The scientific consensus is that, at a global level, extreme weather events will increase in frequency and intensity as the twenty-first century progresses (for an overview of extreme weather, see Box 1). Land and ocean temperatures have increased on a global scale since pre-industrial times (Hoegh-Guldberg et al. 2018). Last year, 2019, was one of the world’s three warmest years since records began; data show that 2014–2019 were the six warmest years since records began in the late 1800s (Blunden & Arndt, 2020; NOAA, 2020). The rate of temperature increase is not expected to be uniform across the globe and some regions have already exceeded an annual average rise of 1.5 ℃; the changes are particularly noticeable in the Arctic in the cold season and in mid-latitude regions in the warm season. Globally, the trend for rainfall is for increased occurrence of once-rare events (Chen et al., 2020).
Hulme et al. (2001) estimated that the African continent warmed on average by 0.5 ℃ between 1900 and 2000, although others have documented a global average temperature change of 0.89 ℃ over the same period; the warming is primarily attributed to human activity (Perkins, 2015). More recently, however, the United States National Oceanic and Atmospheric Administration (NOAA, 2020) suggested a considerably greater average increase of 0.12 ℃ per decade for the African continent. Analysing the data on such a decadal basis also indicates that, at both a global and continental level, the rate of temperature increase has been greater in the past few decades. For example, from 1880 and for much of the twentieth century until 1970, the global average rate of increase has been estimated to have been around 0.07 ℃ per decade, increasing to an average rate of 0.18-0.19 ℃ per decade from 1971 to the present day (Blunden & Arndt, 2020). In terms of Africa, the most recent estimates from NOAA (2020) suggest an increase from the 0.12 ℃ per decade before 1981 to a much faster rate of 0.31 ℃ per decade since then. The latest ‘State of the Climate in Africa 2019’ report (WMO, 2020) notes that in 2019, temperatures were averaged across mainland Africa at between 0.56 ℃ and 0.63 ℃ above the 1981–2010 long-term mean. The report says that 2019 was probably the third warmest year on record after 2010 and 2016. Temperatures in excess of 2 ℃ above the 1981–2010 average were recorded in South Africa, Namibia and in parts of Angola.
The upward trend in average annual temperature over Africa is evident from data observations and the continent’s ten hottest years have all been since 2005 (Blunden & Arndt, 2020). At least three regions of Africa experienced temperature anomalies last year, in 2019, according to data analysed from two different datasets (HadCRUT 4.6 and NASA GISS). East, West and Southern Africa recorded land surface temperature increases of between 1–2 ℃ in comparison to the 1981–2010 base period. The increases are greater than the 2019 global land and ocean surface temperature, which fell between 0.44 ℃–0.56 ℃ above the 1981–2010 average (Blunden & Arndt, 2020). Future projections for this century for the African continent are increases in mean surface temperature that exceed the global mean, and an increase in the frequency of exceptionally hot days (see section 4.1).
Multi-model projections (using the World Climate Research Programme (WCRP)’s Coupled Model Intercomparison Project phase 3) for Africa under 1 ℃, 2 ℃, 3 ℃ and 4 ℃ warming scenarios project that average temperatures across the continent will rise more than the global average. The magnitude of changes will increase in line with increased warming scenarios, with the greatest impacts being predicted for 3 ℃ and 4 ℃ scenarios. There are greater uncertainties regarding future rainfall patterns. Models predict that there will be changes to rainfall across the continent, but the nature of those changes (more rain, less rain, more intense rain or changes to the seasonality of rains, for example) is uncertain, especially in tropical regions. The precipitation trends from a number of models, however, project a wetter East Africa and changing rainfall patterns over the Sahel with some models projecting an increase in rainfall over the central Sahel and a decrease over the western Sahel (see Box 2).
Observational data appear to support the projections. The World Meteorological Organisation (WMO, 2020) notes that in 2019, rainfall deficit in Southern Africa during the 2018–2019 season exacerbated an existing drought, but that heavy rainfall in 2019 led to flooding, and the footprints of rainfall from cyclones Idai and Kenneth were clearly visible in the annual precipitation anomalies despite the preceding drought conditions. Erratic rainfall in East Africa meant that an incipient drought was superseded by flooding. In addition to East Africa, much of the Sahel recorded above normal rainfall.
The accumulation of anthropogenic greenhouse gases in the atmosphere, largely from fossil fuel production and use, is of such magnitude that even if all emissions of climate-harmful gases were stopped immediately, there would not be immediate stabilization of atmospheric gases. The reason is firstly because of the complexity of the climate system and carbon cycle, and secondly because the persistence of greenhouse gases and aerosols in the atmosphere varies from just days to thousands of years. To remove all anthropogenic methane would take around 50 years, but to remove all anthropogenic carbon dioxide (CO2) could take several hundred years (Collins et al. 2013, p1106). Clearly, the scenario is hypothetical because it is implausible that all greenhouse gas emissions will cease immediately, but the exercise is useful as an example of a ‘best case’ scenario to highlight the urgent need to slow and stop emissions.
Box 4: African weather drivers 101
El Niño–Southern Oscillation (ENSO) The ENSO influences extreme weather events globally, causing floods in some regions and droughts in others. It is a naturally occurring oscillating interaction between the tropical Pacific Ocean and the atmosphere that is composed of two alternate, opposing phases: El Niño and La Niña. El Niño occurs irregularly about every three to seven years and brings warm, dry air to Southern Africa and cool air and rain to eastern equatorial Africa. The opposite happens in La Niña years – cool air and rain to Southern Africa and warm, dry air to equatorial East Africa.
Indian Ocean Dipole (IOD) The IOD refers to an irregularly alternating sea-surface temperature difference in the waters of the west and east Indian Ocean. A positive Indian Ocean Dipole means that sea temperatures are warmer in the western Indian Ocean region and cooler in the east, bringing heavy rainfall to East Africa. A negative dipole is the opposite and causes drier conditions in East Africa. The effects of the IOD are exacerbated if the dipole is strongly positive or negative, which can bring flash floods or prolonged drought, respectively. The IOD also affects weather systems in Australia and Southeast Asia.
Inter-Tropical Convergence Zone (ITCZ) The ITCZ is a band of clouds that forms across the tropics. In Africa, the ITCZ brings seasonal daily intense rainfall between latitudes of approximately 23.5° N and S. The ITCZ shifts seasonally towards the hemisphere that is warmer in relation to the other but the precise mechanisms that control its position, and rainfall intensity, are unclear
2.1 Heatwaves in Africa
Heatwaves are periods of time in which the ambient or outdoor air temperature is higher than usual. Many people will have experienced what they perceive to be a heatwave, but the definition is subjective. In the scientific literature, the definition of a heatwave is inconsistent – it is not possible to provide a universal definition or metric of a heatwave to cover all global regions. Heatwaves develop when high-pressure synoptic weather systems (an anticyclone) remain in the same location for a longer period than expected, which could be days or even months. Other factors are involved in the formation of heatwaves include low soil moisture and teleconnections with other climate systems (Perkins, 2015)
Observed data using figures from the second half of the twentieth century suggest that heatwave duration and intensity has increased over parts of Africa, most notably parts of Southern, East and Northern Africa. The observational temperature data indicate that much of Africa experienced an increasing trend in ‘cumulative heat’ by 50% per decade between 1950 and 2017 (Perkins-Kirkpatrick & Lewis, 2020). This study used ‘cumulative heat’, a new conceptual metric to assess the duration and intensity of heat waves during a season. For example, if a heatwave is defined as air temperature above 30 ℃, and the temperature recorded is 33 ℃, a temperature anomaly of 3 ℃ is produced. If the period of the heatwave lasted for 5 days, the ‘cumulative heat’ produced will be 15 ℃. On this basis, the authors estimated that the extra (cumulative) heat produced by heatwaves over parts of Africa is increasing by 10 ℃ per decade (Perkins-Kirkpatrick & Lewis, 2020).
Studies using numerical models at regional and global scales project that during the twenty-first century, heatwaves will occur more often, at higher intensities, and last for longer under enhanced greenhouse gas concentrations (Perkins, 2015). Modelling studies project that the continent of Africa will experience an increased number of hot and humid days as the century progresses, with a median increase of 2.5 heatwave events per season over Central and Southern Africa (Perkins-Kirkpatrick & Gibson, 2017). Parts of Africa (together with Central America and the Middle East) are projected to experience the greatest impact of heating and might experience an increase in heatwave duration by 10–12 days per season for every one degree of global heating (Perkins-Kirkpatrick & Gibson, 2017). Another projection (Rohat et al., 2019) suggests that by the 2090s, the number of people living on the continent of Africa who will be exposed to dangerous heat conditions may reach 86–217 billion person-days per year. The variance in the figures is because the computer model used different scenarios using 12 Shared Socioeconomic Pathway (SSP) – Representative Concentration Pathway (RCP) combinations.
But it is not only the infrequent but extreme heat events that are becoming more common; less extreme but also higher-than-usual temperatures are being experienced and can create long-term heat stress. Observational data show that Africa has experienced an increase in mean annual temperature over the past 50–100 years and projections suggest that the trend is set to continue. In comparison to the mean annual temperature in the late twentieth century, the mean annual temperature increase for much of the continent of Africa is expected to exceed 2 ℃, and possibly fall in the range 3 ℃ to 6 ℃, by the end of the twenty-first century if high emissions continue (Niang et al., 2014).
A warming climate over the next 50 years is projected to lead to the displacement of an estimated 1-2 billion people globally (figures specific to African countries are not available), particularly those who live in the desert regions along the equator, as the mean annual temperature in some regions exceeds 29 ℃. According to Xu et al. (2020), introducing strict climate mitigation measures and stopping greenhouse gas emissions, as in the RCP2.6 scenario, will reduce the impact of global heating on the human population in the most vulnerable regions.
Although much research into the nature of heatwaves has been carried out globally over the past 10-15 years, there are still data-poor regions, of which Africa is one (along with Central and South America, and India). Further research is needed to understand the extent to which human activity is causing changes to the systems that cause heatwaves and to investigate how climate variability will affect heatwave formation (Perkins, 2015).
In summary, global observed data have shown an increasing trend in the overall number and frequency of heatwave days from the twentieth and into the twenty-first century. Future projections for Africa through the twenty-first century follow the global trend in that the frequency, intensity and duration of extreme heat events are expected to increase. The data trends that show the long-term increase in ambient temperature will be important to policy makers to develop strategies to cope with extended periods of heatwaves and also periods of intense heatwaves. Recommendations from climate scientists are to monitor heatwave events on a global level for three to four decades to enable the broad trends to be more accurately assessed and better understood. Determining heatwave trends using data from fewer than several decades can be difficult because heatwaves are susceptible to internal climate variability, which means that short-term trends may not indicate long-term changes (Russo et al., 2016; Perkins-Kirkpatrick & Lewis, 2020). Failing to take measures, however, to reduce emissions while such studies are carried out to develop the evidence base would result in the time window for effective action being seriously compressed.
2.2 Drought in Africa
Drought can cause economic loss, bring crop failures, put food security at risk, and can lead to a shortage of safe, clean drinking water. Drought is an extended period of time in which a region receives less precipitation than expected. But attributing a drought – as with other extreme weather events – to just one cause is not straightforward because extreme events are usually caused by several different (albeit interacting) factors (see section 4.0
Attribution research is a growing field of study that can help to evaluate the extent to which an extreme event has been driven by climate change. Research investigating the 2011 drought in East Africa found that anthropogenic climate change increased the risk of failure of the long rains in 2011, which had preceded the drought and led to dry conditions. But the same piece of research found that human influence was not significant in the failure of the 2010 short rains, which also created dry conditions but were greatly affected by La Niña (Lott et al., 2013).
Between 2015 and 2017 the Western Cape province in the southwest of South Africa was affected by three consecutive years of below-average precipitation. This, in turn, led to a serious water shortage in Cape Town with the possibility of a complete failure of the water supply at a point designated as ‘Day Zero’. When the Day Zero event was analysed using a risk-based approach as a way of teasing out the part played by climate change it was estimated that climate change had made this otherwise very rare event more likely by a factor of three (Otto et al, 2018b).
In another example, the 2015–2016 extreme drought event in East Africa severely impacted the food and water security of more than 15 million people in Ethiopia, Kenya, Somalia and Southern Africa (Funk et al., 2018). The extreme drought caused severe food shortages and a nine-million tonne cereal crop deficit in the region, which meant that 28 million people had to rely on humanitarian food aid (Collins et al., 2019). In 2016, a very strong negative Indian Ocean Dipole (see section 3.3), affected the climate over East Africa, which experienced a failure in the seasonal short rains in October–December. During that time, some regions received less than 50% of their normal rainfall (Lu et al., 2018). Climate scientists found that anthropogenic climate change contributed substantially to the 2015–2016 extreme drought over East and Southern Africa by accentuating the natural El Niño impacts. Attributing the 2015–2016 drought entirely to anthropogenic climate change would not be accurate because of the strong natural variability in the ENSO and associated sea surface temperatures. However, research suggests that anthropogenic climate change significantly contributed to the exceptionally warm sea surface temperature during the El Niño and to an approximate 16% and 24% reduction in rainfall over East and Southern Africa (Funk et al., 2016; Funk et al., 2018). Other research also concluded that the drought that so severely affected East and Southern Africa was caused by a lack of rainfall exacerbated by a strong El Niño event that decayed into a weak La Niña (Lu et al., 2018).
The following year, in 2017, an extensive drought across East Africa affected Tanzania, Ethiopia, Kenya and Somalia when the March–June rains failed. Research indicates that exceptionally warm sea surface temperatures in the western Pacific and failure of the rains that caused the drought conditions in East Africa are associated with ENSO variations. Attribution research using climate model simulations indicates that the extreme sea surface temperature difference would be extremely unlikely without climate change driven by human activities (Funk et al., 2019). The humanitarian consequence was that the people living in the worst affected regions experienced near-famine conditions (Collins et al., 2019).
Models used to project drought over Southern Africa suggest that, with increased global heating, the intensity and frequency of drought conditions will increase but not all regions will be affected equally. A study focused on four major Southern African river basins (Orange, Limpopo, Zambezi, and Okavango river basins that were chosen for their economic importance in agriculture, mining, power generation and industry) projected that at 2 ℃ above the pre-industrial baseline (1861–1890) there would be a statistically significant increase in drought intensity over the southwestern coast, and an increase in drought frequency by two events per decade. If the average annual temperature increases further, by 3 ℃ in comparison to the pre-industrial baseline, more than half of South Africa and Namibia may be severe drought ‘hotspots’ (Abiodun et al., 2019).
Modelling studies and observational data suggest that the tropics (the meteorological term for the moist tropics and dry subtropics at roughly 30° S and 30° N) are increasing in size and are expanding in a polewards direction. Data suggest that the tropics have widened by around 0.5° of latitude per decade since 1979 (when routine satellite observations became possible). A change in the size of the tropics could lead to changes in the rain belt and expansion of subtropical desert, as well as affecting frequency of drought and wildfires. The reasons for the expansion of the tropics is a subject of current scientific research. One early theory was that the cause was predominantly human-driven, although more recent analysis has suggested that several additional factors might be involved; as well as anthropogenic greenhouse gas emissions and other pollutants, natural variability may also contribute. The expectation is that if greenhouse gas emissions continue then global heating will become the dominant cause driving the expansion of the tropics (Staten et al., 2018). The expansion of subtropical deserts could impact billions of people globally who live in semi-arid regions by affecting livelihoods through impacts on agricultural yields and the availability of freshwater.
2.3 Rainfall in Africa
Projections for future rainfall patterns over the African continent are more uncertain than those for future temperature changes. In other words, climate scientists are more confident in the accuracy of climate models to project future temperature changes than future precipitation changes (see Box 2).
Neither the new-generation climate modelling using Climate Model Intercomparison Project 6 (CMIP6) (11 outputs analysed), nor the previous generation CMIP5 (29 outputs analysed), reached firm conclusions on the amount of change expected to rainfall over the Sahel during the twenty-first century. Greater confidence in precipitation projections over the Sahel (and other regions) will probably only be achieved with greater understanding of global circulation (Monerie et al., 2020).
That said, the general consensus is that under RCP8.5, Southern and Northern African regions are projected to experience decreases in mean annual rainfall by the mid- to late twenty-first century. In contrast, Central and East Africa are likely to experience increases in mean annual rainfall under RCP8.5 from around 2050 onwards. Projections for future rainfall patterns over the Sahel and West Africa are more uncertain because different models have produced different outcomes. However, some regional-scale models project an increase in the number of extreme rainfall days over West Africa and parts of the Sahel during the twenty-first century (Niang et al., 2014; Dosio et al., 2020).
Another climate modelling study, investigating changes to extreme weather events in Africa expected from the middle of the century, projects an increase in the frequency and intensity of rainfall over the Sahel during the summer rainy season, and an increase in the frequency, duration and intensity of rainfall over East Africa (Han et al., 2019). The modelling also projected reduced rainfall and increased duration of dry periods in southeastern Africa. The study has limitations because it used only one regional climate model; if such findings are subsequently confirmed by other models and studies then confidence in them would be increased.
To mitigate the impact of unpredictable and/or extreme rainfall events in future decades, the scientific consensus is to try and restrict the average global temperature rise to 1.5 ℃ (rather than >2 ℃) above pre-industrial temperatures. Doing so is likely to reduce the number of extreme precipitation days in many regions around the world, including Africa (Chen et al., 2020).
2.3.1 Surface water and runoff
Regions subject to extreme rainfall events may experience flooding and accelerated soil runoff. Although the primary driver of extreme rainfall events and altered rainfall patterns is climate change, human activity such as land clearance exacerbates the problems that result by, for example, enabling greater soil erosion, which leads to soil runoff blocking drainage channels.
Land-use change such as deforestation for agriculture, pasture and timber can make a significant impact on landscapes and can impact livelihoods. For example, in a study focused on the Olifants Basin in northeast South Africa, urbanisation and agriculture were identified as causing the greatest changes in surface water runoff, water yield and evapotranspiration. Land use and land cover change analysed using Landsat data from 2000 to 2013 found significant changes in the study area (that drains an area of approximately 50,000 km2) in the balance between urban areas, agricultural lands and rangelands. Rangeland is an area of open land that is not used for growing crops or for agricultural practices, that may be used for hunting or grazing animals and typically is covered with natural grasses and shrubs. Urban areas increased from 13% in 2000 to 23% in 2013 and agricultural land increased from 15% in 2000 to 35% in 2013. By contrast, rangeland decreased from 69% in 2000 to 37% in 2013. The study used a model called the Soil and Water Assessment Tool to simulate the hydrological impact of the land-use and land cover change on surface runoff, water yield, lateral flow and groundwater. The most significant impact of the modelled changes in land use that took place between 2000 and 2013 was water runoff, which increased by 46.9% over the period. In figures, that is an increase of 14.52mm surface runoff water on average, annually, across the Olifants basin in 2013 compared to 2000. A decrease in the average annual groundwater recharge from 34mm in 2000 to 22mm in 2013 was attributed to increased surface runoff, less soil infiltration and higher evapotranspiration. The authors noted that other studies found similar effects of urbanisation (Gyamfi et al., 2016).
Urban environments tend to lack trees and vegetation and can be adversely affected by extreme weather events that can cause runoff (following heavy rainfall), dust storms (in periods of drought) and high heat. Coastal areas and settlements on rivers are vulnerable to sea level rise and flooding from sudden high volumes of water such as intense periods of rainfall during tropical cyclones. Measures proposed to make cities more resilient include reforestation of coastal areas to protect against storm surges and to help absorb excess rainfall, encouraging urban farming and forestry to absorb water and heat and to provide cooling shade (Kareem et al., 2020).
2.3.2 Storms
Storms that form over land are potentially devastating if they coincide with areas of habitation or agriculture because of flooding and run-off. Extreme rainfall events are predicted to increase in frequency globally with climate change.
Equatorial Africa, over the Congo Basin and the Sahel regions, are known to experience intense storms; Sahelian storms, for example, occur seasonally (June to September) in a narrow band between 10°–18°N across West Africa during the West African Monsoon. Although defining an ‘intense storm’ is not straightforward, it is generally accepted that the greater the convective vertical velocity, the more intense the storm (Zipser et al., 2006).
Rainfall over the Sahel varied during the twentieth century, with a wet period in the 1950s and 1960s, followed by drought in the 1970s and 1980s, whereas rainfall patterns since the 1990s have varied between years. Research has found that the frequency of the most intense storms – called ‘mesoscale convective systems’ – has increased over the Sahel in the past four decades, according to analysis of satellite data from 1982–2016 (Taylor et al., 2017). Mesoscale convective systems are a collection of intense storms that can last for at least 12 hours. In the Sahel, mesoscale convective systems can be vast, exceeding 25,000km2, and have been shown on satellite data to coincide with extreme rainfall. The drivers of these intense Sahelian storms are unclear and are likely to be a combination of factors that include wind shear and drying air, coupled with warming temperatures over the Sahara.
Projections suggest a rise in extreme daily rainfall over the Sahel if the connection with the warming Sahara continues, but more research is needed to fully explore this scenario (Taylor et al., 2017). The relevance of these findings is that, with increased frequency and intensity of storms and rainfall, the risk of damaging flash floods also increases. In contrast, current data suggest that the frequency of intense storms is not expected to increase over the Congo Basin.
2.4 Tropical storms and cyclones in Africa
Tropical cyclones are storms that originate over the ocean when the surface water reaches or exceeds around 26 ℃. Tropical storms are associated with extreme rainfall and destructive high winds that can cause damaging coastal storm surges. The tropical cyclones that most frequently affect the African continent are generated in the southwest Indian Ocean basin and make landfall in Mozambique or Madagascar. Tropical storms that affect Africa also form on the Arabian Sea, making landfall in Somalia.
Tropical cyclones are graded from 1 (which have a diameter of 50-100km and sustained wind speeds of 119-153km/h) to category 5 (up to 500km in diameter with wind speeds exceeding 249km/h).
Some of the latest high-resolution computer models have projected that fewer tropical cyclones will form in the Southern Africa region as the climate warms but those that do form will be more intense. This is in line with projections suggesting that, although the global average number of tropical storms will decrease by 6–34% by 2100 (Knutson et al., 2010) as the troposphere is expected to hold more water vapour and latent heat than at present, the globally averaged intensity of tropical cyclones will increase by 2–11% over the same period. Increased intensity of tropical cyclones could enhance the risk to coastal communities of damage from storm surges, which may, in turn, be exacerbated by future sea level rise (Walsh et al., 2016).
As a consequence of fewer tropical cyclones, however, the Limpopo River Basin and southern, central and northern Mozambique are expected to receive less rainfall (Muthige et al., 2018).
The total number of tropical cyclones forming annually in the southern Indian Ocean is projected to decrease by 26% in the late twenty-first century in comparison to the present day (1982–2005), but the number of intense category 4 and 5 storms is projected to increase by 64% according to a computer modelling study. The study used the results of a multimodel ensemble of CMIP5 models for the RCP4.5 scenario to force a high resolution atmosphere model and subsequently a hurricane model to estimate changes in hurricane activity (Knutson et al., 2015), though the study did not project how many tropical storms would make landfall. For comparison, the global number of category 4 and 5 tropical cyclones forming annually in the late twenty-first century is projected to increase by 28% from the present-day baseline. The same study projected an average increase of 8.5% in rain rate, or the quantity of rainfall, by the late twenty-first century for all tropical cyclones forming over the southern Indian Ocean in comparison to the present day baseline (compared to a global projected increase of 14%). The authors suggest that their findings are in agreement with other studies (Knutson et al., 2015).
Tropical cyclones are also affected by wider weather systems. The Mascarene High, for instance, is part of the weather system that determines the path taken by tropical cyclones over the Mozambique Channel and Southern Africa, which can cause widespread destruction when they make landfall (Xulu et al., 2020). See also section 3.2.1.
Tropical cyclones that form in the North Indian Ocean and make landfall may affect East Africa. Future projections (2070–2100) using modelling scenarios with RCP8.5 predict an increase in the genesis of tropical cyclones in the Arabian Sea, but only a small increase of 0.5 tropical cyclones per decade that will make landfall and impact East Africa in the future scenario (Bell et al., 2020).
Nevertheless, cyclones that do make landfall and bring extreme heavy rainfall frequently cause severe damage to homes and infrastructure and may lead to outbreak of disease. Last year, 2019, was exceptionally active for southwestern Indian Ocean cyclones including two of the strongest known cyclone landfalls on the east coast of Africa. In March 2019, one of the most severe tropical cyclones ever recorded in the southern hemisphere made landfall in southeast Africa (WMO 2020). Cyclone Idai caused extensive flooding in Mozambique, Malawi and Zimbabwe, damaging more than 100,000 homes and killing more than 1,200 people – although millions of people are thought to have been affected (Blunden & Arndt, 2020). The human health consequences were serious: cases of diarrhoeal disease were reported at the end of March, and at the beginning of April in excess of 1,400 suspected cholera cases were reported (Chen & Azman, 2019). The estimated cost of damage to local infrastructure was approximately US$2.2 billion (Blunden & Arndt, 2020). The following month, in April 2019, Cyclone Kenneth landed on Cabo Delgado province in northern Mozambique, killing at least 45 people, left around 40,000 homeless and led to a cholera epidemic because the entire sewerage system was destroyed (Cambaza et al., 2019).
2.5 Wildfires in Africa
Fires are integral to some ecosystems and are a natural phenomenon on all continents except Antarctica. Geological records show that fire in tropical savannahs, including in Africa, spread across those regions approximately 7 to 8 million years ago. Fire has therefore been a phenomenon across the African continent for millions of years, and some plants have evolved to require natural fires to trigger germination (Brown et al., 1994).
Routine domestic use of fire by humans began around 50,000 to 100,000 years ago, and the use of fire specifically to manage plants and wildlife is believed to have begun tens of thousands of years ago. In more recent history, in line with the rise in industrialised economies, fire has been used to clear forests for agricultural land. But fires can become uncontrolled, especially during extreme drought events, and that raises the question of whether humans or the climate are more influential in determining fire patterns (Bowman et al., 2009).
Landscape fires can be defined as wild and prescribed forest fires, tropical deforestation fires, peat fires, agricultural burning and grass fires. The interactions between fires, land use and climate are therefore complex.
It is worth noting that, whatever the underlying causes of landscape fires, burning biomass produces harmful toxins including fine particulate matter (or PM2.5, which are particles with a diameter of less than 2.5 micrometers), which contains black carbon components, harmful organic compounds and some inorganic species. One study (Johnston et al., 2012) used satellite-derived observational data of fire activity, to estimate average annual human exposure to landscape fire smoke for the period 1997–2006. The study estimated average annual mortality associated with exposure to landscape fire smoke to be approximately 339,000 worldwide, and disproportionately affected low-income countries. In sub-Saharan Africa, an average of 157,000 annual deaths were due to exposure to landscape fire smoke, making it the world’s most severely affected region in relation to premature deaths from landscape fire smoke (ahead of estimations for Asia at 110,000 annual deaths and South America’s 10,000 annual deaths).
2.5.1 Savannah
Scientists have used data from satellite observations beginning in the 1980s to analyse the extent of fire in global ecosystems and measure the impact of human activity on the occurrence of fires. Savannah ecosystems have the highest frequency of naturally occurring fire because of the alternate wet (which is the growing period) and dry seasons (van der Werf et al., 2008). In Africa, more than 80% of the burned area is in the savannah. However, there has been a decrease in savannah fires in Africa this century. A study that analysed 15 years of satellite data between 2002 and 2016 attributed the decrease in the area burned to increased moisture (rainfall), which reduced the flammability and spread of fire, particularly in the savannah regions (Zubkova et al., 2019).
In its Fifth Assessment Report, the IPCC noted that there were still significant uncertainties about the changes to vegetation cover across the African continent because of complex interactions of the weather systems and the impact of fire and grazing. As computer modelling continues to become more sophisticated and additional data become available, there should be fewer uncertainties in how vegetation responds to climate change (Niang et al., 2014).
2.5.2 Wetlands
Very little research has been undertaken to date on fires in Africa’s wetlands. However, the use of fire by humans as an approach to management and conservation has been investigated in the KwaMbonambi wetlands in South Africa (Luvuno et al., 2016). Evidence from that study suggests that, as Africa’s wetlands have been shaped in the past by the use of fire, any change in the fire regime as a result of climate change would be likely to induce further changes in wetland ecosystems. However, there are currently too few data available to draw any generalised conclusions about the likelihood of natural fires in African wetlands, or of the trends in and impacts of the use of fire as a management tool.