Background
How climate influences malaria
Climate variability and climate change: some basic concepts
Direct effects of climate on malaria
Temperature
Rainfall
Humidity
Covariations between temperature, rainfall and humidity
Indirect effects of climate on malaria
Impacts of climate variations across timescales
Extreme weather and climate events
-
Logistics One of the primary effects of extreme weather is disruption to the normal functioning of society. The practicalities of accessing diagnostics, drugs, vector control and vaccines become particularly challenging during periods of extreme weather. Roads and other transport lines may be out of service, complicating the transport of supplies and personnel, and compromising access to vulnerable populations, especially in remote rural areas.
-
Supply chains As supply chains become increasingly monopolized, the continued provision of key commodities such as vaccines, drugs and ITNs becomes more vulnerable to disruption, both from climate shocks and non-climate factors like political unrest.
-
Vector control Pooling water following periods of rainfall generally increases the availability of suitable breeding sites, but in some cases heavy rainfall has been observed to reduce vector abundance by flushing out and diluting existing sites. The nonlinearity of these mechanisms means that changes to the frequency, intensity and location of extreme rainfall episodes may have unpredictable effects on overall malaria transmission.
-
Vaccine supply If a malaria vaccine (currently in pilot implementation) is deployed widely in the future, power outages and interruptions to normal storage and transport during extreme weather are likely to threaten the ‘cold chain’ [37, 38]. During heat waves, such breaks rapidly expose vaccines to high temperatures, particularly in tropical developing countries where power supply is often intermittent and may be dependent on hydropower.
Seasonality
Interannual climate variability
Decadal climate shifts
Long-term trends
-
The expansion and contraction of climatically-suitable malaria zones as a result of long-term trends would be relevant for all components of a malaria control and eradication programme, potentially requiring a scale-up of new disease control efforts in previously unaffected areas and a review of resource allocation in newly eliminated areas. However, climate variability on all timescales means that these changes will not occur incrementally over time, and thus will need to be managed proactively from year to year.
-
Upward trends in the concentration of carbon dioxide (CO2) in the atmosphere may have chemical effects on malaria risk, separate from their effects on altered climate and weather patterns. Elevated concentrations of carbon dioxide may be linked to delayed larvae development and increased mortality in woodland areas through an alteration in the chemical and nutritional quality of leaf litter [45].
-
Trends in atmospheric CO2 may also affect the nutrition provided by staple crops, with consequences for malaria morbidity and mortality [46‐48]. Rising CO2 concentrations also increase rates of plant growth [49]. Although this plant fertilisation effect provides a slow benefit to agricultural productivity, which may indirectly influence malaria control and eradication through indirect socioeconomic effects, it must be balanced against the complex and more immediate effects of changing weather patterns, seasonality and longer-term variability.
Cross-cutting climate issues affecting malaria control and eradication
Functioning health systems
-
Population access Extreme weather and climate events can disrupt operations, making it difficult for staff to reach work, for supplies to be transported and for people to access health care. Depressed household incomes following weather shocks, particularly in communities heavily dependent on agriculture, may prevent people from seeking medical attention when needed. As transmission dynamics evolve over time, in part influenced by climate, the constitution of vulnerable groups may change. For example, regions transitioning from endemic to epidemic will need to address a shift in susceptible age groups from children under 5 years to people of all ages [50]. Similarly, livelihoods, migration patterns and urbanisation are influenced by environmental change and bring new challenges to which health systems must respond to ensure that vulnerable populations can be identified and accessed.
-
Government funding Currently, governments of endemic countries fund about 30% of malaria control activities [50]. Unless substantial progress towards adaptation is made, the economic impacts of climate change are likely to result in a squeeze on domestic funding for malaria control. Funding for health services may be particularly hard-pressed following extreme weather and climate shocks that cause a drop in Gross Domestic Product (GDP) and tax revenues.
Accurate risk mapping
Effective insecticides
Climate action in other sectors
Climate impacts on food security
Assessing the impacts of climate change on future malaria risk
Climate-driven models of malaria transmission
-
First, any future climate change scenario is contingent on the projected trajectory of greenhouse gas and aerosol emissions which drive climate change (and these trajectories themselves rest on a host of assumptions).
-
Second, while climate models have been shown to reproduce some important aspects of the observed climate system, mostly on larger scales, they have many documented failings [8, 68‐72]. There is thus considerable uncertainty about how the climate system will respond to the external forcing introduced by greenhouse gas emissions. Projections among models can differ dramatically, especially on regional or national scales, with some even predicting opposite changes in rainfall in many parts of the world [40].
-
Third (and often overlooked), the timing of natural climate variations on interannual-to-decadal timescales cannot be predicted by climate change projections. Projections thus cannot be used to predict whether a particular year or decade is going to be wet or dry, but only to examine the general trend. In any given season, year or decade, the climate we experience can differ markedly from the projected trend, even if the trend turns out to be correct (Fig. 2). A projection that East African rainfall is likely to increase by the end of the twenty-first century says nothing about the trajectory between now and that future date. Even if average conditions become wetter, as predicted, fluctuations in climate will still result in periods of drought, flooding, and everything in between. The next 10–30 years (often called near-term climate change) is of great interest for planning, but poses exceptional challenges from a climate prediction standpoint. On this timeframe, interannual-to-decadal variability is crucial, but climate projections cannot be used to predict the timing of these fluctuations. In most places, rainfall trends are less significant on this timeframe, so the precipitation projections are of limited value. Warming trends, accompanied by intensifying heat extremes, are more prominent on these timescales, are expected to continue and should be factored in to long-term research and development activities. However, the precise location and timing of important changes (e.g. reaching key temperature transmission thresholds in new regions, or possible changes in seasonality) are highly uncertain.
-
Fourth, climate models deliver data at a spatial resolution of approximately 50–100km2. Clearly, rainfall, temperature and other climate conditions change on much smaller spatial scales, particularly in mountainous areas which are an important boundary transmission zone. Model outputs, therefore, cannot be used directly to inform practical decision-making at national and local scales.
-
Finally, climate model errors and uncertainty about current and future greenhouse gas and aerosol concentrations are addressed using ensembles: multiple simulations run using different initial conditions, emissions trajectories and climate models in an attempt to sample the full range of possible future climates. Studies projecting the effects of climate change on malaria must therefore examine the full spread of these ensembles. However, even the full ensemble of simulations is not designed to represent the true range of possible futures; that would be an impossible task. Thus, while it is hoped that the true future climate will lie somewhere within the range of projections, there is certainly no guarantee.
Implications for designing a malaria eradication strategy
-
Projections should only be used on continental to global scales, never to plan local responses or draw conclusions about future disease incidence at local scales
-
Projections can only be used to infer broad climate conditions over periods of at least 30 years, never in a particular year or decade
-
The possibility of outcomes occurring outside the range of projected futures should be accounted for explicitly in adaptation planning
-
Projections should be regarded as plausible future scenarios, not predictions
-
Projections should never be used without consulting a climate scientist to perform a thorough evaluation of the climate models on the timescale and for the locations of interest
Climate-proofing a malaria eradication strategy
Using climate information for decision making
Investment 2–5 decades | Carbon emissions mitigation strategies |
Malaria eradication strategies | |
Major infrastructure investment | |
Workforce development | |
Strategic planning 6–20 years | Research and development of medical countermeasures (e.g., drugs, vaccines) and vector control tools (e.g., new insecticides) |
Improving the nutritional content of crops | |
Health facility investments | |
Curriculum development | |
Policy cycles 2–5 years | 4- to 5-year political cycle |
Health service re-organization | |
2- to 5-year research grant cycle | |
Planning cycles < 2yrs | Annual planning and commissioning cycle |
Demand for visible ‘quick wins’ from funders | |
Seasonal preparedness and response < 4 months | Seasonal planning cycle |
Epidemic/disaster preparedness and response | |
Weekly facility management < 1 week | Weather disaster preparedness and response |
Patient scheduling for non-urgent cases |
Box 1: Practical entry points for managing long-term climate trends
Practical recommendations
1. Begin with managing short-term climate risks to control and elimination activities
2. Incorporate climate into monitoring and evaluation of malaria control efforts
3. Invest in monitoring and surveillance systems for climate and malaria
4. Iteratively review and update malaria eradication strategies
5. Long-term investments
6. Build capacity and partnerships
Research priorities
-
Diagnosing the role of climate vs. non-climate factors in the success and failure of previous malaria interventions in order to better target future efforts and justify funding for specific malaria control strategies [activities 1 & 2].
-
Regular evaluations of the malaria suitability of recent climate conditions in marginal and endemic transmission zones, in conjunction with other lines of evidence such as case rates, to identify areas of emerging risk and to course-correct eradication plans as needed (e.g. every 5 years) [activities 1, 2 & 4].
-
Testing the sensitivity of malaria control and elimination strategies to plausible, hypothetical, changes in future climate (including the potential expansion of malaria zones). Stress tests can identify key vulnerabilities within programmes that could be reduced, both to improve current performance and to make future plans more robust to climate change uncertainty, without relying on projections [activities 4 & 5].
-
If important long-term decisions must be made that are highly sensitive to future climate conditions, analyses should be conducted to assess appropriate levels of confidence in the projections on the timescales and in the locations relevant for the decisions in question. Such assessments require bespoke analyses to diagnose where the models perform well, where they fail and why, supported by an understanding of the scientific plausibility of the projections. They cannot be shortcut through a one-size-fits-all approach [8, 81] [activity 5].
-
Investigating methods to depict and communicate plausible future climate scenarios in ways that facilitate robust decision making within disease programming [activities 4, 5 and 6].
-
Identifying case studies that illustrate effective and less effective uses of climate information for malaria control, with particular consideration to how successful examples might be scaled up [activities 1, 2, 4, 5 & 6].