Climate change is one of the biggest challenges that the African continent and its inhabitants are facing this century. Regions will be affected differently; the continent is vast, spans several distinct climate zones. Complex meteorological drivers are at play, including the Inter-Tropical Convergence Zone, the El Niño Southern Oscillation, the West African Monsoon and the Indian Ocean Dipole, all of which may be impacted to some degree by climate change.

This report, Weathering the Storm, builds upon the 2013 Greenpeace Facing the Weather Gods report, which broadly concluded that climate change impacts for the African continent could be severe by the end of the twenty-first century and that the need to make deep cuts to global greenhouse gas emissions is urgent. Facing the Weather Gods concluded that although the general trend would be for a warmer and drier continent, some countries and regions would be affected more profoundly than others. The expectation in 2013 was that the continent would experience higher temperature rises than the global average, and increasing variance in rainfall over the tropics would lead to more extreme precipitation events, which could impact around 25% of the continent. These broad conclusions still stand, but during the intervening years the science has become more sophisticated and this report addresses those areas that most urgently point to the need for action.

Among the key climate projections in relation to the African continent from the most recent (Fifth) Assessment Report from the International Panel on Climate Change (Stocker et al. 2013, Table TS.2) are:

The ‘State of the Climate in Africa 2019’ report by the World Meteorological Organisation (WMO, 2020) points to temperature and precipitation as being the two key indicators that characterise the current climate in Africa and which continuously affect living conditions on the continent. Confirming the broad warming trends that have been observed over most of Africa, the WMO report also points out that in 2019, Northern and Southern Africa were much drier than normal while much of the Sahel and western central Africa were much wetter. Added to this was the fact that rains were more erratic than normal. Near-term predictions for 2020–2024 suggest:

The ‘State of the Climate in Africa 2019’ report also highlights the impacts of extreme weather on agriculture and points out that:
“After decades of decline, food insecurity and undernourishment are on the rise in almost all sub-regions of sub-Saharan Africa.”

As we enter the third decade of the twenty-first century, it is appropriate to revisit and update the findings and observations made in Facing the Weather Gods.
The intervening years have seen improvements in science and climate modelling, and a wealth of published literature (for example, Scholes et al., 2015; Sylla et al., 2016; Girvetz et al., 2019) on global and regional climate change, as well as vast datasets detailing demographics, climate and land-use change.

In many parts of Africa, the impacts of climate change – heatwaves with greater intensity, duration and frequency, together with droughts, more intense storms, more extreme rainfall events and crop failures – will be exacerbated by a combination of growing population, urbanisation and lack of access to information and resources (including money) to protect homes from extreme heat and from floods. Temperatures over the central interior regions of Southern Africa have been rising at about twice the average global rate of temperature increase over the past five decades (Engelbrecht et al., 2015).

As the climate continues to warm due to continuing greenhouse gas emissions from the burning of fossil fuels and other human activities, the world’s traditional weather patterns are predicted to change. At the same time, it must be acknowledged that the global and regional dynamics that influence weather patterns remain uncertain, particularly so in a climate changing world. As a consequence, the models that make climate projections carry a high degree of uncertainty (Cook et al., 2014). What is clear, however, is that all weather, and the systems that drive it, are taking place in a world that has already experienced climate change.

Greenpeace’s updated report, Weathering the Storm, aims to assess the current state of scientific knowledge regarding the trends and drivers of extreme weather events in Africa, by inter alia:

  1. reviewing the available data on the intensity and frequency of extreme weather events;
  2. providing updated projections for the future that are based on the latest climate models; and
  3. discussing the implications of extreme weather events on human health, food security, resilience to extreme events, biodiversity and human conflicts.

Efforts have been made to limit the use of complex technical terms in this report. However, explanations of commonly used terms and concepts are included in section 7.0.

1.1 Africa in context

Africa is the world’s second largest continent and covers a total land area of approximately 30,365,000 square kilometres (11,724,000 square miles). This is around 20% of the total global land mass. The continent is bisected by the Equator, which means that much of the land mass is in the tropics.

Noteworthy physical characteristics are the Sahara desert in the northwest, the world’s second largest rainforest in the Congo Basin of Central Africa and the 6,400 km (4,000-mile)-long East African Rift System. Africa has many mineral resources but the economies of most countries are dominated, in terms of employment-share, by subsistence agriculture, the productivity of which is expected to be severely negatively impacted by climate change and overexploited soil. In spite of its large landmass, Africa’s population was an estimated 1.35 billion in 2020, or 16% of the global population. The most populated regions of the continent are near lakes and along river basins, in coastal West Africa, Northern Africa and some highland areas. The population density is lowest in the desert and savanna regions (;; Niang et al., 2014).

Sub-Saharan Africa is home to more than half of the world’s extreme poor, amounting to approximately 400 million people. Most of those people live in rural areas and work in agriculture (Porciello et al., 2020). The significance of an increase in population figures is that climate extremes such as heat and rainfall are likely to disproportionately affect those who are least equipped to cope with the impacts of such events. As the twenty-first century progresses, the population of sub-Saharan Africa is expected to increase from 1.06 billion in 2019 to 3.7 billion in 2100 (UN, 2019a). The population increase is projected to be driven by a decrease in mortality of children under 5 years old and an increase in life expectancy (males and females combined) from around 63 years in 2020 to around 75 years in the 2090s. In addition, the continent of Africa (together with Asia) is predicted to experience the fastest rate of global urbanisation during the twenty-first century. Currently, 43% of Africa’s population (424,000,000 people) is urban; by 2050 this figure is predicted to increase to almost 60% (1,258,000,000 people). However, urbanisation is not expected to be uniform across all African countries (UN, 2019b).

Africa crosses the equator and extends to the northern and southern latitudes and as a result its mix of climatic conditions is extremely varied. Climatic zones include humid tropical rainforest in the equatorial regions, seasonally arid tropical regions, desert and also subtropical Mediterranean regions (Hulme et al., 2001). The continent encompasses equatorial tropical forest ecosystems, tropical and subtropical woodlands and savannas, tropical grasslands, arid shrublands and desert vegetation (Midgley & Bond, 2015). The different climates make projecting future climate changes a challenge for climate scientists. Challenges in making future climate projections are also because future greenhouse gas emissions are difficult to predict, and different climate systems may also react in unpredictable ways to climate forcing.

Box 1: What is extreme weather?

No universally accepted definition of ‘extreme weather’ exists, even though an extreme weather event might be described as such by those who experience unusual weather. The nature of ‘extreme’ is relative to the normal or prevailing conditions. For example, a heatwave in Lesotho would have different temperature readings to a heatwave in Mali even though the magnitude of change from the usual temperature might be similar.

Extreme weather events are also described as rare events (and humans are not well adapted to cope with them) or severe events (that create loss or damage to infrastructure and/or ecosystems). Extreme weather events generally involve a number of variables that include a combination of climate variables and drivers and the location of such an event will determine its overall impact (Fig. 1) (Stephenson, 2008).

Extreme weather events can reflect changes in one or more underlying drivers, and the frequency and severity of events can change as a result (Scholes et al., 2015).

For the purposes of this report, the term ‘extreme weather’ is used to indicate weather conditions regarded as unusual or exceptional for the time of year or season and that may have negative social and/or environmental consequences. The extreme weather events covered in this report are consistent with the meteorological perspective: high temperatures, droughts, floods and cyclones. Extreme weather may cause death, starvation, damage to ecosystems, housing, infrastructure and agriculture, and may lead to evacuation or migration of inhabitants or crop failure. Events and issues associated with or caused by periods of extreme weather that are covered in this report include those to human health, food and water security, biodiversity, fire and locust swarms.

Figure 1. A schematic describing rare, or extreme, weather events. Adapted from: Stephenson, 2008.

Box 2: An overview of Africa’s weather systems

The weather systems and phenomena that most affect climate variability in Africa are the West African Monsoon (Lafore et al., 2011), El Niño Southern Oscillation, sea surface temperature in the Indian and Atlantic oceans, and tropical cyclones (that affect East Africa, Southern Africa and the Madagascan coastal regions) (Christensen et al., 2013, section 14.8.7). Climate scientists are concerned because global warming-induced changes to major climate systems – including to drivers of African weather such as the West African Monsoon and the El Niño Southern Oscillation – could pass a tipping point this century that could induce rapid, irreversible climate changes (Lenton, 2011). A brief description of these climate systems is in Box 4.
The Intergovernmental Panel on Climate Change (IPCC) AR5 climate scenarios project that mean annual temperatures across Africa will increase during the twenty-first century and that changes in temperature will be greater over Northern and Southern Africa than Central Africa. Modelled projections of rainfall patterns through the twenty-first century to 2100 are less certain than those relating to temperature, but generally agree that Northern and Southern Africa will become significantly drier from around the middle of this century, whereas Central and Eastern Africa regions are likely to experience increases in mean annual rainfall. In the eastern regions the projection is for an overall wetter climate with more intense wet seasons. Future rainfall patterns for West Africa have varied projected outcomes (Niang et al., 2014). The projections broadly agree with those made in AR4. The IPCC AR5 report (as with AR4) approaches the continent of Africa on a continental or regional basis when it discusses observational and projected climate trends.

Box 3: An overview of climate modelling

Future climate projections are made using computer models. There are two broad categories of climate models:
(i) dynamical models, that are based on the physical climate processes; and
(ii) statistical models that are based on observational data.

Dynamical models are computer programs that simulate the chemical, physical and biological processes that control climate. These models can be ‘atmosphere-only’, ‘ocean–atmosphere’ or ‘Earth System’ models, depending on which parts of the climate system are included (atmosphere, oceans, land, biosphere and cryosphere). Each model type has advantages and disadvantages for researchers. For example, atmosphere-only models are not able to simulate how the ocean and atmosphere interact. This means that the model runs quickly and that uncertainty from the ocean model does not influence the result. However, atmosphere-only models are limited by the assumption that the climate change process being investigated does not influence the state of the ocean (Stone et al., 2019) and therefore cannot capture feedback effects.

By contrast, statistical models do not replicate chemical, physical and biological processes. Instead, statistical models are derived from the analysis of past weather patterns. The models are generated by deriving relationships between different climate parameters in meteorological archives. They can then be projected forward in time to indicate how these parameters might evolve. Statistical models reduce the complexity of the model needed to make a forecast. They are especially useful for predicting local weather patterns that have complexities not included in dynamic models. A key limitation of statistical climate modelling is that the model may not correctly represent climates that are significantly different to the meteorological archives used to design them.

Model accuracy
Climate scientists usually measure the accuracy of a model by comparing their simulation with observational data. For example, scientists test the climate models that are used to project future climates by simulating the present climate or past climate – the assumption being that if the model accurately simulates present and historical climates then the likelihood is that it will be reliable in projecting a future climate scenario (Xulu et al., 2020).

Studies are now also beginning to evaluate climate model predictions made decades ago to subsequent observations of what then actually occurred. One study concluded that, in general, projections published over the past five decades have been accurate in predicting the changes to global mean surface temperature which have since been observed (see Box 5) (Hausfather et al., 2020).

The uncertainties in climate model results stem from three principal areas: scenario uncertainty (for example, uncertainty in the future atmospheric carbon dioxide concentration); natural variability (the day-to-day or decade-to-decade variation in weather and climate); and model uncertainty (models are never perfect representations of the climate system). In comparison to other global regions, Africa has a lack of observational weather data from which to assess climate trends (Han et al., 2019). The incomplete knowledge of the climate system, especially in Africa, contributes to model uncertainty. Some climate models have known deficiencies in simulating the mean rainfall patterns and variability in Africa’s weather patterns (James & Washington, 2013). In South Africa, for example, the models currently in use for operational purposes were developed in countries outside of Africa. Accordingly, the models may not be adequately ‘tuned’ to simulate local conditions (Bopape et al., 2019).

Researchers can estimate uncertainty by comparing the results of many different models, each run multiple times. Models are run many times to introduce small changes to the initial conditions in each model run. This produces different outcomes for each run of the same model. When the same result is produced by multiple climate models they can be more confident in the prediction. This is because each climate model is different and is likely to have different inherent errors or biases. Therefore, if different models produce the same result, that result is less likely to be a consequence of a deficiency in the model and more likely to represent a real climate effect. Model Intercomparison Projects (MIPS) are also used extensively in climate research, for example CMIP3 and CMIP5.

Developing climate models
Work to improve climate modelling aims to understand the discrepancies between the climate projections generated by model simulations and observational data collected from the field. This provides new understanding of the climate system allowing models to be improved. For example, CMIP3 and CMIP5 model projections overestimated the rain in East Africa’s short rains season, which led researchers to doubt the projections made by models for later in the century (Yang et al., 2014). Subsequently, analysis of recent and current weather observations led researchers to conclude that the sea surface temperature of the western Indian Ocean is closely related to the rains over East Africa (Yang et al., 2015).

Choose your model
Different models have known strengths and weaknesses, and the best model will depend upon the region in question and the type of projection being made. Global climate models focus on the overall picture but may not accurately represent regional areas because the resolution is too coarse, which is why regional models are favoured for smaller areas of a country or continent. Dynamic models used to project the East African climate have been good at predicting the short rains and the Indian Ocean Dipole but not good at predicting extreme weather events. As computer climate modelling becomes more sophisticated there will probably be fewer uncertainties in projections of future climate changes (Nicholson, 2017).

A future climate with no human intervention
Climate scientists can create computer models to simulate the most likely scenario in a world that has not experienced anthropogenic greenhouse gas emissions. Such simulations include data from the pre-industrial period (circa 1850) and natural forcings such as volcanic aerosols and solar irradiance (Christidis & Stott, 2014). This allows scientists to estimate how our future climate might have looked without human intervention.

An increasing body of attribution studies are evaluating the extent to which human activity is influencing the climate. These studies seek to distinguish between events that are the result of human-driven climate change and events that may result from natural climate variability and without human intervention.