAGEREF _Toc451678965 \h 20 HYPERLINK \l "_Toc451678966" 3.7Impact of wealth PAGEREF _Toc451678966 \h 20 HYPERLINK \l "_Toc451678967" 3.8Air Pollution PAGEREF _Toc451678967 \h 21 HYPERLINK \l "_Toc451678968" 4.Favorable Mortality Consequences PAGEREF _Toc451678968 \h 23 HYPERLINK \l "_Toc451678969" 4.1Positive impacts of climate change on future mortality PAGEREF _Toc451678969 \h 23 HYPERLINK \l "_Toc451678970" 4.2Direct positive impacts PAGEREF _Toc451678970 \h 23 HYPERLINK \l "_Toc451678971" 4.3Indirect positive impacts PAGEREF _Toc451678971 \h 24 HYPERLINK \l "_Toc451678972" Adaptation measures PAGEREF _Toc451678972 \h 24 HYPERLINK \l "_Toc451678973" Mitigation efforts PAGEREF _Toc451678973 \h 25 HYPERLINK \l "_Toc451678974" 5.Insured / Pension Effects PAGEREF _Toc451678974 \h 28 HYPERLINK \l "_Toc451678975" 5.1Direct effects PAGEREF _Toc451678975 \h 28 HYPERLINK \l "_Toc451678976" 5.2Indirect effects PAGEREF _Toc451678976 \h 28 HYPERLINK \l "_Toc451678977" 5.3Beyond the mortality impact PAGEREF _Toc451678977 \h 29 HYPERLINK \l "_Toc451678978" 6.Quantitative Analysis and Modeling PAGEREF _Toc451678978 \h 30 HYPERLINK \l "_Toc451678979" 6.1Use of scenarios PAGEREF _Toc451678979 \h 31 HYPERLINK \l "_Toc451678980" 6.2Dealing with non-stationary risk PAGEREF _Toc451678980 \h 32 HYPERLINK \l "_Toc451678981" 7.Case studies PAGEREF _Toc451678981 \h 34 HYPERLINK \l "_Toc451678982" 8.Need for further research PAGEREF _Toc451678982 \h 35 HYPERLINK \l "_Toc451678983" 9.References PAGEREF _Toc451678983 \h 36Executive Summary41.Introduction51.1Objectives51.2Acknowledgments51.3Roadmap to the report52.Background72.1Climate change – types of effects72.2Systems thinking92.3Climate components102.4Regional effects112.5Age related effects (possibly later chapter?)122.6Relative wealth (possibly later chapter?)122.7Positive and negative impacts132.8Measuring the impact133.Adverse Mortality Consequences163.1Background163.2Increase in certain diseases163.3Drought and resulting famine173.4Increased and more frequent severe weather conditions193.5Increase in natural disasters193.6Coastal flooding and rising sea levels203.7Impact of wealth203.8Air Pollution214.Favorable Mortality Consequences234.1Positive impacts of climate change on future mortality234.2Direct positive impacts234.3Indirect positive impacts24Adaptation measures24Mitigation efforts255.Insured / Pension Effects285.1Direct effects285.2Indirect effects285.3Beyond the mortality impact295.4Use of scenarios in actuarial work295.5Dealing with non-stationary risk316.Quantitative Analysis and Modeling337.Case studies358.Need for further research369.References37Executive Summary[later]IntroductionFuture climate change represents a long-term peril to our planet. The effects of this change will arise from several factors, reflecting adverse effects on water, air, weather, oceans, and ecosystems, from significant increases in greenhouse gases in the Earth’s atmosphere and oceans, which will consequently result in many other changes in our environment, including increases in temperature, changes in precipitation including heavy rains and longer-lasting droughts, increases in the frequency and intensity of extreme weather events and weather-caused conditions, and rises in sea levels – affecting glacier size and water and air quality. These represent a range of sudden and slowly developing factors.The ultimate effects of these climatic changes include consequences on human health and life – the effect on mortality is the focus of this report. It does not address other effects of climate change, such as adverse consequences to human health (except to the extent directly related to mortality), property damage, and death and reduction in diversity of other living species, (for example the bleaching of coral resulting from warmer and more acidic oceans). This report was prepared by the Resources and Environment Working Group of the International Actuarial Association (IAA). It does not represent an official opinion of the IAA.ObjectivesThe objectives of this report include providing background information and raising the awareness of actuaries and others regarding a key component of the effects of climate change. Secondarily, it is hoped to provide such information to national actuarial associations and policy-makers, as applicable. One conclusion is that further research is clearly needed to discern more particular effects over the short, intermediate and long-term future. Such research will need to be multi-disciplinary in nature, given the complexity of the issues.Acknowledgments[to be filled in later]Roadmap to the reportThis report first provides a brief overview of this issue in chapter 1, Background, covering the range of significant effects of climate change on our environment and human health and life. It is then organized as follows:Chapter 2, Adverse effects on mortality, covering the major sources of additional deaths, including the types of events and conditions precipitating death and attributable diseases likely to be significant in the future. Chapter 3, Favorable effects on mortality, covering the major sources of reduced deaths due to climate change and adaptation/mitigation efforts to reduce climate change. To some extent, these offset the effect of adverse climatic trends. Chapter 4, Quantitative analysis, providing an assessment of the net effect of climate change on mortality. This reflects the wide range of uncertainties associated with the future, especially tail risk associated with possible enormous one-off conditions and exponential or cascading effects of excess greenhouse gas emissions. Both stochastic effects and stress testing approaches may be of use in analysis of alternative possible futures. Chapter 5, Insured effects, addresses the effect of financial programs serviced by actuaries in contrast to the effect of climate change on overall populationChapter 6, Case studies, which study the effects of climate change on {TBD}.Chapter 7, Next steps, which outlines some of the further analysis and research which would provide further insight into likely future effects of the climate of the future. BackgroundClimate change – types of effectsRobert Glasser, the UN Special Representative for Disaster Risk Reduction said “It is clear that weather and climate are implicated in 90 per cent of major disaster events attributed to natural hazards. Droughts, floods, storms and heatwaves have the potential to undermine many developing states’ efforts to eradicate poverty. Climate change is adding to pre-existing levels of risk fueled by exposure and socio-economic vulnerability”.It is the current scientific consensus that climate change has occurred since the industrial revolution began and will continue to affect the atmosphere and oceans over the next several centuries. This has primarily been due to an increase in greenhouse gas emissions. However, international action arising in part as a result of the commitments emanating from the agreement reached at the 2015 United Nations’ Climate Change Conference, COP21 , significant adaptation and mitigation activities will be undertaken that may offset some of the adverse effects of the accumulation of greenhouse gases. Modelling the impacts of climate change on future levels of mortality presents formidable challenges. This is due the interconnected nature of the system being modelled. The Intergovernmental Panel on Climate Change frames it as a combination of: the physical environment, ecology (“natural processes”), and institutions. These three factors combine to impact on the relationship between climate and health. They also interact, each with the other. Thus a change in temperatures (physical) may support the conditions that give rise to an increase in the population of a disease vector (natural process) in a country that may or may not have the institutional capacity to deal with such a change (institutions).A specific example would be higher temperatures leading to the spread of the ranges of deer, mice, and ticks—“the ecologic causal chain that brings Lyme disease to humans” (John Balbus, National Institute of Environmental Health Sciences, USA). This might be described as a slow onset impact of climate change. Similarly, as a result of significant efforts, death from malaria has, according to the World Health Organisation (WHO), experienced a 60 percent reduction in new cases over the last 15 years. This progress may be at risk as temperature increases and rainfall patterns shift over the long-term future, offset by changing per capita wealth as a proxy to education and access to healthcare.Equally, severe storms, such as Hurricane Katrina (U.S.), can overload the functionality of health-care systems, which in turn may have negative consequences in particular for the elderly and those who have chronic medical conditions. According to Molly Brown, of the U.S. National Aeronautics and Space Administration, an analysis conducted in the wake of Superstorm Sandy showed that what most affected human welfare and public health factors such as generators below sea level and unexpected cascading power outages. Hurricane Katrina might be described as a “complex emergency”. 800,000 people were displaced, and therefore impacts were felt in a wide number of cities, not just restricted to New Orleans.Thus, climate impacts are mediated by other physical, environmental and socio-economic factors. Any approach to modelling will therefore of necessity benefit from being multi-disciplinary, involving complex systems thinking. There are many scientific disciplines and tools in addition to climate science necessary to understand the interactions among the three factors and to model the associated health outcomes, including hydrology, geography, ecology, agriculture, sociology, economics, biomedical science, and clinical medicine.Figure SEQ Figure \* ARABIC 1 – impact of climate change on human health Source: George Luber, US Center for Disease Control and PreventionSystems thinkingComplex inter-dependent sets of models will require systems thinking. This is clearly true of the models of slow onset drivers of mortality impacts such as long-term temperature changes or change in rainfall patterns. But it is also relevant in the context of extreme weather-related events. According to Georges Benjamin, of the American Public Health Association “many health systems are designed for just-in-time management for the mean; we cannot afford to build for extremes.” This implies that increased number and / or severity of weather shocks has the potential to critically overload albeit temporarily local or even national health systems. Such shocks will have immediate and medium term consequential impacts which need to be taken into account.Figure SEQ Figure \* ARABIC 2 – a simplified system representationMitigation and adaptationClimate changeDirect impactsFloodHurricaneSea levelIndirect impactsZoonosesFood securityMigrationHealthcare institutionsPublic infrastructureMortality and health outcomesDemographic, social, economic consequencesSocio-economic and political driversTechnology and GHG emissionsClimate systemMitigation and adaptationClimate changeDirect impactsFloodHurricaneSea levelIndirect impactsZoonosesFood securityMigrationHealthcare institutionsPublic infrastructureMortality and health outcomesDemographic, social, economic consequencesSocio-economic and political driversTechnology and GHG emissionsClimate systemAs indicated in REF _Ref450134757 \h \* MERGEFORMAT Figure 2Figure 2, the number of deaths around the globe, among other adverse consequences such as health and property damage, are one of the effects of this process. Greenhouse gas (GHG) emissions contribute to the amount of greenhouse gasses in the atmosphere, which can take a very long period of time to naturally dissipate. These, in addition to other drivers, contribute to climate change. This in turn directly and indirectly, result in deaths, ill health and property damage. Climate componentsFour major components of the effects of climate change and related environment influences will influence the level of future deaths:Temperature. Possibly the most advertised effect of climate change is the effect on average global temperature (usually measured from the average temperature level experienced immediately prior to the industrial revolution), influencing many facets of water, air and land temperature. An increase in extreme heat and possibly a decrease in extremely cold weather may also be increasingly common (although see “extreme events” below for one-off local swings in temperature in either direction). Warmer temperatures can cause more evaporation of water. Precipitation. Droughts in already arid regions will continue, may well spread and could increase in severity. Excessive precipitation can result in flooding. In other areas, water scarcity can adversely affect human health. Changes in precipitation can lead to both more intense individual downpours but also swings into drought conditions. The threat from all this is not just to what people drink but what they eat through agriculture, as the human activity that consumes the most water is agriculture.Extreme events. An increase in both the frequency and severity of extreme natural events are anticipated. These events include hurricanes, tornados, flooding and windstorms. These not only affect mortality, but also cause infrastructure and property damage.Sea levels. A rise in sea levels can affect water quality, cause coastal flooding and insect related diseases. It could impact on food security and may have direct and indirect consequences for mass migration.In addition, a wide range of secondary effects are possible, including deteriorating air quality, excessive amounts of ozone and particulate matter, and food and water insecurity including malnutrition. Secondary effects may also include mass migration / human conflict, and changes in insect population and other disease vectors. In addition, economic damages may also result in increased challenges to human longevity.Regional effectsAlthough it is clear that climate change will happen in the future, many of its characterizations and consequences have been expressed in terms of global averages (e.g., average global temperature rises). Even more dramatic changes will arise on a regional, if not local basis.Figure SEQ Figure \* ARABIC 3 – IPCC projections under two future scenariosThe IPCC (Intergovernmental Panel on Climate Change) gives the above projections in its 5th Assessment Report (Working Group 1, Summary for Policymakers). Broadly speaking RCP 2.6 is a greenhouse gas emissions scenario, whereby very substantial mitigations are rapidly put in place. RCP 8.5 by contrast shows a projection assuming little change in emissions levels. In either case, the point is that there will be substantial regional variation. Some regions will experience wider, more violent temperature and precipitation swings and significant adverse natural events. Of course, everyone is exposed to such risks; nevertheless, everyone will be exposed to different hazards, some of them adverse.It has been often noted that these changes will arise to different extent or even in opposite directions by region and demographic groups. Thus, for example, in the United States, the Northeast and Midwest are expected to experience significant increases in precipitation, while the southwest will receive less precipitation and increased drought.Age related effects (possibly later chapter?)It is also expected that climate change will affect age segments to a different extent. For example, the oldest and youngest may be most at risk to many mortality hazards. The oldest will likely be most affected in the case of weather extremes (particularly heat). Many of the youngest will in experience stunting arising from malnutrition, which can in turn expose those affected to other diseases. Children and infants will also suffer higher rates of mortality from diarrhoeal diseaseRelative wealth (possibly later chapter?)The effects of many climate change effects will especially affect more vulnerable populations, e.g., those with lower income and areas with weaker health infrastructures, many of which have not contributed to the causes of climate change in the first place. This is expected to be particularly true of Africa and South Asia.Positive and negative impactsClimate change will have some positive impacts on human health. There are likely to be reductions in cold-related mortality and morbidity in high-income populations. The most recent assessment report of the IPCC concludes, however, that the impacts on health of more frequent heat extremes greatly outweigh the benefits of fewer cold days, and that the few studies of the large developing country populations in the tropics, point to effects of heat, but not cold, on mortality (Smith et al., 2014). These are discussed in chapter 4. Measuring the impactOf course, not all increased deaths from natural causes can be attributed to climate change. For instance, there have always been climate or weather related deaths, and there will continue to be. As a result, it is important to attempt to assess the marginal effect of climate change from these intermediate causes.As set out above, this is a highly complex projection problem crossing over into many different scientific disciplines, with multiple feedback loops from the socio-economic through to planning and infrastructure, adaptation and mitigation, technology and so on.The most sophisticated attempts to examine impacts of climate change on mortality have not attempted to model all of the above. Rather, they have concentrated on a limited set of specific health models for a range of health outcomes known to be sensitive to climate change.For example, WHO (2014) examined the following climate related factors:heat-related mortality in elderly peoplemortality associated with coastal floodingmortality associated with diarrhoeal disease in children aged under 15 yearsmalaria population at risk and mortalitydengue population at risk and mortalityundernutrition (stunting) and associated mortality.It did not attempt an overall “prediction” of the possible impacts of climate change on health or mortality. A main limitation is the inability of current models to account for major pathways of potential health impact, such as the effects of:major heatwave eventsriver floodingother extreme weather eventswater scarcityconsequential increases in migration or conflictconsequential economic damage.Indeed, we are not aware of any major studies examining the mortality impacts of potential discontinuities in climatic, social or ecological conditions. Nor are there extensive studies of what is referred to as slow-onset events, such as increasing sea height or acidification levelsInstead, the mortality forecasts are based on empirical models of specific key mortality causes examining observed mortality trends in relation to major drivers such as socioeconomic development, education and technology, together with projections of the future trajectories of these drivers on a national scale. This is done with regression equations that quantify the current and historical relationships between mortality and a set of independent variables such as gross domestic product (GDP) per capita, years of education and time (which is assumed to be a proxy for health benefits arising from technological developments). For example, malaria is a modeled as a function which incorporates temperature, precipitation and GDP per capita.Figure SEQ Figure \* ARABIC 4 – Example of a general model examining impacts of climate change on diet-related mortalitySource: Springmann et al. (2016)The projections therefore attempt to look relative or comparative levels of mortality in the future on the footing of “with and without climate change”. Thus, by looking at different future GHG emissions pathways (or “radiative concentration pathways”) and by using different climate models, a range of “before and after” climate scenarios can be generated. These are combined with world population projections and different future economic-growth scenarios. In addition, assumptions are required for future levels of adaptation responses.Existing models and considerations for their development and use are discussed further in chapter 6. Adverse Mortality ConsequencesBackgroundIt has been estimated that some 12.6 million deaths globally are attributable to the environment, about 23% of all deaths and 22% of the disease burdens in Disability-Adjusted Life Years (DALYs). However, although it is expected that climate change will have a significant effect on society in the decades ahead, there is a great deal of uncertainty regarding the nature of these impacts, only a fraction of these deaths can be attributed to climate change. This chapter will cover areas for which additional deaths will arise, while the next chapter areas for which there will be a reduction in deaths. In sum, there is considerable potential for increased deaths from the consequences of climate change. They will certainly have a more adverse effect on vulnerable populations and subpopulations, ranging from those in less developed countries to those who cannot afford to move to less affected areas or who cannot afford adaptation / mitigation tools. The very young and very old are also at more vulnerable, as they may be more sensitive to some of these effects. Increase in certain diseasesThe incidence of certain diseases will certainly be affected and will likely increase as a result of the overall warming of our planet. For instance, Diarrheal disease. Compared with a future without climate change, it has been estimated that about 48,000 deaths annually, especially in children younger than age 15, can be attributed to climate change by the year 2030. Malaria and dengue. These diseases are due to mosquitoes that are sensitive to climate change. Although focused in Africa, they are not limited to those areas. The number at risk of malaria will increase by some 5 per cent or 150 million if temperatures rise 2oC to 3oC higher than pre-industrial levels. There have been significant reductions of deaths from these conditions over the last several decades. However, the WHO estimates that, compared with a future without climate change, by 2030 there will be an additional 60,000 deaths per annum due to malaria alone.Gastrointestinal-tract illnesses and infections due to increases in water temperature. Climate change may be associated with staple food shortages, malnutrition, and food contamination of seafood from chemical contaminants, biotoxins, pathogenic microbes, and of crops by pesticides.Asthma, respiratory and lung diseases, such as reduced lung function, coughing and wheezing. Respiratory allergies and diseases may become more prevalent because of increased human exposure to pollen (due to altered growing seasons), molds (from extreme or more frequent precipitation), air pollution and aerosolized marine toxins (due to increased temperature, coastal runoff, and humidity) and dust (from droughts).Cardiovascular and respiratory diseases due to increased ozone and other air pollutants. Climate change may exacerbate existing cardiovascular disease by increasing heat stress, increasing the effect of airborne particulates, and changing the distribution of zoonotic vectors that cause infectious diseases linked with cardiovascular diseases. Cancer. Many potential direct effects on cancer risk, such as increased duration of exposure and intensity of ultraviolet (UV) radiation.Drought and resulting famineMany areas of the world will experience extreme droughts. This lack of water will seriously affect agriculture and the production of food. Drought and severe lack of rainfall can have severe health ramifications. Two of the impacts are:Food security. Severe decreases in water availability can contribute to severe loss of food and liquid input. The obvious result is significant increases in malnutrition and undernutrition, which in children can lead to stunting. Both immediate deaths and a long-term increase in mortality can be the result. Conflicts and violence. Drought, especially when the result of extreme increases in temperature, have been known to result in conflicts and violence. The World Bank (2016) report expresses the view that the likelihood of future conflicts may increase significantly at or near a 4oC global temperature increase compared to pre-industrial climate, although this has already occurred, although it can be difficult to attribute a conflict to a single cause. “In Syria, a devastating drought beginning in 2006 forced many farmers to abandon their fields and migrate to urban centers. There’s some evidence that the migration fueled the civil war there” says Aaron Wolf, a water management expert at Oregon State University, who frequently visits the Middle East.” The Economist, discussing the Arab Spring noted that “… Middle East and North Africa depend more on imported food than anywhere else. Most Arab countries buy half of what they eat from abroad and between 2007 and 2010, cereal imports to the region rose 13%, to 66m tons. Because they import so much, Arab countries suck in food inflation when world prices rise. In 2007-08, they spiked, with some staple crops doubling in price. In Egypt local food prices rose 37% in 2008-10”. Projecting deaths from possible future conflicts sparked at least in part by mass migration or other pressures arising as a result of changes in climate has not to the best of our knowledge been attempted. Even if that modelling were carried out, there would still be questions of attribution. How many of the deaths caused by warfare in Syria could reasonably be said to have been the initial four year drought which may have created in part the conditions in which warfare became inevitable? And, a further layer of questioning would ask, “how much of that drought was attributable to (anthropomorphic) climate change”? The same questions of attribution can then be asked of those who have died fleeing the conflict, or those that have died because of excessive strain on medical facilities as a consequence of large influxes of people. The difficulty of such questions notwithstanding, it does appear likely that increased pressure, whether caused by drought or sea level rises, will give rise to increased mortality through conflict.In the WHO report of 2014, one of the most substantial health effects of climate change in 2050 was undernutrition (an additional 85 000 deaths per annum on the central economic scenario). However, an article in the Lancet published March 2016 went on to examine not just the impact of climate on calories consumed / available, but also the composition of future diet. This compositional effect is said to have greater impact than all the climate factors considered by the WHO combined.The model developed by the team in the Lancet report projected that by 2050, climate change will be associated with 529,000 annual climate-related deaths worldwide (95% CI 314,000–736,000), representing a 28% (95% CI 26–33) reduction in the number of deaths that would be avoided because of changes in dietary and weight-related risk factors between 2010 and 2050.Twice as many climate-related deaths were associated with reductions in fruit and vegetable consumption than with climate-related increases in the prevalence of malnutrition. Most of the climate-related deaths were projected to occur in South and East Asia. This model was based on broadly comparable greenhouse gas emissions and future economic growth projections.The Global Burden of Disease study reported that in 2010, the greatest number of deaths, worldwide and in most regions including developing countries, was attributable to dietary risk factors associated with imbalanced diets, such as those low in fruits and vegetables and high in red and processed meat. This represents a shift, as communicable disease (often associated with malnourishment, poverty and poor sanitation) used to be the dominant cause.Increased and more frequent severe weather conditionsIncreasing average global temperatures will likely increase the frequency and severity of extreme heat conditions and may moderate the frequency of cold weather incidences. Extreme heat surges can especially affect the elderly and most vulnerable. Deaths will increase from heat stress (for example in 2003 during the European heat wave, there were an additional 70,000 deaths). There has been a great deal of study of heat waves – there can be many deaths, especially of the elderly. However, the use of air conditioning can reduce these effects. But it is important to keep in mind that significant increases in air conditioning can increase electricity demand which in turn can contribute to more greenhouse gas emissions, which will then exacerbate the underlying climate change cycle.On its central or base case economic assumptions, the WHO projected 38,000 extra deaths per annum by 2030 as a result of heat. This figure relates to over 65 year olds only, and assumes a fairly high level (50%) of adaptation. With no adaptation this would have been 92,000 annually. These figures rise to a central estimate of around 100,000 extra deaths per annum by 2050 (with 50% adaptation) or over 250,000 with no adaptation. The increases are not distributed evenly: impacts are greatest in the South, East and Southeast Asia regions.Increase in natural disastersExtreme events may result in both direct deaths and consequential environmental degradation. It is expected that there will be a significant increase in natural disasters, including the number of them and their intensity. Examples of sudden events, include hurricanes, tornadoes, cyclones, wild fires and storms. Extremely dry conditions can result in increased wild fires, such as the one that forced abandonment of In addition, there will likely be an increase in intensity of these natural disasters. For example, although not agreed by all, the el Ni?o of 2014-16 was the most intense on record – if due to climate change, certain of the deaths caused by this warming of the Pacific Ocean could be attributed to climate change.Increased refugees, destabilization of communities, reduced access to life-support systems and resulting conflicts will all result in increases in deaths.Coastal flooding and rising sea levelsWater-related deaths will arise from two sources: specific disasters, such as coastal flooding and slow-onset conditions, such as rising ocean levels. If known far enough in advance, and if economically possible, there may be migration away from exposed areas (such as low-lying areas next to the sea or from islands, such as those in the Louisiana delta area in the United States in 2016). Assuming sufficient warning, immediate deaths may be minimal, although consequential effects of crowded emigration areas (e.g., water-borne and infectious diseases and violence) can also impact on mortality. If in the process medical facilities and resources are destroyed, there will be likely be resulting slower delivery of health services (as well as adverse effects on mental health), which will inevitably result in increased mortality. In extreme situations, the effects of water-related events could result in migration of masses of people, with the consequential increase in disease and conflict.Impact of wealthClimate change will affect the air, water, food and shelter in many areas of the world. The most vulnerable may not be able to afford the adaptation and mitigation tools that may be available to avoid these effects. For example, there are over half a million premature deaths per year due to air pollution in India, although a switch to cleaner energy and away from coal is likely to reduce this horrific effect. Food and water-borne infectious diseases are also likely to increase – technology will be used to fight the sources of these diseases, although it is unlikely that they will be able to eliminate them.It has long been understood by actuaries that wealth and income are closely correlated to increased longevity. It is therefore unsurprising that the WHO (2014) noted “(Climate change) impacts are greatest under a low economic growth scenario because of higher rates of mortality projected in low- and middle-income countries”.However, they go on to say that even under optimistic scenarios of future socioeconomic development and with adaptation measures assumed to be effective, climate change is still projected to have substantial adverse impacts on future mortality.They conclude “This indicates that avoiding climate-sensitive health risks is an additional reason to mitigate climate change, alongside the immediate health benefits that are expected to accrue from measures to reduce climate pollutants, for example through lower levels of particulate air pollution. It also supports the case for strengthening programmes to address health risks including undernutrition, diarrhoea, vector-borne disease and heat extremes, and for including consideration of climate variability and change within programme design. The strong effect of socioeconomic development on the projections of future risks emphasizes the need to ensure that economic growth, climate policies and health programmes particularly benefit the poorest and most vulnerable populations”.The WHO work indicate that the most substantial impacts of climate change on health are projected to be caused by undernutrition and infectious diseases (diarrhoeal disease and malaria). In 2030, sub-Saharan Africa is projected to have the greatest burden of mortality impacts attributable to climate change. By 2050, South Asia is projected to be the region most affected by the health effects of climate. However, this work only focused on a number of specific drivers of mortality and did not look at a number of major factors such as impacts on economic development, conflict, mass migration, river flooding, or increased incidence of fast onset (extreme) events such as major heatwaves or hurricanes.Air PollutionIn addition to and in part as a result of climate change, a large number of premature deaths are caused by air pollution. Since a large part of the contributing causes of climate change is also involved with air pollution, it is worth mentioning here some background of this risk, taken from Lelieveld (2016). This study that used a global atmospheric chemistry model, assessed the effect of seven emission sources, including ozone and fine particulate matter (diameter smaller than 2.5 micrometres) on global premature mortality.Lelieveld estimated that premature deaths from outdoor air pollution will increase from 3.3 million per annum in 2010 to 6.6 million per annum by 2050, to a large extent due to increases in population and economic activity. This is in addition to estimates of 3.54 million premature deaths per annum from indoor pollution (e.g., from residential heating and ovens). Immediate causes of death that result from air pollution include chronic obstructive pulmonary disease (COPD), acute lower respiratory disease, cerebrovascular disease, ischaemic heart disease and lung cancer. Although the relative importance of these causes differ significantly by country and region, the top seven sources of death globally are: residential and commercial (31%), agricultural (20%), natural sources such as desert sand (18%), power generation (14%), industry (7%), land transportation (5%), and biomass burning (5%). Premature deaths from outdoor air pollution are heavily concentrated in Asia, with China (about 41%) and India (about 20%), Pakistan (about 4%), Bangladesh (3%) and Nigeria (3%). This number of premature deaths is quite large, but the sources of premature deaths from both outdoor and indoor air pollution points to a number of sources of possible mitigation, including reduced use of coal, more efficient agricultural practices and more fuel-efficient cars/trucks. It appears that although air pollution is decreasing in many more developed cities, it is moving in the wrong direction in developing cities, including in countries such as China, India and Iran. Favorable Mortality ConsequencesPositive impacts of climate change on future mortality While the most likely overall impact of climate change is a worsening of mortality outcomes, there are a number of positive outcomes that may arise either directly from future changes in temperatures and precipitation but also indirectly resulting from the mitigation and adaptation efforts that governments and other stakeholders will take in response to the challenges ahead. As with the impacts of climate change, these trends will significantly vary by region, social class and income. However, despite these possible positive outcomes, overall the likely health effects of climate change are likely to be negativeThis section considers likely positive outcomes. Firstly, it will consider direct positive impacts solely resulting from climate change ahead. Secondly, it will look at the indirect impacts referred to above.Direct positive impactsThis section analyses possible impacts by type.Warmer winter temperatures.While increasing the number of heatwaves, average winter temperatures will also increase with likely less cold related deaths. However, there will still be extreme cold events with some debate regarding whether these will increase, remain stable or decrease.Better agricultural outcomes in certain regions due to temperature. The migration of warm weather will make it possible to grow certain crops in some regions where it was previously difficult or impossible. An anecdotal example – although not a food staple – would be the increasing areas where grapes can now be grown. On its own this may locally improve health outcomes but needs to be considered in conjunction with the increasing number of extreme events including more variable precipitation patterns. This will therefore be offset to some extent by such factors as more extreme events, the loss of agricultural land to sea level rises (particularly in Southeast Asia) and increased drought. It is estimated that climate variation explains a round one third of global crop yield variabilityCO2 Fertilisation effectIn addition to the temperature effect above, there is also a “CO2 fertilisation effect”, which increases carbohydrate production in plants with improved growth and yield as CO2 levels rise.While the effect has been known about for a number of years, it has recently been highlighted in a paper called “Greening of the Earth and its Drivers" in the journal Nature Climate Change. This showed significant “greening” or increased leaf production data from NASA and NOAA satellites over the past 33 years. Green leaves produce sugars using energy in the sunlight to mix carbon dioxide (CO2) drawn in from the air with water and nutrients pumped in from the ground. These sugars are the source of food, fibre and fuel for life on Earth. More sugars are produced when there is more CO2 in the air – this is called CO2 fertilization.However, some studies have shown that plants acclimatize, or adjust, to rising CO2 concentration and the fertilization effect diminishes over time," according to the report’s co-author Dr. Philippe Ciais, Associate Director of the Laboratory of Climate and Environmental Sciences, Gif-suvYvette, France and contributing lead author of the Carbon chapter for the IPCC Assessment Report 5.Indirect positive impactsThis section looks at both mitigating and adaptation efforts whose effects will likely be to reduce mortality.Adaptation measuresAdaptation is essentially the use of (sometimes new) technology to reduce the impact of climate events. Examples include buildings / infrastructure that can withstand greater levels of sudden events such as flood and storms; or which can be put back to use more rapidly following such an event (e.g., “wet-proofing” buildings in flood prone areas). Increased access to air conditioning in hot temperatures is a further example. Increased use of air conditioning has reduced the impact of hotter weather and extreme heat events and therefore the effect on mortality of climate change in hot areas of the world. However, this also has an impact on the climate itself. In addition, living in unaired air-conditioned buildings is thought to be connected to the increase in the number of allergies over recent decades. Thus adaptation measures may themselves have side effects.While migration away from the most badly affected areas may be possible for some, those who are most vulnerable often do not have the resources to move or may lose much of their wealth when migrating. Therefore whether there is a net positive impact on mortality is debatable.Mitigation effortsMitigation in the context of climate change means reducing emissions. Measures taken to reduce climate change emissions are the most likely area where actions may reduce mortality. Particularly important are those leading to a reduction in either household air pollution (which causes some 4.3 million deaths per year) and ambient air pollution which causes around 3.7 million deaths per year. These mitigation efforts have arguably another positive impact – in contributing towards a more efficient use of limited resources, the increase in prices of such resources resulting from their rarefaction - which in general would hit the least well of most - are mitigated.Transport. Increasing car use (and ownership) and a move away from public transport is a significant contributor to climate change .Therefore reduced use of cars, favouring public transport, encouraging car sharing, cycle use and walking (e.g., Nashville which is putting back pavements initially removed to create more room for car use ) are all examples of effective measures that can be taken. A number of positive impacts are likely to arise from such measures. Firstly, a reduction in air pollution directly improves population health. Secondly, increasing physical activity leads to direct health benefits. Finally, mental well-being is improved; indeed it is widely accepted that there is a direct positive correlation between exercise and mental well-being (Helliwell, Layard and Sachs, 2013). In addition, the switch to low impact transport options leads to fewer road traffic accidents and injuries freeing up health resources for other uses. The switch from petrol to electric powered cars can also have positive impacts depending on how electricity is generated but the ‘grey’ energy used to construct the vehicle also needs to be taken into account. Energy. Reduced fossil fuel use and a move to renewable energy and more energy efficiency measures is a large positive contributor to population health. The mining and burning of coal causes contributes to air pollution which has significant health impacts including respiratory infections, cardio vascular diseases, asthma and cancer with between 20% and 40% due to environmental reasons.Solar home systems in BangladeshSince 2003, the installation of solar home systems (“SHS”) has been a priority of the Government of Bangladesh and contributes to a number of objectives including meeting renewable energy targets and poverty reduction. By 2013, nearly 2 million households had been equipped. The project has:Reduced the number of households without energy, a significant factor in reducing poverty.Partially replaced the use of less powerful, polluting and climate-change contributing kerosene lamps. SHS households consume less than 1 litre of kerosene per month, compared to almost 3 litres per month by households without SHS. There are also direct health benefits; SHS adoption reduces respiratory disease of women by 1.2 per cent.Reduced the time of fuel collection (traditionally carried out by women and children).Encouraged a number of small scale projects (e.g. households sell mobile phone charging facilities).Increased the hours during which children can study for school and access to information (e.g. health related) obtained via internet, TV or radio.Has created an estimated 100,000 directly related jobs.Resulted in an increase in food consumption (with positive health impacts).The World Bank estimates that the accrued benefits of a solar unit exceed its cost by 210 per cent.Source: Samad et al. (2013).Solar home systems in BangladeshSince 2003, the installation of solar home systems (“SHS”) has been a priority of the Government of Bangladesh and contributes to a number of objectives including meeting renewable energy targets and poverty reduction. By 2013, nearly 2 million households had been equipped. The project has:Reduced the number of households without energy, a significant factor in reducing poverty.Partially replaced the use of less powerful, polluting and climate-change contributing kerosene lamps. SHS households consume less than 1 litre of kerosene per month, compared to almost 3 litres per month by households without SHS. There are also direct health benefits; SHS adoption reduces respiratory disease of women by 1.2 per cent.Reduced the time of fuel collection (traditionally carried out by women and children).Encouraged a number of small scale projects (e.g. households sell mobile phone charging facilities).Increased the hours during which children can study for school and access to information (e.g. health related) obtained via internet, TV or radio.Has created an estimated 100,000 directly related jobs.Resulted in an increase in food consumption (with positive health impacts).The World Bank estimates that the accrued benefits of a solar unit exceed its cost by 210 per cent.Source: Samad et al. (2013).Reducing fossil fuel use has another potentially positive impact on population health – redirecting fossil fuel subsidies to financing social protection. Subsidies for fossil fuels are estimated at approximately USD 500 to USD 600 billion annually. Contrary to public perception, the level of fossil fuel subsidies are significantly higher than those for renewable energy which amount to around USD 100 billion annually (IEA, 2013). Fossil fuel subsidies are also regressive; the majority of the benefit of such subsidies accrues to the better off with the wealthiest 20 per cent of the population enjoying some 43 per cent of the benefit from fossil fuel subsidies, while the poorest quintile gets only 7 per cent (Arze del Granado, Coady, and Gillingham, 2010). Removing such subsidies and replacing them with targeted social programmes, redistributive measures improve and the correct incentives to reduce fossil fuel consumption are put in place. The IMF (2013) notes that if negative externalities from energy consumption are factored in, subsidies actually cost around USD 1.9 trillion annually – some 8 per cent of government revenues worldwide – which if reformed, could provide much needed financing for social security extension. A World Bank (2012) study indicates that those most in favour of retaining such subsidies are indeed those who most benefit – the rich – and that with detailed explanation of the merits and design of a social programme the worse off are generally open to the change (e.g. in Morocco).Agriculture and diet. Agriculture is a key contributor to climate change. The trend to more meat eating is particularly harmful to the environment and produces significant amounts of climate change gases. At the same time, the increase in consumption of sugar, dairy products and meat has provoked a number of important and expensive health effects (e.g. cancer, diabetes and other obesity related diseases) which put pressure on medical systems and leads to significant losses in productivity. Healthier nutrition initiatives e.g., reduction in eating red meat , soda taxes and promoting physical activity are all effective ways to improve population health whilst at the same time, likely to have related positive impacts on climate change. The overall impacts of mitigation measures on mortality and population health are difficult to quantify as they will depend on the interaction of a number of factors. However, where measures have been taken, results have been positive. There is particular potential for those less well-off who, without intervention measures, are likely to be worse hit by climate change.Insured / Pension EffectsDirect effectsIn general, it appears likely that climate change will have a larger direct mortality effect on vulnerable populations, whether in developed or developing countries. Since the population who are more likely be covered by life insurance or pension programs are more likely to be among the less-vulnerable, those participating in these programs may be less affected.Nevertheless, there may be exceptions. For instance, although the affluent may live, on average, in buildings with better storm-protection than those who are less-well off, they may still live in a highly concentrated area. If, for instance, a hurricane or tornado hits that area, there may be a higher number of deaths in that concentrated area than otherwise. In some ocean-front property areas, the better-off may be able to afford to live there, thus being more exposed to risks related to rising sea levels, for example as a result of increasingly severe storm surges or slow onset due to melting glaciers. For vulnerable populations living in areas with increased risk of mosquito infestation or drought, there may still be actuarial issues despite the lack of traditional pension / life insurance penetration. In Africa, for example, there has been a rapid rise in mobile phone usage for day-to-day transactions. This in turn has driven an explosion in microinsurance, often on a “loyalty” basis as part of a phone package. For example, it is said that EcoLife Zimbabwe reached 20% of the adult population within 7 months on a loyalty model. The spread of mobile insurance via mobile phone (including life, accident, hospital cover as well as funeral plans) is not limited to Africa. Telenor Talkshawk Pakistan / MicroEnsure reached 400,000 in two months also with a loyalty scheme. Other distribution models also have proved successful, whether “freemium” distribution (also via mobile) or through other “traditional” microinsurance routes. Millions of people now have coverage, including some of the poorest and most vulnerable. Clearly, an increasing number of deaths attributable to climate change (or more properly a drag on mortality improvements that otherwise might have been anticipated) will have actuarial ramifications in these markets.Indirect effectsMortality impacts indirectly related to climate change include, for example, deaths caused by a sustained surge in demand beyond the capacity of a healthcare system to cope. As discussed in Section REF _Ref450924204 \r \h 2.1, in the U.S. Hurricane Katrina and caused massive displacement of people and consequential overloading of healthcare services which were not designed to carry significant amounts of spare capacity.As with the direct effects, it seems probable that the more vulnerable (poorer) sections of society will be most vulnerable, which implies that South and Southeast Asia and large parts of Africa will feel the indirect mortality consequences most severely. However, Katrina and Sandy serve as powerful reminders that these mortality effects may not be limited to developing economies.Beyond the mortality impactIn discussing the potential impact of climate change related effects on pension and insurance programs, we have limited our remarks above to mortality effects. However, if the projections are borne out, with the potential for significant discontinuities to arise in the social and economic fabric of countries across the world, there will of course be other and possibly more profound impacts arising in relation to investment strategies of financial institutions such as insurance and retirement programs, and indeed the political backdrop within which these programs operate. Discussion of these effects is outside the scope of this paper.Quantitative Analysis and ModelingTo understand the direct effect of climate change induced temperature changes we need: 1) a model of how changes in temperature affects the mortality rates experienced by a particular population, and; 2) a time series model of the expected changes in average daily temperature for a particular location. A time series model is needed because daily mortality rates might be expected to depend on the temperature experienced by people in the preceding days, not just on the day on which mortality is measured Models of the direct temperature effects attempt to replicate the short-term effect of changes in weather-related temperature on a population (for example that of a large city) by modelling the changes in mortality associated with changes in the average daily temperature. Studies taking this approach have found that the curve of mortality versus temperature has a “J” shape with a minimum at a particular temperature (the “optimum” temperature). The rate at which mortality increases as temperature increases above the optimum is typically greater than the rate of increase as temperature falls below the optimum. A Key finding is that the optimum temperature varies with the location studied, but it is above the average ambient temperature in most places ADDIN ZOTERO_ITEM CSL_CITATION {"citationID":"irfdYJYC","properties":{"formattedCitation":"(Gasparrini et al. 2015)","plainCitation":"(Gasparrini et al. 2015)"},"citationItems":[{"id":34,"uris":[""],"uri":[""],"itemData":{"id":34,"type":"article-journal","title":"Mortality risk attributable to high and low ambient temperature: a multicountry observational study","container-title":"The Lancet","page":"369-375","volume":"386","issue":"9991","source":"CrossRef","URL":"","DOI":"10.1016/S0140-6736(14)62114-0","ISSN":"01406736","shortTitle":"Mortality risk attributable to high and low ambient temperature","language":"en","author":[{"family":"Gasparrini","given":"Antonio"},{"family":"Guo","given":"Yuming"},{"family":"Hashizume","given":"Masahiro"},{"family":"Lavigne","given":"Eric"},{"family":"Zanobetti","given":"Antonella"},{"family":"Schwartz","given":"Joel"},{"family":"Tobias","given":"Aurelio"},{"family":"Tong","given":"Shilu"},{"family":"Rockl?v","given":"Joacim"},{"family":"Forsberg","given":"Bertil"},{"family":"Leone","given":"Michela"},{"family":"De Sario","given":"Manuela"},{"family":"Bell","given":"Michelle L"},{"family":"Guo","given":"Yue-Liang Leon"},{"family":"Wu","given":"Chang-fu"},{"family":"Kan","given":"Haidong"},{"family":"Yi","given":"Seung-Muk"},{"family":"Sousa Zanotti Stagliorio Coelho","given":"Micheline","non-dropping-particle":"de"},{"family":"Saldiva","given":"Paulo Hilario Nascimento"},{"family":"Honda","given":"Yasushi"},{"family":"Kim","given":"Ho"},{"family":"Armstrong","given":"Ben"}],"issued":{"date-parts":[["2015",7]]},"accessed":{"date-parts":[["2016",2,23]]}}}],"schema":""} (Gasparrini et al. 2015)Temperature effects on mortality resulting from climate change were reviewed in the IPCC Fourth Annual Assessment Report ADDIN ZOTERO_ITEM CSL_CITATION {"citationID":"AHAbV6YL","properties":{"formattedCitation":"{\\rtf (\\uc0\\u8220{}ar4-wg2-chapter8.pdf\\uc0\\u8221{} 2016)}","plainCitation":"(“ar4-wg2-chapter8.pdf” 2016)"},"citationItems":[{"id":48,"uris":[""],"uri":[""],"itemData":{"id":48,"type":"article","title":"ar4-wg2-chapter8.pdf","URL":"","accessed":{"date-parts":[["2016",4,27]]}}}],"schema":""} (“ar4-wg2-chapter8.pdf” 2016), and more recently in WHO (2016).A large scale study of the mortality risk attributable to high and low ambient temperatures was published in the Lancet in 2015 ADDIN ZOTERO_ITEM CSL_CITATION {"citationID":"rL2iXz9w","properties":{"formattedCitation":"(Gasparrini et al. 2015)","plainCitation":"(Gasparrini et al. 2015)"},"citationItems":[{"id":34,"uris":[""],"uri":[""],"itemData":{"id":34,"type":"article-journal","title":"Mortality risk attributable to high and low ambient temperature: a multicountry observational study","container-title":"The Lancet","page":"369-375","volume":"386","issue":"9991","source":"CrossRef","URL":"","DOI":"10.1016/S0140-6736(14)62114-0","ISSN":"01406736","shortTitle":"Mortality risk attributable to high and low ambient temperature","language":"en","author":[{"family":"Gasparrini","given":"Antonio"},{"family":"Guo","given":"Yuming"},{"family":"Hashizume","given":"Masahiro"},{"family":"Lavigne","given":"Eric"},{"family":"Zanobetti","given":"Antonella"},{"family":"Schwartz","given":"Joel"},{"family":"Tobias","given":"Aurelio"},{"family":"Tong","given":"Shilu"},{"family":"Rockl?v","given":"Joacim"},{"family":"Forsberg","given":"Bertil"},{"family":"Leone","given":"Michela"},{"family":"De Sario","given":"Manuela"},{"family":"Bell","given":"Michelle L"},{"family":"Guo","given":"Yue-Liang Leon"},{"family":"Wu","given":"Chang-fu"},{"family":"Kan","given":"Haidong"},{"family":"Yi","given":"Seung-Muk"},{"family":"Sousa Zanotti Stagliorio Coelho","given":"Micheline","non-dropping-particle":"de"},{"family":"Saldiva","given":"Paulo Hilario Nascimento"},{"family":"Honda","given":"Yasushi"},{"family":"Kim","given":"Ho"},{"family":"Armstrong","given":"Ben"}],"issued":{"date-parts":[["2015",7]]},"accessed":{"date-parts":[["2016",2,23]]}}}],"schema":""} (Gasparrini et al. 2015) and attracted significant comment ADDIN ZOTERO_ITEM CSL_CITATION {"citationID":"ryMTVTlD","properties":{"formattedCitation":"{\\rtf (\\uc0\\u8220{}Climate and Health: Mortality Attributable to Heat and Cold - The Lancet\\uc0\\u8221{} 2016)}","plainCitation":"(“Climate and Health: Mortality Attributable to Heat and Cold - The Lancet” 2016)"},"citationItems":[{"id":52,"uris":[""],"uri":[""],"itemData":{"id":52,"type":"webpage","title":"Climate and health: mortality attributable to heat and cold - The Lancet","URL":"(15)60897-2/abstract","accessed":{"date-parts":[["2016",5,4]]}}}],"schema":""} (“Climate and Health: Mortality Attributable to Heat and Cold - The Lancet” 2016) [Add other links]. There is clearly more, and high-profile, work to be done in this area.The statistical model underlying this paper ADDIN ZOTERO_ITEM CSL_CITATION {"citationID":"gue7WjeZ","properties":{"formattedCitation":"(Gasparrini 2013)","plainCitation":"(Gasparrini 2013)"},"citationItems":[{"id":38,"uris":[""],"uri":[""],"itemData":{"id":38,"type":"article-journal","title":"Distributed lag linear and non-linear models for time series data","container-title":"Document is available at R project: . r-project. org/web/packages/dlnm/.(Accessed: 4 th May 2015)","source":"Google Scholar","URL":"","author":[{"family":"Gasparrini","given":"Antonio"}],"issued":{"date-parts":[["2013"]]},"accessed":{"date-parts":[["2016",2,23]]}}}],"schema":""} (Gasparrini 2013) is open source and published as a package for the R language ADDIN ZOTERO_ITEM CSL_CITATION {"citationID":"ULwuWqH3","properties":{"formattedCitation":"{\\rtf (\\uc0\\u8220{}Distributed Lag Linear and Non-Linear Models in R: The Package Dlnm | Gasparrini | Journal of Statistical Software\\uc0\\u8221{} 2016)}","plainCitation":"(“Distributed Lag Linear and Non-Linear Models in R: The Package Dlnm | Gasparrini | Journal of Statistical Software” 2016)"},"citationItems":[{"id":29,"uris":[""],"uri":[""],"itemData":{"id":29,"type":"webpage","title":"Distributed Lag Linear and Non-Linear Models in R: The Package dlnm | Gasparrini | Journal of Statistical Software","URL":"","shortTitle":"Distributed Lag Linear and Non-Linear Models in R","accessed":{"date-parts":[["2016",2,16]]}}}],"schema":""} (“Distributed Lag Linear and Non-Linear Models in R: The Package Dlnm | Gasparrini | Journal of Statistical Software” 2016). The same type of statistical model underlies the quantitative modelling in WHO (2014). ADDIN ZOTERO_ITEM CSL_CITATION {"citationID":"b5ZCOfty","properties":{"formattedCitation":"(Honda et al. 2014)","plainCitation":"(Honda et al. 2014)"},"citationItems":[{"id":20,"uris":[""],"uri":[""],"itemData":{"id":20,"type":"article-journal","title":"Heat-related mortality risk model for climate change impact projection","container-title":"Environmental Health and Preventive Medicine","page":"56-63","volume":"19","issue":"1","source":"PubMed Central","abstract":"Objectives\nWe previously developed a model for projection of heat-related mortality attributable to climate change. The objective of this paper is to improve the fit and precision of and examine the robustness of the model.\n\nMethods\nWe obtained daily data for number of deaths and maximum temperature from respective governmental organizations of Japan, Korea, Taiwan, the USA, and European countries. For future projection, we used the Bergen climate model?2 (BCM2) general circulation model, the Special Report on Emissions Scenarios (SRES) A1B socioeconomic scenario, and the mortality projection for the 65+-year-old age group developed by the World Health Organization (WHO). The heat-related excess mortality was defined as follows: The temperature–mortality relation forms a V-shaped curve, and the temperature at which mortality becomes lowest is called the optimum temperature (OT). The difference in mortality between the OT and a temperature beyond the OT is the excess mortality. To develop the model for projection, we used Japanese 47-prefecture data from 1972 to 2008. Using a distributed lag nonlinear model (two-dimensional nonparametric regression of temperature and its lag effect), we included the lag effect of temperature up to 15?days, and created a risk function curve on which the projection is based. As an example, we perform a future projection using the above-mentioned risk function. In the projection, we used 1961–1990 temperature as the baseline, and temperatures in the 2030s and 2050s were projected using the BCM2 global circulation model, SRES A1B scenario, and WHO-provided annual mortality. Here, we used the “counterfactual method” to evaluate the climate change impact; For example, baseline temperature and 2030 mortality were used to determine the baseline excess, and compared with the 2030 excess, for which we used 2030 temperature and 2030 mortality. In terms of adaptation to warmer climate, we assumed 0?% adaptation when the OT as of the current climate is used and 100?% adaptation when the OT as of the future climate is used. The midpoint of the OTs of the two types of adaptation was set to be the OT for 50?% adaptation.\n\nResults\nWe calculated heat-related excess mortality for 2030 and 2050.\n\nConclusions\nOur new model is considered to be better fit, and more precise and robust compared with the previous model.","URL":"","DOI":"10.1007/s12199-013-0354-6","ISSN":"1342-078X","note":"PMID: 23928946\nPMCID: PMC3890078","journalAbbreviation":"Environ Health Prev Med","author":[{"family":"Honda","given":"Yasushi"},{"family":"Kondo","given":"Masahide"},{"family":"McGregor","given":"Glenn"},{"family":"Kim","given":"Ho"},{"family":"Guo","given":"Yue-Leon"},{"family":"Hijioka","given":"Yasuaki"},{"family":"Yoshikawa","given":"Minoru"},{"family":"Oka","given":"Kazutaka"},{"family":"Takano","given":"Saneyuki"},{"family":"Hales","given":"Simon"},{"family":"Kovats","given":"R. Sari"}],"issued":{"date-parts":[["2014",1]]},"accessed":{"date-parts":[["2016",2,16]]},"PMID":"23928946","PMCID":"PMC3890078"}}],"schema":""} (Honda et al. 2014).The fact that the model has been published as an R package means that the whole modelling process is transparent and could be reproduced by anyone with access to the data. Actuaries would therefore engage with the modelling of temperature related mortality in various ways:Attempting to reproduce the results of existing research groups;Using the same model to calculate different results;Present results from the DLNM in different ways, perhaps ones which are more relevant to actuarial applications or which are designed to provoke discussion of different aspects of the results. The R modelling ecosystem includes tools which make creating and sharing interactive visualisations relatively easy: here is an example HYPERLINK "" ;Applying the model to different data sets;Extending or otherwise adapting the DLNM model;Comparing the DLNM with completely different modelling approaches.Use of scenariosModelling climate related impacts on mortality is discussed in chapter REF _Ref450992706 \r \h 6. A powerful alternative to modelling is the use of scenarios to illustrate possible impacts. Actuaries in all sectors who carry out any form of modelling work need to subject their projections to sensitivity testing. Typically in relation to mortality this might include variations in the underlying base table, or the (rate of) projected improvements. Another common approach is to add or subtract a year to the calculated age.In 2014, the UK’s National Association of Pension Funds (since renamed the Pensions and Lifetime Savings Association or PLSA) commissioned a study to examine possible future trends in longevity, based in part on analysis of 2.5 million pensioners from some of the UK’s largest occupational pension schemes. It considered a number of possible future scenarios, shown in the table below taken from their report. The following table highlights the potential climate change scenarios used.In the PLSA’s scenario of increasing costs of energy, economic growth is severely impacted and consequentially health service provision is challenged. Reduced access to and increased costs of food are a further consequence. The poorer sectors of UK society would as a result be unable to afford a trio of basic needs: heating, medicine and a balanced diet. They also posit two consecutive harsh winters early on in the projection. Their results are shown below.This particular scenario therefore led to a lowering of life expectancy versus the central assumptions, with a greater impact on the poorer sections of society. However, it is not the specific scenario that is of most interest, but rather the technique used. Actuaries in different countries could construct similar scenario-based sensitivity analysis of their model results. If combined with consistent economic scenarios this could potentially prove to be a far more powerful test of a model or of a proposed solvency level that simply altering the rate of longevity improvement / discount rate by varying amounts.Dealing with non-stationary riskWhilst not a direct link to insurance or pensions disciplines, it is instructive to realise that the climate scientists looking at the impact of climate on, say, grain production, are dealing with non-stationary risk. This is illustrated in the following diagram taken from The Global Food Security programme (2015).The diagrams below show how distributions for the annual yield in cereal crops are expected to change. On each row the left hand (blue) diagram shows the current observed yield. The middle diagram shows the projected yields for the period 2011-40. The final right hand diagram shows how this distribution / volatility change over longer periods. The top row shows the effect including the effect of CO2 fertilisation (see section REF _Ref451072621 \r \h 4.2 above), and the bottom row excludes this effect (as recently questions have been raised as to its real magnitude).This analysis suggests that what is referred to as an extreme food production shock in the late 20th century will become more common in the future. These data indicate that a 1-in-200 year event for the climate in the late 20th century equates to a loss of approximately 8.5% (top left) and over the next decades (2011-2040) a 1-in-200 year event is about 15% larger in magnitude and equivalent to the loss of 9.8% of calorie production. Furthermore, according to the model, an event that we would have called 1-in-100 years over the period 1951-2010 may become as frequent as a 1-in-30 year event before the middle of the century.Similar techniques may be needed to project for example impacts of climate change on mortality or economic development.Case studiesUnited Kingdom[others]Need for further researchReferences“ar4-wg2-chapter8.pdf.” 2016. Accessed April 27. .“Climate and Health: Mortality Attributable to Heat and Cold - The Lancet.” 2016. Accessed May 4. (15)60897-2/abstract.“Distributed Lag Linear and Non-Linear Models in R: The Package Dlnm | Gasparrini | Journal of Statistical Software.” 2016. Accessed February 16. Economist, 17 March 2012. “Pollution in India: Gasping for Air”.Environmental Health Perspectives, National Institute of Environmental Health Sciences. 2010. "A Human Health Perspective on Climate Change, a Report Outlining the Research Needs on the Human Health Effects of Climate Change. The Interagency Working Group on Climate Change and Health. April 2010. Financial Times November 17, 2015.Gasparrini, Antonio. 2013. “Distributed Lag Linear and Non-Linear Models for Time Series Data.” Document Is Available at R Project: . R-Project. org/web/packages/dlnm/. (Accessed: 4 May 2015). , Antonio, Yuming Guo, Masahiro Hashizume, Eric Lavigne, Antonella Zanobetti, Joel Schwartz, Aurelio Tobias, et al. 2015. “Mortality Risk Attributable to High and Low Ambient Temperature: A Multicountry Observational Study.” The Lancet 386 (9991): 369–75. doi:10.1016/S0140-6736(14)62114-0.Honda, Yasushi, Masahide Kondo, Glenn McGregor, Ho Kim, Yue-Leon Guo, Yasuaki Hijioka, Minoru Yoshikawa, et al. 2014. “Heat-Related Mortality Risk Model for Climate Change Impact Projection.” Environmental Health and Preventive Medicine 19 (1): 56–63. doi:10.1007/s12199-013-0354-6. The Intergovernmental Panel on Climate Change (IPCC). 2014. 5th Assessment Report (Working Group 1).Lelieveld, J., J.S. Evans, M. Fnais, P. Giannodeki, A. Puzzer. 2016. "The Contribution of outdoor air pollution sources to premature mortality on a global scale". Nature. v. 525, 17 September 2015. 367-371.National Research Council. Proceedings of the National Research Council’s Standing Committee on Emerging Science for Environmental Health Decisions workshop in November 2014. Nature, 2014 : Ray, Gerber, MacDonald, West.Samed, H.A., S.R. Khandker, M. Asaduzzaman, M. Yunus. 2013. "The Benefits of Solar Home Systems". Policy Research Working Paper 6724. The World Bank. Smithsonian Magazine (June 2013). “Is a Lack of Water to Blame for the Conflict in Syria?”Springmann, Marco, D. Mason, D’Croz, S. Robinson, T. Garnett, H. Charles, J. Godfray, D. Gollin, et.al. 2016. “Global and regional health effects of future food production under climate change: a modelling study.” The Lancet. Vol. 387, No. 10031, 1937–1946. The Global Food Security programme. 2015. “Final Project Report from the UK-US Taskforce on Extreme Weather and Global Food System Resilience”. UKWHO. 2014. "Quantitative Risk Assessment of the Effects of Climate Change on Selected Causes of Death, 2030s and 2050s.” . World Health Organisation, 2015. Fact sheet 266, September 2015.WHO. 2016. “Preventing disease through healthy environments"World Bank (2016). “High and Dry: Climate Change, Water and the Economy”. Publishing and Knowledge Division. World Bank. ................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download