1. Chapter 7: Modeling the Health Impacts of Climate Change
Protecting our Health from Climate Change: a Training Course for Public Health Professionals
Chapter 7: Modeling the Health Impacts of Climate Change

2. Overview: This Module
Defines and discusses the scenarios used for projecting climate change; and
Reviews approaches taken for modeling the potential health impacts of climate change

3. Selected scenario of temperature change to 210020
Modeled
temperature
change 19 18
2020s
2080s
2050s
Selected scenario of temperature change to 2100 17 16 15
Projecting the health impacts of climate change presents challenges different from considering the future impacts of other risk factors. The “exposure” (in terms of changing temperature and precipitation patterns) will change over time, with high uncertainty about the rate and extent of impacts in a particular region. In addition, there will be significant changes in the demographic structure of most populations, technologies in 2100 will differ from those of today, and socioeconomic development may change the world as much as from 1900 to today. Standard epidemiologic analyses are not designed to deal with these complexities. 14 13
1860
1900
1950
2000
2050
2100
Year 3

4. Estimating Future Health Impacts of Climate Change
Expert judgment
Simple extrapolation
Mathematical/statistical modeling
Bivariate
Multivariate
Fully integrated
The potential future health impacts of climate change can be estimated using:
Expert judgment that considers current health burdens, the driving forces for those burdens, and trends that are likely to affect the health burdens over time. For example, the warming associated with climate change will provide opportunities for more rapid replication of many water- and foodborne pathogens. Rural areas in mountainous areas without sustainable access to improved sanitation can be expected to experience an increase in diarrheal diseases if no additional interventions are implemented.
Current trends can be extrapolated to future periods, assuming no major changes in those trends due to climate change or socioeconomic development. For example, continued increases in temperature could result in the mosquitoes that carry dengue fever, malaria, and other vectorborne diseases to continue to expand their geographic range in mountainous areas.
Models, based on biological properties of disease transmission dynamics or on statistical associations between environmental variables and health outcomes, are being increasingly used to gain insights into how climate change could affect future patterns of climate-sensitive health outcomes. Models range from simple (i.e. considering only one key variable that will change with climate change, such as temperature extremes, to fully integrated models that incorporate all known processes of significance). 4

5. Mathematical/Statistical Models
Simplified representation of a more complex, dynamic relationship
Reduce complexities and background noise to a simpler mathematical representation
Necessarily “wrong” (incomplete, simplified), but useful for:
Insights into processes
Indicative estimates of future impacts
Enhancing communication to peers, public, and policy-makers
Models are simplified representations of complex, dynamic relationships. Models aim to identify key processes for the association between climate change and health, to further insights into how changing weather patterns could affect the geographic range, seasonal length, and incidence of health outcomes. The goal of a “good” model is to provide insights into possible future changes in health outcomes with enough confidence for decision-makers to plan for possible interventions to avoid, prepare for, and effectively respond to the health risks of climate change. For example, several models suggest that a changing climate will provide an opportunity for various vector species to increase their geographic range in mountainous areas in the coming decades. Public health institutions and agencies can use that information to plan for where and when to alter current surveillance programs. 5

6. Models Models are useful Models do not predict
Consistent framework for structuring scientific knowledge
Explore interactions and feedbacks
Particularly if the relationship is strong, or involves a clear threshold above which a outcome event is very likely
Models do not predict
Limited knowledge of all factors driving an outcome
Policy-makers must understand that models estimate changes in probability
Models are difficult to validate
As noted in the slide, models provide consistent frameworks for exploring interactions and feedbacks, but do not predict what will happen because of limited knowledge of all factors that affect an outcome, including how those factors will change over temporal and spatial scales.
The general circulation models that are used to project climate change rely on scenarios of how many people there will be in the world, how wealthy they will be, and what kinds of technology they will use. 6

7. Scenarios
Coherent, internally consistent depictions of pathways to possible futures based on assumptions about economic, ecological, social, political, and technological development
Scenarios include:
Qualitative storylines that describe assumptions about the initial state and the driving forces, events, and actions that lead to future conditions
Models that quantify the storyline
Outputs that explore possible future outcomes if assumptions are changed
Consideration of uncertainties
Scenarios have been developed for the Intergovernmental Panel on Climate Change (IPCC) of pathways to future worlds. The definition of a scenario is provided, as well as the components of a scenario. IPCC scenarios will be discussed on subsequent slides. 7

8. Goals of Scenarios
To provide policy relevant analyses of possible consequences of mitigation policies
To better understand the potential impacts of climate variability and change
To facilitate the development and implementation of effective and efficient adaptation strategies, policies, and measures to reduce negative impacts
Scenarios have been used for the three goals listed. Much of the analysis with scenarios has focused on the first bullet – understanding the possible consequences of mitigation policies. There has been increasing use of scenarios to project the health impacts of climate change, as summarized in the Human Health chapter of the IPCC 4th Assessment Report report. There has been limited use of scenarios to explore adaptation options, although there is increasing interest in doing so. 8

9. SRES Reference Scenarios
IPCC sponsored 40 emissions scenarios for GHGs, sulfur dioxide, and other gases — A1, A2, B1, B2
The scenarios are published in a Special Report on Emissions Scenarios (SRES)
Six have been used for detailed climate calculations
A1B, A1FI, A1T, A2, B1, B2
The Standardized Reference Emission Scenarios were developed by the IPCC to provide consistent scenarios for use by the climate change community. Each envisions a different pathway of world development, as explained in the following slides.
Nakicenovic et al. 2000 9

10. SRES Reference Scenarios (cont.)
Global A1 B1
Social and
Environmental
Economic A2 B2
Regional
The SRES were developed by the IPCC as alternative images of how the future might unfold and are appropriate tools with which to analyze how driving forces may influence future emission outcomes and to assess the associated uncertainties. Four different narrative storylines were developed to describe consistently the relationships between emission driving forces and their evolution. Probabilities or likelihood were not assigned to the individual scenarios. There is no single most likely, or best guess, scenario. None of the scenarios represents an estimate of a central tendency for all driving forces or emissions. None of the scenarios includes policy changes to address climate change.
Each SRES storyline assumes a distinctly different direction for future development, such that the four storylines differ in increasingly irreversible ways. The storylines were created along two dimensions – global vs. regional development patterns and whether economic or environmental concerns were primary. It is important to note that the scenarios do not cover all possible future worlds. For example, there is no SRES world in which absolute incomes are constant or falling.
Population
Economy
Technology
Energy use
Land use
Environment
Driving forces 10

11. A B 1 2 DRIVING THEMES Economics Environmentalism Globalization
Market forces
Economic and technological convergence
Sustainable development 2
Regionalization
Slower economic growth
This table summarizes the driving themes within each of the four families of scenarios shown on the previous figure.
Standardized Reference Emission Scenarios (SRES) were developed by the IPCC as alternative images of how the future might unfold. Four different narrative storylines were developed to describe the relationships between greenhouse gas emission driving forces and their evolution. Probabilities or likelihood were not assigned to the individual scenarios. There is no single most likely, or best guess, scenario. None of the scenarios represents an estimate of a central tendency for all driving forces or emissions.
Each SRES storyline assumes a distinctly different direction for future development, such that the four storylines differ in increasingly irreversible ways. The storylines were created along two dimensions – global vs. regional development patterns and whether economic or environmental concerns would be primary. It is important to note that the scenarios do not cover all possible future worlds. For example, there is no SRES world in which absolute incomes are constant or falling. The A2 and B2 storylines are frequently used in modeling health impacts.
The IPCC recommends that a range of SRES scenarios from more than one storyline be used in any analysis. The smallest subset recommended to capture the range of uncertainties associated with driving forces and emissions is the three scenario families A2, B1, and B2, plus three groups within the A1 scenario family: A1B, A1FI (fossil fuel intensive variant of the A1 storyline), and A1T (variant of the A1 with increased technology development).
For example, the A2 storyline describes a very heterogeneous world with an underlying theme of self-reliance and preservation of local identities. Fertility patterns across regions vary slowly, resulting in continuously increasing global population. Economic development is primarily regionally oriented and per capita economic growth and technological change are fragmented and slower compared with the other scenarios.
The B2 storyline describes a world in which the emphasis is on local solutions to economic, social, and environmental sustainability. It is a world with continuously increasing global population (at a rate slower than A2), intermediate levels of economic development, and less rapid and more diverse technological change than in the B1 and A1 storylines. 11

12. A B 1 2 DRIVING THEMES Economics Environmentalism Globalization
Market forces
Economic and technological convergence
Sustainable development 2
Regionalization
Slower economic growth
The A1 storyline and scenario family describes a future world of very rapid economic growth, global population that peaks in mid-century and declines thereafter, and the rapid introduction of new and more efficient technologies. Major underlying themes are convergence among regions, capacity building and increased cultural and social interactions, with a substantial reduction in regional differences in per capita income. The A1 scenario family develops into three groups that describe alternative directions of technological change in the energy system. The three A1 groups are distinguished by their technological emphasis: fossil intensive (A1FI), non fossil energy sources (A1T), or a balance across all sources (A1B).
The A2 storyline and scenario family describes a very heterogeneous world. The underlying theme is self reliance and preservation of local identities. Fertility patterns across regions converge very slowly, which results in continuously increasing population. Economic development is primarily regionally oriented and per capita economic growth and technological change more fragmented and slower than other storylines.
The B1 storyline and scenario family describes a convergent world with the same global population, that peaks in mid-century and declines thereafter, as in the A1 storyline, but with rapid change in economic structures toward a service and information economy, with reductions in material intensity and the introduction of clean and resource efficient technologies. The emphasis is on global solutions to economic, social and environmental sustainability, including improved equity, but without additional climate initiatives.
The B2 storyline and scenario family describes a world in which the emphasis is on local solutions to economic, social and environmental sustainability. It is a world with continuously increasing global population, at a rate lower than A2, intermediate levels of economic development, and less rapid and more diverse technological change than in the B1 and A1 storylines. While the scenario is also oriented towards environmental protection and social equity, it focuses on local and regional levels.

13. SRES: Population Complete globalization Emphasis on material wealth
Emphasis on sustainability and equity
Emphasis on material wealth
These graphs show the population projections for the four families of scenarios. Population projections depend on assumed fertility and mortality rates, which become increasingly uncertain further into the future. As can be seen in the figures, population peaks then declines in the A storylines; continues to increase in the B1 scenario; and increases then stabilizes in the B2 scenario.
Strong regionalization 13

14. Population Projections
Population projections for the A2 and B2 scenarios were from the UN population projections in 1998 (high and medium projections)
UN Population Division 2002 Revision included further consideration of the impact of the HIV/AIDS epidemic and projected a lower population in 2050 by 0.4 billion people (total 8.9 billion people; medium growth)
If correct, there will be 400 million fewer people in 2050 engaging in activities that burn fossil fuels, etc., thus inflating the estimated cumulative CO2 emissions
Some demographers attach a probability of more than 90% that actual population will be lower than the trajectory adopted in the A2 scenario
The population projections for the A2 and B2 scenarios were from the UN population projections in However, as detailed in the slide, these projections were recognized as too high by 2002, which means the associated emissions also may be high. 14

15. Projections of GDP Depend on
Assumed rate of population growth
Specific economic assumptions made about growth and the implementation of technological changes
The characteristics of the economic model used to project GDP
Assumptions about future exchange rates
Projected future gross domestic product (GDP) is quite uncertain because it depends on (1) the assumed rate of population growth, (2) specific economic assumptions made about growth and the implementation of technological changes, (3) the characteristics of the economic model used to project GDP, and (4) assumptions about future exchange rates. Downscaling adds further uncertainty. The different integrated assessment models used to project GDP for the SRES storylines produce a range of up to 30% in global GDP for a given storyline. 15

16. Emphasis on sustainability
SRES: Economic Growth
Complete globalization
Emphasis on material wealth
Emphasis on sustainability
and equity
All of the SRES storylines describe futures that are more affluent than today, with gross world product rising 10- to 26-fold. Along with the increasing income is a narrowing of income differences among world regions; this implies very high growth rates for all currently developing countries – even those, particularly in Africa, that have shown negative to little growth over the past few decades.
Strong regionalization 16

17. Per Capita Income Ratio (High to Low Countries) in SRES
GDP grows in all countries, from 10- to 26-fold
There is a narrowing of income differences comparing high income to low income countries
1990
16.1 A1 B2
2020
9.4
7.7
2050
6.6
4.0
2100
4.2
3.0
These per capita income ratios show the dramatic narrowing of income differences assumed in two of the scenarios. In the 1990s, high income countries had 16-times the GDP of low-income countries. This ratio is projected to fall to a 3-4 fold difference by the end of the century. 17

18. Economic Growth in SRES
1990 income per capita income
$900 in low income countries
$19,100 in OECD countries
Projections for low income counties
A2 = $3,900 in 2050 to $11,000 in 2100
B2 = $8,100 in 2050 to $18,000 in 2100
Projections for OECD countries
A2 = $34,600 in 2050 to $58,500 in 2100
B2 = $39,200 in 2050 to $61,000 in 2100
For example, 1990 income per capita in developing countries was $900 in 1990 (at 1990 prices and exchange rates). In the A2 scenario, this is projected to increase to $3,900 in 2050 and to $11,000 in In the B2 scenario, the projections are for $8,100 in 2050 and $18,000 in For comparison, growth in the OECD countries was projected to increase from $19,100 in 1990 to $34,600 in 2050 in the A2 scenario and to $39,200 in the B2 scenario; for 2100, the income per capita was projected to be $58,500 (A2) and $61,000 (B2).
Another issue with the SRES storylines is that they were produced for world regions, so there are considerable difficulties in moving to finer spatial scale resolution. A major assumption is that all parts of a region would change at the same rate, including population, GDP, and land cover. 18

19. SRES Fossil CO2 Emissions
This graph shows emissions of CO2 from fossil fuels under the different scenarios. Emissions are projected to continue to be high under the A1 family of scenarios, peak and then fall under the B1 family of scenarios, and to follow intermediate trajectories for the other scenarios. Again, the projected emissions depend on how many people there will be in the world, where they will live, the technologies they will use, and how wealthy they will be. Currently, world emissions are above the A1FI trajectory. 19

20. SRES CO2 Concentration Projections
Based on emissions in the previous slide, these are the resulting atmospheric concentrations of CO2. It can be seen how much higher emissions are under the A1FI (fossil fuel intensive scenario) than the B1 scenario. 20

21. Sources of Uncertainty
Full range of “not improbable” futures captured?
Model uncertainty
Were appropriate models chosen?
Are assumptions and associations likely to remain constant over time?
Rate, speed, and regional extent of climate change
Policy uncertainty
Changes in economic development, technology, etc.
How populations in different regions will respond
Effectiveness of mitigation and adaptation strategies and policies
As you can imagine, there is a wide range of uncertainties associated with these scenarios, from whether the full range of possible futures was captured, to uncertainties about assumed rate and extent of change in demographics, economic growth, etc., to uncertainties about the consequences of current and future policy choices. 21

22. Projected Heat-Related Deaths in Adults > 65, Due to Higher Mean Annual Temperatures, Australia 2100
+ ??? + ???
High GHG emissions
Low GHG emissions
+ 14,000
+ 6,900
+ 11,900
Possible synergistic effect of temperature and aging
(especially at higher temperatures than previously encountered)
+ 6,300
Estimated deaths due to very hot days in 2100
Combined (additive) effect of temperature + aging
Independent effect of aging
+ 2,100
Three examples of the use of scenarios are provided in this and the following slides.
Woodruff et al. used scenarios in their modeling of projected heat-related deaths in older adults due to higher mean annual temperatures under low and higher greenhouse gas emission scenarios. From a baseline of 1,100 annual deaths related to heat, and considering only the independent effect of temperature, very hot days in 2100 were projected to increase the number of deaths by 600 under a scenario of lower greenhouse gas emissions and by 2,100 under a scenario of higher emissions. These projections held demographic change constant.
The second bar shows the independent effect of an ageing population on projected numbers of deaths (i.e. temperature was held constant). The increased number of deaths projected for the higher emission scenario is due to a larger population.
The third bar shows the combined (additive) effect of temperature and aging.
The fourth bar speculates that if there are synergistic interactions between higher temperatures and aging, that the number of deaths in 2100 may be higher than the additive effects of temperature and aging.
+ 600
Independent effect of temperature
Deaths due to very hot days in 2000
Baseline (current) no. of annual deaths related to heat = 1,100
Woodruff et al. 2005 22

23. Malaria in Zimbabwe
Cases by month
Patterns of stable transmission follow pattern of precipitation and elevation (which in turn influences temperature)
> 9,500 deaths and 6.4 million cases between 1989 and 1996
Recent high-altitude outbreaks
This example shows modeling of whether future climate will be suitable for stable malaria transmission in Zimbabwe in The left graph shows the markedly seasonality in malaria cases, with the largest number of cases reported in Nov-June; the peak of cases is in March.
Malaria continues to significantly affect African health, society and economy. The World Health Organization estimated that in 2002, 1, of the 1,272,393 deaths due to malaria occurred in Africa (89.3%).
In sub-Saharan Africa, malaria remains the most common parasitic disease and is the main cause of morbidity and mortality among children less than five years of age and among pregnant women.
Roughly 75% of the deaths from the direct effects of malaria occur in children. This estimate could double if the indirect effects of malaria (including malaria-related anemia, hypoglycemia, respiratory distress and low birth weight) are included when defining the burden of malaria.
Malaria has been shown to decrease economic growth in severely malarious countries by 1.3% per year. The result is that a disease such as malaria creates something of a vicious cycle for poorer countries. Malaria is likely to persist without the creation and maintenance of an appropriate public health infrastructure. However, the economic development necessary for these improvements is directly hindered by the presence of malaria itself. Changes in climate have the potential to make it even more difficult for poor countries to reduce the burden of malaria.
Source: South African Malaria Research Programme
Ebi et al. 2005 23

24. Climate and Stable Malaria Transmission
Climate suitability is a primary determinant of whether the conditions in a particular location are suitable for stable malaria transmission
A change in temperature may lengthen or shorten the season in which mosquitoes or parasites can survive
Changes in precipitation or temperature may result in conditions during the season of transmission that are conducive to increased or decreased parasite and vector populations
Climate is a primary determinant of whether the conditions in a particular location are suitable for stable malaria transmission. Small changes in temperature or precipitation can result in large changes in malaria transmission in areas that are currently marginal for transmission.
Some areas of Zimbabwe, particularly the northern and southern lowveld regions, have year round malaria transmission with peaks in the austral summer months. Stable transmission is partially defined by temperature and precipitation, with altitude a proxy indicator of temperature. Zimbabwe has dramatic elevation ranges that correlate with maximum and minimum temperatures. This heterogeneity coupled with interannual climatic variability results in constantly shifting fringe areas that are prone to malaria outbreaks, similar to many countries in southern and eastern Africa. High altitude regions, which are also the areas of densest human population, are currently malaria-free due to climatic constraints and to a long policy of “barrier spraying” in the transitional elevation zone.
Ebi et al. 2005 24

25. Baseline
In Zimbabwe, as in other countries, patterns of stable malaria transmission are associated with temperature and precipitation. The model was based on the project, Mapping Malaria Risk in Africa (MARA/ARMA), that determined where climate was suitable for stable malaria transmission (based on average temperature, precipitation, and frost).
MARA/ARMA uses three variables to determine climatic suitability for a particular geographic location: mean monthly temperature, winter minimum temperature, and total cumulative monthly precipitation. An important distinction between this model and others is that the MARA/ARMA decision rules were developed using fuzzy logic to resolve the uncertainty in defining distinct boundaries to divide malarious from non-malarious regions. Rather than using a Boolean designation of climatically suitable or not, an area can be assigned a fuzzy logic suitability value ranging from zero (not suitable) to one (suitable). A value of one means that malaria transmission is most likely stable. A value of zero means that transmission is very unstable, with malaria either absent or with rare epidemics. Values between zero and one (0.1–0.9) represent a gradient from unstable to increasingly stable transmission. For all variables, assignment of fuzzy logic values between zero and one were based on a sigmoid curve. The red areas are the regions where the climate is suitable for stable malaria transmission. Orange and yellow colors show areas where malaria epidemics are possible. It can be seen that the capital, Harare, which is in the mountain range that cuts across Zimbabwe from left to right), is currently in an area without a suitable climate for malaria.
Future climate projections were created in the program COSMIC, the output of which was the change in mean temperature (°C) and monthly precipitation from 1990 to These changes in mean temperature and precipitation were added to the baseline climatology for each of the roughly 14,000 grid cells in Zimbabwe different scenarios of climate in the year To represent a range of possible future climates, 16 climate projections were generated from the COSMIC program to create Zimbabwe-specific projections of monthly precipitation and mean temperature starting in 1990 and carried out to 2100.
Ebi et al. 2005 25

26. 2025
Under a mid-range climate change scenario (UKMO), this shows how climate suitability could change by The next slide shows how climate suitability could change by 2050.
Ebi et al. 2005 26

27. 2050
If this scenario is realized, then Harare may have malaria outbreaks within a few decades. Climate, of course, is not the only determinant of malaria outbreaks; other factors include the effectiveness of surveillance and control programs, the status of and access to health care, drug resistance, land use change and others.
Ebi et al. 2005 27

28. Climate Change and Malaria: SRES Climate and Socioeconomic Scenarios
MIASMA 2.2
HadCM3 with A1F1, A2, B1, B2
0.5°by 0.5°grid
Downscaled to national level
Re-aggregated by region
Expert judgment of adaptive capacity (SES, current malaria control)
The third example also is of malaria. More modeling has been conducted for the potential impacts of climate change on malaria than for any other infectious disease.
This study used temperature changes from four climate scenarios to drive the program MIASMA, which uses biological relationships to project the impact of climate change on the geographic range and seasonality of malaria. The study used expert judgment of current malaria control and socioeconomic development to assess the capacity of countries to address possible climate-related changes in malaria distribution and length of the season of transmission.
Van Lieshout et al. 2004 28

29. Climate Change and Malaria under Different Scenarios (2080)
Increase: East Africa, Central Asia, Russian Federation
Decrease: Central America, Amazon
[Within current vector limits] A1 A2
These maps show the results for changes in the seasonal length of the malaria transmission season. The red areas show the areas where there may be a more than 2-month increase in the length of the malaria transmission season; orange shows an approximate 2-month increase; blue shows areas where there may be a 2-month decrease; and green areas show where there may be a more than 2-month decrease. Overall, the study projected increases in the seasonal transmission season in East Africa, Central Asia, and the Russian Federation. Decreases are projected in Central America and the Amazon.
Recommended reading: 2007 IPCC reports: The Physical Science Basis, FAQ 8.1 page 117 B1 B2
Van Lieshout et al. 2004 29