Climate Change Scenarios and African Climate Change
Summary and Keywords
Accurate projections of climate change under increasing atmospheric greenhouse gas levels are needed to evaluate the environmental cost of anthropogenic emissions, and to guide mitigation efforts. These projections are nowhere more important than Africa, with its high dependence on rain-fed agriculture and, in many regions, limited resources for adaptation. Climate models provide our best method for climate prediction but there are uncertainties in projections, especially on regional space scale. In Africa, limitations of observational networks add to this uncertainty since a crucial step in improving model projections is comparisons with observations. Exceeding uncertainties associated with climate model simulation are uncertainties due to projections of future emissions of CO2 and other greenhouse gases. Humanity’s choices in emissions pathways will have profound effects on climate, especially after the mid-century.
The African Sahel is a transition zone characterized by strong meridional precipitation and temperature gradients. Over West Africa, the Sahel marks the northernmost extent of the West African monsoon system. The region’s climate is known to be sensitive to sea surface temperatures, both regional and global, as well as to land surface conditions. Increasing atmospheric greenhouse gases are already causing amplified warming over the Sahara Desert and, consequently, increased rainfall in parts of the Sahel. Climate model projections indicate that much of this increased rainfall will be delivered in the form of more intense storm systems.
The complicated and highly regional precipitation regimes of East Africa present a challenge for climate modeling. Within roughly 5º of latitude of the equator, rainfall is delivered in two seasons—the long rains in the spring, and the short rains in the fall. Regional climate model projections suggest that the long rains will weaken under greenhouse gas forcing, and the short rains season will extend farther into the winter months. Observations indicate that the long rains are already weakening.
Changes in seasonal rainfall over parts of subtropical southern Africa are observed, with repercussions and challenges for agriculture and water availability. Some elements of these observed changes are captured in model simulations of greenhouse gas-induced climate change, especially an early demise of the rainy season. The projected changes are quite regional, however, and more high-resolution study is needed. In addition, there has been very limited study of climate change in the Congo Basin and across northern Africa. Continued efforts to understand and predict climate using higher-resolution simulation must be sustained to better understand observed and projected changes in the physical processes that support African precipitation systems as well as the teleconnections that communicate remote forcings into the continent.
Keywords: African climate change, Sahel precipitation, South Africa climate change, East Africa climate change, Sahel climate change, West Africa climate change, African climate change projections, predicting African climate change, climate models
Predicting future climate change in Africa is vital because of the high dependence of the population on rain-fed agriculture, the lack of robust infrastructure for water resources in many areas, and often limited resources for climate change adaptation. But these projections present special challenges for the African continent since our understanding of weather and climate systems is hampered by a shortage of observations.
Methodology for generating climate change scenarios is discussed in the next section to provide a framework for discussing climate change predictions over Africa, including the crucial issues of regional specificity and uncertainty. This material is followed by a review of our current understanding of how climate may change over West Africa and the Sahel, East Africa, and southern Africa (Figure 1).
Approaches to Predicting 21st Century African Climate Change
Three methods are used to generate information about future climate expectations. One involves perspectives drawn from the study of paleoclimates. It is clear from the paleoclimate record that regional climates, including those across Africa, have changed drastically on time scales ranging from thousands of years to millennia due to such climate-forcing factors as volcanic activity and variations in the Earth’s orbital parameters. The study of past climates uses paleoclimate proxy data to estimate past temperatures and precipitation levels. This information is extracted from the analysis of tree rings, reconstruction of ancient lake boundaries, ocean and lake sediment cores, and coral as well as historical records. Analysis of ice cores from Greenland and Antarctica provides information about atmospheric concentrations of greenhouse gases for the last 800,000 years or more. These data provide perspective on how much climate has changed in the past, and can indicate the cause of the change. A striking example is the African Humid Period, which featured a significantly wetter climate across much of the present-day Sahara Desert (Claussen, 2009; Claussen et al., 1999; deMenocal et al., 2000; Patricola & Cook, 2007; Timm et al., 2010). A great advantage of using paleoclimate proxies to study climate change is their reliance on these observations. However, the paleoclimate analog approach is limited for projecting future climate because the forcing of climate change—the rapid (on geological time scales) increase in atmospheric greenhouse gas concentrations—is unprecedented in the geological history of the planet. In addition, with so much complexity and nonlinearity in the climate system, it is not clear that relationships among climate variables in past climates will be preserved in the future.
Another method that is used to project future climate is the extrapolation of the present-day statistical description of climate. This approach involves the development of statistical models of climate based on observed relationships (e.g., Mtongori et al., 2016). It is regularly used for seasonal forecasting, where the primary cause of variations is changes in solar radiation with the seasons, accompanied by regular changes in sea surface temperatures and large-scale circulation systems. The accuracy of this approach is less certain for predictions of future decades because (a) the climate forcing—namely, increasing atmospheric greenhouse gases—is not part of the observed climate variations that the statistical model is based upon, and (b) statistical distributions of the predictor variables may not be the same in the future climate as in the past.
Dynamical climate prediction is generally seen as the most reliable method for predicting future climate. This method uses computer models that are governed by the laws of physics and represent a compilation of our understanding of how the climate system works. Climate models are based on the set of physical principles embodied in the Navier-Stokes equations. This well-established set of nonlinear partial differential equations governs the physical state of the ocean and the atmosphere, including temperature and circulation, on all space and time scales.
A version of the Navier-Stokes equations, designed to treat hydrodynamic flow on the surface of a sphere, is known as the primitive equations. This set of equations forms the core of climate and weather prediction models, and it is based on the principles of conservation of mass, energy, and momentum. Basic textbooks on atmospheric, ocean and climate dynamics include derivations of these equations from first principles (e.g., Cook, 2013; Holton & Hakim, 2013; Marshall & Plumb, 2008; Vallis, 2006). The set of primitive equations includes the following:
• Newton’s second law of motion. The acceleration of a parcel of air or water occurs when a net force acts on the parcel. The primary forces that accelerate (or decelerate) the wind and ocean currents are pressure gradient, Coriolis, and frictional forces.
• The first law of thermodynamics. Energy cannot be created or destroyed in a closed system. If the climate becomes warmer, there must be a reason for the increase in temperature since temperature measures a form of energy (the kinetic energy of individual molecules). This equation relates temperature change to solar heating, the greenhouse effect, conduction, and phase changes of water vapor (e.g., the release of latent heat when water condenses in the atmosphere).
• Conservation of air mass.
• Conservation of water vapor in the atmosphere.
• Conservation of salt in the oceans.
In addition to these core equations, climate models include a consideration of the land surface. The treatment of the surface in climate models ranges from a relatively simple surface heat balance calculation that accounts for the exchange of energy and momentum between the physical surface and the atmosphere to sophisticated land surface models with multiple soil levels and interactive vegetation. Climate models also include physical parameterizations for processes that operate on space scales that are smaller than the model resolution, such as diffusion and, depending on the model’s grid spacing, convection and cloud formation.
Several types of climate models are defined and commonly referred to using acronyms. A global model that produces the simultaneous solution of the primitive equations is known as a general circulation model, or GCM. GCMs typically have horizontal resolution of 100–200 km, on the order of 50 vertical levels, and integration time steps of roughly 20 minutes. This means that the governing equations are solved on something like 10 million grid points each time step, or billions of times to simulate one year. GCMs are a primary application run on the world’s most powerful supercomputers.
Versions of GCMs that solve the primitive equations in only the atmosphere, with prescribed sea surface temperatures derived from observations, are termed atmospheric GCMs (AGCMs). Similarly, models that solve the governing equations in only the ocean, with prescribed momentum (wind) and thermal forcing from the atmosphere, are known as ocean GCMs (OGCMs). Coupled ocean-atmosphere GCMs (CGCMs, or AOGCMs) bring these two components of the climate system together in an interactive solution, and typically also include dynamic sea ice and vegetation models. At the forefront of GCM development is the addition of interactive chemical budgets that track, for example, the exchange of carbon among the atmosphere, ocean, and land surface components. These are referred to as Earth System Models (ESMs).
Another class of climate models that solves the primitive equations on regional grids at higher resolution than is possible for GCMs is the regional climate model, or RCM. RCMs can capture a more accurate representation of the physical processes that determine climate due to the increase in resolution and the ability to choose physical parameterizations that work well in the region and/or season of interest. They are especially useful in regions with complex and/or steep topography, and for generating climate change information on space scales that are most relevant for impacts analysis. Many RCM simulations use a grid spacing of tens of kilometers, but when this resolution is increased to 3 or 4 km convection can be explicitly calculated by the governing equations. This eliminates the need for parameterized convection in the model and provides improved simulations, for example, of intense rainfall events and diurnal cycles. The trade-off for this increased resolution is limitations on global connectivity and the need to make assumptions about conditions on the model’s lateral boundaries.
The design of climate model simulations is crucial, and decisions should be made in the design based on the goals of the study. For generating climate predictions due to increasing greenhouse gas levels, both equilibrium and time-dependent model integrations are performed. In an equilibrium simulation, the model’s climate state is established with a multidecade simulation of climate with present day or preindustrial levels of atmospheric aerosols, CO2, and other greenhouse gases (most notably, methane). The model is then run again, but with increased CO2 levels; a doubling or quadrupling of CO2 is common. Differences between the two runs are identified as the CO2-induced climate change. Simulations such as these were typical before about 1990, and they formed the basis of our concern that increasing CO2 levels will significantly alter climate.
Time-dependent climate model simulations are run for decades or centuries with time-dependent greenhouse gas levels specified in the model atmosphere. Observed values are used for the past, and levels are increased into the future using projected emissions scenarios. Multiple emissions scenarios are generated by the United Nations’ Intergovernmental Panel on Climate Change (IPCC, 2013) in the course of their periodic evaluations of global climate change. The most recent of those evaluations is the Fifth Assessment Report, known as “AR5,” and the GCM simulations that were run to contribute to the AR5 are called the CMIP5 (Climate Model Intercomparison Project 5) models. The Fourth Assessment Report, AR4, preceded AR5, and the GCM’s runs to support that publication are collectively referred to as the CMIP3 models.
Several intercomparison projects have also been organized for RCM simulations, some with applications to Africa. Various RCMs are paired with boundary conditions from various GCMs to produce a multimodel ensemble simulation of present day and future climate. Some projects also explore the dependence of the simulation on choices of parameterizations used in the model. These coordinated projects result in sets of simulations with large spreads in their simulation of the present-day climate, and a similarly large spread in future climate projections (e.g., Mariotti et al., 2014; Paeth et al., 2011) that can even disagree with the GCM used to drive the RCM’s lateral boundaries (Dosio & Panitz, 2016). In a set of GCM simulations, differences in the representation of climate are due, in large part, to different choices in parameterizations of convection, clouds, and land surfaces. RCMs will have the same source of differences, but also differences due to lateral boundary prescriptions and domain choice.
Another approach to using RCMs to project future climate is to use them as independently as possible from GCMs. In a control or present-day simulation, lateral and surface boundary conditions can be taken from observations. Then, for future simulations, GCM integrations can be used to generate anomalies (i.e., differences between the present day and future climate in the GCM model). These anomalies can then be added to the observed present-day boundary conditions for simulations of the future. In addition, CO2 levels in the regional climate model would be increased according to one of the IPCC emissions scenarios. This methodology has proven useful for projecting future regional climates in various parts of the world, and for understanding the processes of climate change.
Confidence in model simulations is estimated by various methods, but none is a guarantee that a projection is accurate. In comparisons among different models, confidence is enhanced when projections from different models are in agreement. Statistical analysis is used to establish the significance of the results—that the signal is emerging from the noise (weather and natural climate variability) of the model simulation—to ensure that one is evaluating climate change and not natural variability. In another approach, the physical processes of climate change are evaluated and compared with models of natural variability or, more recently, observed climate change to evaluate realism.
The generation of multiple realizations of future climate, known as ensemble members, is also used to evaluate confidence in projections. An ensemble can be composed of multiple runs of the same model, or one or more integrations of different models, to form a multimodel ensemble. The intent is to suppress internal (natural) variability and the vagaries of individual models to clarify the response to greenhouse gas forcing.
African Climate Change Projection: Challenges and Predictions
Because of the distortion of Mercator (and similar) projection maps that are commonly used in schools and text books, the size of Africa is often unappreciated. Africa is the second largest continent, after Asia. The continent is larger than the United States, China, India, and most of Europe combined. This enormous size, and Africa’s essentially symmetric latitudinal expanse into the Northern and Southern Hemispheres from 37 ºN to 34 ºS, means that there are many different climate regimes across the continent. Climate change over such a large expanse must be approached regionally. Here the current state of knowledge of climate change projection for West Africa and the Sahel, East Africa, and southern Africa (Figure 1) is discussed.
West Africa and the Sahel
The West African monsoon system is the primary source of moisture for West Africa and the Sahel. This region includes the countries of the Guinean Coast (Cameroon, Nigeria, Benin, Togo, Ghana, Ivory Coast, Liberia, Sierra Leone, Guinea, Guinea Bissau, Gambia, and Senegal) and the western and central African Sahel (Chad, Niger, Burkina Faso, Mali, and Mauritania).
The annual cycle of rainfall over West Africa is displayed in Figure 2.
Maximum rainfall rates occur close to the equator in January through April. Higher rainfall rates, in excess of 10 mm/day, develop along the Guinean coast (~4 °N) in early May (Guo & Adler, 2004; Nguyen et al., 2011; Okumura & Xie, 2004; Thorncroft et al., 2011). Coastal precipitation persists through June, with the location of the precipitation maximum moving gradually from 3 °N to 5 °N. By the middle of June the precipitation maximum is close to 5 °N, and it remains there with diminishing rainfall rates until the third week in July. At the same time, rainfall rates in the Sahel increase over a wide region from about 9 °N to 14 °N, with maximum rates of 9 mm/day in September near 11 °N. The composite effect is a sudden movement of the rainfall maximum from the Guinean coastal region into the Sahel in July that is known as the “monsoon jump” (Cook, 2015; Hagos & Cook, 2007; Le Barbé et al., 2002; Ramel et al., 2006; Sijikumar et al., 2006; Sultan & Janicot, 2000). Rainfall in the Sahel begins to decrease in late September, and the maximum moves smoothly back to about 2 °N by the middle of December.
There are two primary flows of moisture onto the continent that hydrate the West African monsoon system. One is the northward monsoon flow across the Guinean Coast. This flow carries moisture from the Gulf of Guinea to the countries of the Guinean coast primarily in May and June. After the rainfall maximum moves into the Sahel in late June or early July, moisture transport from the subtropical North Atlantic near 10 ºN by the West African westerly jet plays a more prominent role (Grodsky et al., 2003; Pu & Cook, 2010, 2012).
Land/sea temperature differences play a central role in generating this climatological moisture transport onto the continent. However, rainfall requires not only a supply of low-level moisture, but also mechanisms for moisture convergence and lift. Surface temperature distributions on the continent play a crucial role here. Of primary importance is the seasonal development of the Saharan heat low for the proper simulation of the precipitation climatology and its variations (Biasutti et al., 2009; Lavaysse et al., 2009; Skinner et al., 2012).
Confidence in the ability of GCMs and RCMs to produce reliable predictions of climate change in West Africa depends, in part, on their ability to simulate the present day climatology. Because an accurate simulation of the present day climatology does not guarantee a projection of the future will be accurate, it is seen as a necessary but not sufficient condition. As seen in Figure 3, a regional model simulation with prescribed SSTs (Vizy et al., 2013) can capture the seasonality well, but rainfall rains tend to be stronger than observed.
CMIP5 CGCMs, with calculated SSTs, tend to misrepresent the seasonality of rainfall. Some CGCMs are not able to generate the monsoon system in the sense that the precipitation maximum is not shifted into the Sahel in July (Cook & Vizy, 2006). Model deficiencies are not exclusively, or even primarily, related to resolution issues, but also to the choice of physical parameterizations of convection, cloud formation, land surface processes, and radiation.
Expected Future Climate
Numerous studies have been conducted with both GCMs and RCMs to advance our understanding about how climate is likely to change over this region. Biasutti et al. (2008), Druyan (2011), and others examined CMIP3 projections that were discussed in the AR4. They concluded that a consensus on climate change projections for West Africa did not emerge from this research activity.
Progress was made with the CMIP5 models in the level of agreement for climate change predictions over West Africa. Figure 4 shows projected July—September (JAS) surface temperature differences averaged over the late 21st century (2081–2100; equivalent to 20 ensemble members) from an RCM with 30-km resolution (Vizy et al., 2013) and 5 CMIP5 CGCMs driven by a business-as-usual emissions scenario, assuming no global action to reduce carbon emissions, from the IPCC known as RCP8.5 (Riahi et al., 2011).
(Representative Concentration Pathway 8.5 is an emissions scenario under which the added radiative forcing from increases in greenhouse gases reaches 8.5 W/m2 in 2100.) Temperatures across the northern Sahel/southern Sahara in the RCM simulation increase across West Africa, with strongest warming in the north (Monerie et al., 2013; Vizy et al., 2013). Temperatures increase by over 6 °K in some regions of northern Sudan and Niger, and by 3 to 4 °K between the equator and 10 °N. With less warming along the Guinean coast, meridional temperature gradients in the Sahel are increased. The CMIP5 CGCM projections are similar in structure—each predicts the greatest warming over the Sahara and an intensification of the Sahel meridional temperature gradient, although magnitudes vary among the models. For mid-century (2041–2060; not shown), weaker differences are simulated (e.g., about 2 °K warming in the north) but the structure of the temperature anomalies is the same as for the end of the century.
End-of-century precipitation differences projected by the same models for JAS are shown in Figure 5 along with low-level (925 hPa) wind vectors and normalized geopotential height anomalies (Vizy et al., 2013).
(Geopotential height anomalies are normalized by subtracting domain-averaged differences. Removing the mean change in geopotential heights in the warmer climate better focuses attention on the changes in gradients that lead to circulation changes.) Precipitation anomalies that are not statistically significant at the 95% confidence interval are not shown. Summer precipitation is projected to increase across the Sahel in each simulation (see also Patricola & Cook, 2010, 2011) in association with enhanced moisture transport by the low-level westerly flow including the West African westerly jet. The enhanced eastward low-level flow is, in part, geostrophic (parallel to the geopotential height lines, with magnitude proportional to the meridional geopotential height gradient). It is a response to the amplified heating over the Sahara that occurs in each simulation (Figure 4). Again, results for the mid-21st century are similar but weaker. At this time, only about half of the models produce statistically significant precipitation differences from the late-20th century, but the amplified warming over the Sahara and the increase in meridional geopotential height gradients are consistent across the models.
Confidence in the projected increase in Sahel rainfall under greenhouse gas forcing is supported by the agreement among the models, the evaluation of statistical significance, and our ability to understand the physical basis for the precipitation increase, since dry continental regions are expected to warm more quickly than wet continental regions and the oceans (Sutton et al., 2007). Observations also support this confidence, since an amplified warming over the Sahara is currently observed (Cook & Vizy, 2015; Vizy & Cook, 2016).
As seen in Figures 2 and 3, another important rainy season is during May and June along the Guinean coast. Unlike the summer rainy season in the Sahel, the model simulations do not produce a consensus about change in this region during boreal spring (Figure 4). Even though an amplified Sahara warming occurs during May and June (not shown), the amplification is weaker and the Guinean coast is farther removed from its effects.
Most if not all of these modeling studies do not account for the impact of land cover changes on the future projections. Paeth and Thamm (2007) and Paeth et al. (2009) demonstrate the potential impact of accounting for land cover changes in their regional climate modeling studies for the first half of the 21st century. Without accounting for land cover changes, rainfall is projected to increase slightly over the Sahel in their model, but not so when land changes are factored in. That being said the sensitivity of the results to land cover changes is thought to be model dependent, and additional study is needed along with accurate projections of how land use may change in the future.
Increases in rainfall in the Sahel are expected to be delivered, at least in part, by increasing numbers of intense events (e.g., Cretat et al., 2014; Sylla et al., 2015). The recent observed recovery of rainfall in the Sahel has also been accompanied by more extreme rainfall events than in the past (Sanogo et al., 2015), pointing to the possibility that this recovery is related to increasing greenhouse gas levels.
The Horn of Africa is the easternmost extension of the continent into the Arabian Sea, and is composed of the countries of Ethiopia, Eritrea, Somalia, and Djbouti. Its nearly 2 million square kilometers is home to roughly 115 million people, and it is one of the most food insecure regions on the planet (FAO, 1978). That insecurity is, in part, related to its vulnerability to climate variability and change (Busby et al., 2014; O’Loughlin et al., 2012). To the south of the Horn of Africa is equatorial East Africa, consisting of the nations of Kenya, Tanzania, Uganda, Rwanda, Burundi, and South Sudan. Taken together, the Horn of Africa and equatorial East Africa regions are known as the Greater Horn of Africa.
The distribution of rainfall over East Africa, including its seasonality, is complicated. Figure 6a–d show the seasonal rainfall climatology throughout East Africa as observed by the TRMM satellite, which provides complete coverage of rainfall over the region and is found to be in good agreement with ground-based precipitation measurements over Africa (e.g., Pfeifroth et al., 2016; Roca et al., 2010).
Some regions of East Africa—most notably, southern Ethiopia, southern Somalia, most of Kenya, and northern Tanzania—experience two wet seasons. The boreal spring rainy reason is known as the long rains, and the boreal fall season is called the short rains. This bimodal precipitation distribution is often attributed to “two passes” of the intertropical convergence zone (ITCZ) during the year. However, this is an oversimplification of both the distribution of rainy seasons (Hermann & Mohr, 2011) and the physical processes that determine the region’s seasonality. These seasonal characteristics are not zonally uniform or symmetric about the equator, as they would be if solar forcing was the only factor involved. For example, while Kenya experiences two rainy seasons, the climate directly to the west along the equator is a humid regime, with rainfall all year around supporting the Congo Basin rainforests.
Flanking these equatorial regions with bimodal rainy seasons lie regions with a single warm-season precipitation maximum. The northern half of Ethiopia, Uganda, and most of Tanzania experience a single wet season, sometimes with a mid-season break in precipitation.
Accurate simulation of East Africa in both GCMs and RCMs is complicated by the importance of such factors as regional topography; connections with the Indian monsoon (Camberlin et al., 2001; Vizy & Cook, 2003); intense circulation features such as the Somali (Findlater, 1969; Kinuthia & Asnani, 1982; Riddle & Cook, 2008) and Turkana (Nicholson, 2016) low-level jets; and teleconnections with central and West Africa. In general, atmosphere-only models with prescribed sea surface temperatures from observations produce reasonable simulations, while coupled models do not (Liebmann et al., 2014). Typical concerns in CGCM representations of this region, shown in Figure 6e–l, are long rains that are too weak and short rains that are too strong (Otieno & Anya, 2013; Rowell et al., 2015; Yang et al., 2015a). Yang et al. (2015a) compare coupled and atmosphere-only versions of the MRI GCM and relate the improved simulation in the atmosphere-only model to improvements in the low-level moist static energy and convective instability, ultimately related to having an accurate representation of western Indian Ocean SSTs. RCM simulations, with higher resolution to better capture the region’s important orography, an optimized choice of physical parameterizations, and prescribed sea surface temperatures, can provide reasonably accurate simulations of the East African rainy seasons on regional scales. Figure 6m–p show seasonal rainfall distributions from the 30-km resolution simulations described in Vizy et al. (2013, 2015). Unlike most GCM simulations, these regional simulations capture the observed stronger rains in boreal spring than in boreal fall in regions with bimodal precipitation distributions.
Expected Future Climate
In southern Ethiopia and southern Somalia, with two rainy seasons, observations indicate a trend in which the long rains are becoming weaker (Lyon & DeWitt, 2012), with especially damaging impacts ensuing when this reduction is paired with a failure of the short rains. Funk et al. (2014) relates low rainfall in boreal spring with various Pacific and Indian Ocean sea surface temperature indices. For 2014 in particular, they find a close association with Western Pacific SST gradients (Funk et al., 2015), similar to Yang et al. (2015b). Liebmann et al. (2014) relate the observed decline in the long rains throughout the eastern Horn of Africa to an increase zonal sea surface temperature gradient between Indonesia and the central Pacific Ocean.
Cook and Vizy (2013) evaluated output from 90-km resolution RCM simulations over East Africa to predict and analyze precipitation changes caused by greenhouse gas increases. Two six-member ensembles are used, representing the late 20th century and the mid-21st century. The 20th-century simulation uses boundary conditions from reanalysis climatology, and these are modified for the mid-21st-century simulation using anomalies from coupled GCMs. Results are similar in 20-year RCM simulations with 30-km resolution. In these simulations, the long rains are weaker in the mid-21st-century simulation compared with the 20th-century simulation due to an anomalous dry, anticyclonic flow that develops over the Arabian Peninsula and the northern Arabian Sea.
There is also evidence that the short rains in boreal fall are strengthening in the Horn of Africa, but the statistical significance of these trends is not firmly established (Liebmann et al., 2014). Sensitivity to the occurrence of El Niño and La Niña events (El Niño/Southern Oscillation, ENSO) has been identified (e.g., Gamoyo et al., 2015; Indeje et al., 2000), introducing pronounced interannual variability of the short rains and making a definitive identification of trends more difficult.
Lott et al. (2013) examined large ensembles of the 2010/2011 East African rainy season composed of hundreds of GCM simulations with and without enhanced CO2 and modified SSTs. They conclude that anthropogenic forcing increases the probability that the long rains fail, but does not strongly influence the short rains, which are more closely tied to ENSO or, more generally, Indian Ocean SSTs (Lyon, 2014; Nicholson, 2015). Marthews et al. (2015) used a 0.44 ° resolution RCM embedded in a GCM to generate pairs of simulations of the 2014 long rains season with and without enhanced CO2 to investigate the role of greenhouse gas forcing on recent East African (Ethiopia, Somalia, and Kenya) climate. Their statistically-based analysis did not reveal a role for greenhouse gas forcing in the failure of the long rains in 2014. Kent et al. (2015) finds that there is a large spread in GCM projections for this region under atmospheric greenhouse gas forcing.
Similar to the Horn of Africa region, observational analysis suggests that there are negative trends in the long rains in equatorial east Africa (e.g., Maidment et al., 2015). Schmocker et al. (2016) examined rainfall records of 30 or more years from 50 stations near Mount Kenya and found that 70% of the stations are experiencing this change. Schmocker et al. (2016) also find an increasing trend in the short rains. This pattern of change is consistent with projections for the mid- and late 21st century over the entire Greater Horn of Africa region using RCM simulations (Cook & Vizy, 2013). In these simulations, the reduction in the short rains is a secondary response to precipitation enhancement over the Congo basin. The boreal fall short rains season is lengthened in the 21st-century simulation in southern Kenya and Tanzania in association with a northeastward shift of the South Indian convergence zone (SICZ) (Cook, 2000).
Southern Africa is defined here as the countries south of about 10 ºS, including Angola, Zambia, Malawi, Mozambique, Madagascar, Zimbabwe, Botswana, Namibia, Lesotho, and South Africa. The surface over much of the region lies one kilometer or more above sea level (Figure 1), with elevations above three kilometers associated with the Central Plateau and the Great Escarpment in South Africa.
Most of southern Africa experiences a single rainy season, although Hermann and Mohr (2011) classify small regions near the southern tip of the continent, in central Madagascar, and in parts of coastal Mozambique as having bimodal rainfall seasonality. Highest rainfall rates generally occur during the summer months (Figure 7a) between 10ºS and 20ºS, with precipitation amplified over the topography of Madagascar. The SICZ is apparent, extending off the continent to the southeast into the South Indian Ocean. Rainfall rates exceeding 3 mm/day extend southward to about 33ºS along the east coast. There is a strong east/west rainfall gradient across South Africa, with dry conditions prevailing in the west. These dry conditions in western South Africa extend farther north into Namibia and western Botswana in the fall (Figure 7b). The rainfall distribution in fall is similar to that in summer, but with lower precipitation amounts. Summer brings dry conditions throughout the region (Figure 7c). Spring reestablishes the east/west rainfall gradient over South Africa (Figure 7d), but the east coast to the north (5 ºS–25 ºS) remains dry during this season.
During the single, warm-season rainfall maximum in the northern part of this domain, the large-scale regional circulation is dominated by the Kalahari anticyclone, a mid-tropospheric feature that marks the subsiding branch of the Hadley circulation over southern Africa. This high overlies the Angola thermal heat low (Hachigonta & Reason, 2006; Usman & Reason, 2004) that develops during the warm season as the continent heats up more than the adjacent ocean. Anomalously wet conditions occur in association with enhanced low-level flow from the southwestern Indian Ocean and the warm Agulhas Current, while dry anomalies are associated with increased low-level dry air advection from the South Atlantic Ocean and the cool Benguela Current (Dieppois et al., 2016; Reason, 2002; Vigaud et al., 2009; Washington & Preston, 2006). Thus precipitation anomalies are associated with variations in both the South Atlantic and Indian Ocean subtropical highs, and reliable prediction hinges on our ability to accurately project changes in sea surface temperatures and ocean currents.
Interannual variability of the southern Africa climate reflects a significant sensitivity to ENSO (e.g., Lindesay, 1988; Ropelewski & Halpert, 1987) with drying and lower crop yields typical during an El Niño event. The mechanisms of this connection are known. They are similar to the physical processes that explain the sensitivity of the region to decadal-scale modes of natural variability, and involve shifts in the SICZ and adjustments of the Pacific Walker circulation (Cook, 2000, 2001; Dieppois et al., 2016). Surface temperatures throughout southern Africa also respond to variations of the southern annual mode of variability (Manatsa et al., 2015).
Expected Future Climate
Seasonal precipitation over subtropical southeastern Africa has changed in recent decades, with the onset of the wet season shifting from October to December, an earlier demise, and a higher rainfall maximum (ActionAid, 2006; Manford et al., 2011; Tadross et al., 2005, 2009). Estes et al. (2014) documents a decline in water availability in Malawi and Tanzania in the 1979–2010 period driven by both precipitation decreases and increasing water demand. These changes have driven modifications in agricultural practice throughout the region. As is the case for most of Africa, climate simulation and the detection of a regional climate change signal is hampered by a scarcity of observations. The CMIP5 CGCMs overestimate summer rainfall over southern Africa, perhaps in association with excessive moisture transport by the Angola low (Lazenby et al., 2016), and do not accurately represent the observed association with ENSO (Dieppois et al., 2015). Seasonal precipitation distributions from the 30-km resolution RCM simulations described in Vizy et al. (2015) are shown in Figure 7e–f. These simulations capture the present day climatology and seasonality of southern Africa rainfall quite well, albeit with the high rainfall rates in the tropics that are typical of regional simulation.
A series of RCM projections for 2081–2100 driven by various CGCMs and the IPCC RCP8.5 emissions scenario has been generated by the Coordinated Regional Climate Downscaling Experiment (CORDEX) (e.g., Dosio & Panitz, 2016) for southern Africa (south of 22 ºS). The overestimation of summer rainfall is not as pronounced in the higher-resolution RCMs, and Shongwe et al. (2015) reports that the austral summer rainy season is represented accurately over southern Africa in the RCM ensemble mean. Mean surface warming is projected to be in the 3.5–5K range for both austral summer and winter, with general agreement across both the RCM and CGCM models. There is much less agreement about precipitation changes, however, with opposite-signed changes simulated in different models.
The differences are driven by changes in sea surface temperatures taken from CGCM projections, and CO2 increases according to the IPCC8.5 emissions scenario. During the warm season (Figure 8a) and extending into the fall (Figure 8b), dry anomalies of roughly 1 mm/day are projected for the continental interior between 10 ºS and 20 ºS. The drying is large over western Zambia and the southern Democratic Republic of Congo in the summer (Figure 8a), while rainfall increases occur on the east coasts of Tanzania and northern Mozambique. In the fall, the drying in the interior extends farther east to Malawi and Mozambique. Vizy et al. (2015) find that this fall drying is associated with an earlier demise in the rainy season in Malawi. The early demise is associated with a strengthening of the thermal low over southwestern Africa due to increased surface heating over the Kalahari Desert, analogous to the amplified warming over the Sahara (Cook & Vizy, 2015; Vizy & Cook, 2016). There is clear added value from high-resolution RCMs in simulating and projecting changes in extreme precipitation compared with GCMs (Cretat et al., 2014; Dosio et al., 2015). In the CORDEX ensemble of RCM simulations, precipitation at the end of the 21st century over southern Africa is projected to have a changed distribution within the rainy season. While significant changes in the seasonal total precipitation are not predicted, the number of consecutive dry days and the occurrence of intense rainfall events both increase (Pinto et al., 2016).
This discussion provides a state-of-the-art review of our current understanding of how and why climate may change in the monsoon regions of West and East Africa, the Sahel, and southern Africa. Missing is information about the potential for climate change in the Congo Basin and North Africa. Very little is understood about climate variability and change in these regions—much work remains. Progress hinges on having observations of sufficient quality and quantity for evaluating the performance of climate models, and a concentration on understanding the physical processes of climate change. There is great hope for progress as higher-resolution simulation improves our ability to accurately represent climate—including regional processes that support precipitation systems and teleconnections that communicate remote forcings regionally.
It is important to remember that there are two primary sources of uncertainty in climate prediction. One stems from deficiencies in our understanding of how regional climates work, and these deficiencies are reflected in climate model output when it does not accurately reproduce the observed climate. The other source of uncertainty stems from not knowing how emissions of CO2 and other greenhouse gases will develop through the 21st century and into the 22nd. After the mid-21st century, uncertainty in emissions scenarios begins to dominate uncertainty in model projections. The message is clear: How we proceed as a species to reduce emissions and atmospheric greenhouse gas levels is critical for the stability of the climate and human populations that it sustains.
The author thanks Dr. Edward Vizy for assistance with figures, and Drs. Martin Claussen, Jaya Khanna, Edward Vizy, and an anonymous reviewer for their helpful comments on the manuscript.
ActionAid. (2006). Climate change and smallholder farmers in Malawi. (Unpublished manuscript.) Retrieved from https://www.actionaid.org.uk/sites/default/files/doc_lib/malawi_climate_change_report.pdf.
Biasutti, M., Held, I. M., Sobel, A. H., & Giannini, A. (2008). SST forcings and Sahel rainfall variability in simulations of the twentieth and twenty-first centuries. Journal of Climate, 21, 3471–3486.Find this resource:
Biasutti, M., Sobel, A. H., & Camargo, S. J. (2009). The role of the Sahara low in summertime Sahel rainfall variability and change in the CMIP3 models. Journal of Climate, 22, 5755–5771.Find this resource:
Busby, J. W., Cook, K. H., Vizy, E. K., Smith, T. G., & Bekalo, M. (2014). Identifying hot spots of security vulnerability associated with climate change in Africa. Climatic Change, 124, 717–731.Find this resource:
Camberlin, P., Janicot, S., & Poccard, I. (2001). Seasonality and atmospheric dynamics of the teleconnection between African rainfall and tropical sea-surface temperature: Atlantic vs. ENSO. International Journal of Climatology, 21, 973–1005.Find this resource:
Claussen, M. (2009). Late quaternary vegetation-climate feedbacks. Climate of the Past, 5, 203–216.Find this resource:
Claussen, M., Kubarzki, C., Brovkin, V., Ganopolski, A., Hoelzmann, P., & Pachur, H. J. (1999). Simulation of an abrupt change in Saharan vegetation in the mid-Holocene. Geophysical Research Letters, 26, 2037–2040.Find this resource:
Cook, K. H. (2000). The South Indian Convergence Zone and interannual rainfall variability over southern Africa. Journal of Climate, 13, 3789–3804.Find this resource:
Cook, K. H. (2001). A Southern Hemisphere wave response to ENSO with implications for southern Africa precipitation. Journal of Atmospheric Science, 15, 2146–2162.Find this resource:
Cook, K. H. (2013). Climate dynamics. Princeton, NJ: Princeton University Press.Find this resource:
Cook, K. H. (2015). Role of inertial instability in the West African monsoon jump. Journal of Geophysical Research—Atmospheres, 120, 3085–3102.Find this resource:
Cook, K. H., & Vizy, E. K. (2006). Coupled model simulations of the West African monsoon system: 20th century simulations and 21st century predictions. Journal of Climate, 19, 3681–3703.Find this resource:
Cook, K. H., & Vizy, E. K. (2013). Projected changes in East African rainy seasons. Journal of Climate, 26(16), 5931–5948.Find this resource:
Cook, K. H., & Vizy, E. K. (2015). Detection and analysis of an amplified warming of the Sahara Desert. Journal of Climate, 28, 6560–6580.Find this resource:
Crétat, J., Vizy, E. K., & Cook, K. H. (2014). How well are daily intense rainfall events captured by current climate models over Africa? Climate Dynamics, 42, 2691–2711.Find this resource:
Dieppois, B., Pohl, B., Rouault, M., New, M., Lawler, D., & Keenlyside, N. (2016). Interannual to interdecadal variability of winter and summer southern Africa rainfall, and their teleconnections. Journal of Geophysical Research, 121, 6215–6239.Find this resource:
Dieppois, B., Rouault, M., & New, M. (2015). The impact of ENSO on southern Africa rainfall in CMIP5 ocean atmosphere coupled climate models. Climate Dynamics, 45, 2425–2442.Find this resource:
Dosio, A., & Panitz, H. J. (2016). Climate change projections for CORDEX-Africa with COSMO-CLM regional climate model and differences with the driving global climate models. Climate Dynamics, 46, 1599–1625.Find this resource:
Dosio, A., Panitz, H. J. Schubert-Frisius, M., & Luthi, D. (2015). Dynamical downscaling of CMIP5 global circulation models over CORDEX-Africa with COSMO-CLM: Evaluation over the present climate and analysis of the added value. Climate Dynamics, 44, 2637–2661.Find this resource:
Druyan, L. M. (2011). Studies of 21st century precipitation trends over West Africa. International Journal of Climatology, 31, 1415–1424.Find this resource:
Estes, L. D., Chaney, N. W., Herrera-Estrada, J., Sheffield, J., Caylor, K. K., & Wood, E. F. (2014). Changing water availability during the African maize-growing season. Environmental Research Letters, 9(7), Art. no. 075005.Find this resource:
FAO (1978). Report on the agro-ecological zones project. Vol. 1. Methodology and Results for Africa. World Soil Resources Report No. 48. Rome: FAO.Find this resource:
Findlater, J. (1969). A major low-level air current near the Indian Ocean during northern summer. Quarterly Journal of the Royal Meteorological Society, 95(404), 362–380.Find this resource:
Funk, C., Shukla, S., Blade, I., Liebmann, B., Roberts, J. B., & Husak, G. (2014). Predicting East African spring droughts using Pacific and Indian Ocean seas surface temperature indices. Hydrology and Earth System Sciences, 18(12), 4965–4978.Find this resource:
Funk, C., Shukla, S., Hoell, A., & Livneh, B. (2015). Assessing the contributions of East African and West Pacific warming to the 2014 boreal spring East African drought. Bulletin of the American Meteorological Society, 96(12), S77–S82.Find this resource:
Gamoyo, M., Reason, C., & Obura, D. (2015). Rainfall variability over the East African coast. Theoretical and Applied Climatology, 120, 311–322.Find this resource:
Grodsky, S. A., Carton, J. A., & Nigam, S. (2003). Near surface westerly wind jet in the Atlantic ITCZ. Geophysical Research Letters, 30(19).Find this resource:
Guo, G., & Adler, R. F. (2004). Seasonal evolution and variability associated with the west African monsoon system. Journal of Climate, 17, 3364–3377.Find this resource:
Hachigonta, S., & Reason, C. J. C. (2006). Interannual variability in dry and wet spell characteristics over Zambia. Climate Research, 32, 49–62.Find this resource:
Hagos, S. M., & Cook, K. H. (2007). Dynamics of the West African monsoon jump. Journal of Climate, 20(21), 5264–5284.Find this resource:
Hermann, S. M., & Mohr, K. I. (2011). A continental-scale classification of rainfall seasonality regimes in Africa based on gridded precipitation and land surface temperature products. Journal of Applied Meteorology and Climatology, 50, 2504–2513.Find this resource:
Holton, J. R., & Hakim, G. J. (2013). An introduction to dynamic meteorology (5th ed.). Boston: Elsevier.Find this resource:
Indeje, M., Semazzi, F. M. H., & Ogallo, L. J. (2000). ENSO signals in East African rainfall seasons. International Journal of Climatology, 20, 19–46.Find this resource:
IPCC. (2013). Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change (T.F. Stocker, D. Qin, G.-K. Plattner, M. Tignor, S.K. Allen, J. Boschung, et al. [Eds.]). Cambridge, United Kingdom and New York: Cambridge University Press.Find this resource:
Kent, C., Chadwick, R., & Rowell, D. P. (2015). Understanding uncertainties in future projections of seasonal tropical precipitation. Journal of Climate, 28, 4390–4413.Find this resource:
Kinuthia, J. H., & Asnani, G. C. (1982). A newly found jet in northern Kenya (Turkana Channel). Monthly Weather Review, 110, 1722–1728.Find this resource:
Lavaysse, C., Flamant, C, Janicot, S., Parker, D. J, Lafore, J. -P., Sultan, B., & Pelon, J. (2009). Seasonal evolution of the West African heat low: A climatological perspective. Climate Dynamics, 33(2–3), 313–330.Find this resource:
Lazenby, M. J., Todd, M. C., & Wang, Y. (2016). Climate model simulation of the South Indian Ocean convergence zone: Mean state and variability. Climate Research, 68(1), 59–71.Find this resource:
Le Barbé, L., Lebel, T., & Tapsoba, D. (2002). Rainfall variability in West Africa during the years 1950–1990. Journal of Climate, 15, 187–202.Find this resource:
Liebmann, B., Hoerling, M. P., Funk, C., Bladé, I., Dole, R. M., Allured, D., . . . Eischeid, J. K. (2014). Understanding recent eastern Horn of Africa rainfall variability and change. Journal of Climate, 27, 8630–8645.Find this resource:
Lindesay, J. A. (1988). South African rainfall, the Southern Oscillation, and a Southern Hemisphere semi-annual cycle. International Journal of Climatology, 8(1), 17–30.Find this resource:
Lott, F. C., Stott, P. A., & Cristidis, N. (2013). Can the 2011 East African drought be attributed to human-induced climate change? Geophysical Research Letters, 40(6), 1–5.Find this resource:
Lyon, B. (2014). Seasonal drought in the Greater Horn of Africa and its recent increase during the March-May long rains. Journal of Climate, 27, 7953–7975.Find this resource:
Lyon, B., & DeWitt, D. G. (2012). A recent and abrupt decline in the East African long rains. Geophysical Research Letters, 39(2), Art. No. L02702.Find this resource:
Maidment, R. I., Allan, R. P., & Black, E. (2015). Recent observed and simulated changes in precipitation over Africa. Geophysical Research Letters, 42, 8155–8164.Find this resource:
Manatsa, D., Matarira, C., Mushore, T. D., & Mudavanhu, C. (2015). Southern Africa winter temperature shifts and their link to the Southern Annular Mode. Climate Dynamics, 45, 2337–2350.Find this resource:
Manford, H. N., Kita, I., & Kitamura, Y. (2011). Effect of climate change on rainfed maize production: Assessment of maize production vs a changing rainfall pattern in Malawi. Journal of Rainwater Catchment Systems, 16(2), 25–37.Find this resource:
Mariotti, L., Diallo, I., Coppola, E., & Giorgi, F. (2014). Seasonal and intraseasonal changes of African monsoon climates in 21st century CORDEX projections. Climatic Change, 125(1), 53–65.Find this resource:
Marshall, J., & Plumb, R. A. (2008). Atmosphere, ocean, and climate dynamics: An introductory text. Boston: Elsevier.Find this resource:
Marthews, T. R., Otto, F. E., Mitchell, D., Dadson, S. J., & Jones, R. G. (2015). The 2014 drought in the Horn of Africa: Attribution of meteorological drivers. Bulletin of the American Meteorological Society, 96, S83–S88.Find this resource:
deMenocal, P., Ortiz, J., Guilderson, T., Adkins, J., Sarnthein, M., Baker, L., & Yarusinsky, M. (2000). Abrupt onset of the African Humid Period: Rapid climate responses to gradual insolation forcing. Quaternary Science Reviews, 19, 347–361.Find this resource:
Monerie, P. A., Roucou, P., & Fontaine, B. (2013). Mid-century effects of climate change on African monsoon dynamics using the A1B emission scenario. International Journal of Climatology, 33, 881–896.Find this resource:
Mtongori, H. I., Stordal, F., & Benestad, R. E. (2016). Evaluation of empirical statistical downscaling models’ skill in predicting Tanzanian rainfall and their application in providing future downscaled scenarios. Journal of Climate, 29, 3231–3252.Find this resource:
Nguyen, H., Thorncroft, C. D., & Zhang, C. (2011). Guinean coast rainfall of the West African monsoon. Quarterly Journal of the Royal Meteorological Society, 137, 1828–1840.Find this resource:
Nicholson, S. E. (2015). Long term variability of the East African “short rains” and its links to large-scale factors. International Journal of Climatology, 35, 3979–3990.Find this resource:
Nicholson, S. E. (2016). The Turkana low-level jet: Mean climatology and association with regional aridity. International Journal of Climatology, 36, 2598–2614.Find this resource:
Okumura, Y., & Xie, S. P. (2004). Interaction of the Atlantic equatorial cold tongue and the African monsoon. Journal of Climate, 17, 3589–3602.Find this resource:
O’Loughlin, J., Witmer, F. D. W., Linke, A. M., Laing, A., Gettleman, A., & Dudhia, J. (2012). Climate variability and conflict risk in East Africa, 1990–2009. Proceedings of the National Academy of Sciences, 109, 18344–18349.Find this resource:
Otieno, V. O., & Anya, R. O. (2013). CMIP5 simulated climate conditions of the Greater Horn of Africa. Part 1: Contemporary climate. Climate Dynamics, 41, 2081–2097.Find this resource:
Paeth, H., Born, K., Grimes, R., Podzun, R., & Jacob, D. (2009). Regional climate change in tropical and northern Africa due to greenhouse forcing and land use changes. Journal of Climate, 22(1), 114–132.Find this resource:
Paeth, H., Hall, N. M., Angel-Gaertner, M., Alonso, M. D., Moumouni, S., Polcher, J., . . . Rummukainen, M. (2011). Progress in regional downscaling of West African precipitation. Atmospheric Science Letters, 12, 75–82.Find this resource:
Paeth, H., & Thamm, H. P. (2007). Regional modelling of future climate north of 15 ºS including greenhouse warming and land degradation. Climatic Change, 83, 401–427.Find this resource:
Patricola, C. M., & Cook, K. H. (2007). Dynamics of the West African Monsoon under mid-Holocene precessional forcing: Regional climate model simulations. Journal of Climate, 20(4), 694–716.Find this resource:
Patricola, C. M., & Cook, K. H. (2010). Northern African climate at the end of the 21st century: Integrated application of regional and global climate models. Climate Dynamics, 35, 193–212.Find this resource:
Patricola, C. M., & Cook, K. H. (2011). Sub-Saharan Northern African climate at the end of the 21st century: Forcing factors and climate change processes. Climate Dynamics, 37(5–6), 1165–1188.Find this resource:
Pfeifroth, U., Trentmann, J., Fink, A. H., & Ahrens, B. (2016). Evaluating satellite-based diurnal cycles of precipitation in the African tropics. Journal of Applied Meteorology and Climatology, 55, 23–39.Find this resource:
Pinto, I., Lennard, C., Tadross, M., Hewitson, B., Dosio, A., Nikulin, G., . . . Shongwe, M. E. (2016). Evaluation and projections of extreme precipitation over southern Africa from two CORDEX models. Climatic Change, 135(3–4), 655–668.Find this resource:
Pu, B., & Cook, K. H. (2010). Dynamics of the West African westerly jet. Journal of Climate, 23, 6263–6276.Find this resource:
Pu, B., & Cook, K. H. (2012). Role of the West African westerly jet in Sahel rainfall variations. Journal of Climate, 25, 2880–2896.Find this resource:
Ramel, R., Gallée, H., & Messager, C. (2006). On the northward shift of the West African monsoon. Climate Dynamics, 26, 429–440.Find this resource:
Reason, C. J. C. (2002). Sensitivity of the southern Africa circulation to dipole sea surface temperature patterns in the South Indian Ocean. International Journal of Climatology, 22, 377–393.Find this resource:
Riahi, K., Rao, S., Krey, V., Cho, C., Chirkov, V., Fischer, G., . . . Rafaj, P. (2011). RCP 8.5: A scenario of comparatively high greenhouse gas emissions. Climatic Change, 109, 33.Find this resource:
Riddle, E. E., & Cook, K. H. (2008). Abrupt rainfall transitions over the Greater Horn of Africa: Observations and regional model simulations. Journal of Geophysical Research—Atmospheres, 113(D15109), 1–14.Find this resource:
Roca, R., Chambon, P., Jobard, I., & Kirstetter, P. E. (2010). Comparing satellite and surface rainfall products over West Africa at meteorologically relevant scales during the AMMA campaign using error estimates. Journal of Applied Meteorology and Climatology, 49, 715–731.Find this resource:
Ropelewski, C. F., & Halpert, M. S. (1987). Global and regional scale precipitation patterns associated with ENSO. Monthly Weather Review, 115, 1606–1626.Find this resource:
Rowell, D. P., Booth, B. B., Nicholson, S. E., & Good, P. (2015). Reconciling past and future rainfall trends over East Africa. Journal of Climate, 28, 9768–9788.Find this resource:
Sanogo, S., Fink, A. H., Omotosho, J. A., Ba, A., Redl, R., & Ermert, V. (2015) Spatiotemporal characteristics of the recent rainfall recovery in West Africa. International Journal of Climatology, 35, 4589–4605.Find this resource:
Schmocker, J., Liniger, H. P., Ngeru, J. N., Brugnara, Y., Auchmann, R., & Brönnimann, S. (2016). Trends in mean and extreme precipitation in the Mount Kenya region from observations and reanalysis. International Journal of Climatology, 36, 1500–1514.Find this resource:
Shongwe, M. E., Lennard, C., Liebmann, B., Kalognomou, E. A., Ntsangwane, L., & Pinto, I. (2015). An evaluation of CORDEX regional climate models in simulating precipitation over southern Africa. Atmospheric Science Letters, 16, 199–207.Find this resource:
Sijikumar, S., Roucou, P., & Fontaine, B. (2006). Monsoon onset over Sudan-Sahel: Simulation by the regional scale model MM5. Geophysical Research Letters, 33, L03814.Find this resource:
Skinner, C. B., Ashfaq, M., & Diffenbaugh, N. S. (2012). Influence of twenty-first-century atmospheric and sea surface temperature forcing on West African climate. Journal of Climate, 25, 527–542.Find this resource:
Sultan, B., & Janicot, S. (2000). Abrupt shift of the ITCZ over West Africa and intra-seasonal variability. Geophysical Research Letters, 27, 3353–3356.Find this resource:
Sutton, R. T., Dong, B., & Gregory, J. M. (2007). Land/sea warming ratio in response to climate change: IPCC AR4 model results and comparison with observations. Geophysical Research Letters, 34.Find this resource:
Sylla, M. B., Giorgi, F., Pal, J. S., Gibba, P., Kebe, I., & Nikiema, M. (2015). Projected changes in the annual cycle of high-intensity precipitation events over West Africa for the late twenty-first century. Journal of Climate, 28, 6475–6488.Find this resource:
Tadross, M. A., Hewitson, B. C., & Usman, M. T. (2005). The interannual variability of the onset of the maize growing season over South Africa and Zimbabwe. Journal of Climate, 18, 3356–3372.Find this resource:
Tadross M. A., Suarez, P., Lotsch, A., Hachigonta, S., Mdoka, M., Unganai, L., . . . Muchinda, M. (2009). Growing-season rainfall and scenarios of future change in southeast Africa: Implications for cultivating maize. Climate Research, 40(2–3), 147–161.Find this resource:
Thorncroft, C. D., Nguyen, H., Zhang, C., & Payrille, P. (2011). Annual cycle of the West African monsoon: Regional circulations and associated water vapour transport. Quarterly Journal of the Royal Meteorological Society, 137(654), 129–147.Find this resource:
Timm, O., Joehler, P., Timmermann, A., & Menviel, L. (2010). Mechanisms for the onset of the Africa Humid Period and Sahara greening 14.5 ka-11ka BP. Journal of Climate, 23, 2612–2633.Find this resource:
Usman, T., & Reason, C. J. C. (2004). Dry spell frequencies and their variability over southern Africa. Climate Research, 26(3), 199–211.Find this resource:
Vallis, G. K. (2006). Atmospheric and oceanic fluid dynamics. Cambridge, U.K.: Cambridge University Press.Find this resource:
Vigaud, N., Richard, Y., Rouault, M., & Fauchereau, N. (2009). Moisture transport between the South Atlantic Ocean and southern Africa: Relationships with summer rainfall and associated dynamics. Climate Dynamics, 32, 113–123.Find this resource:
Vizy, E. K., & Cook, K. H. (2003). Connections between the summer East African and Indian rainfall regimes. Journal of Geophysical Research, 108(D16).Find this resource:
Vizy, E. K., & Cook, K. H. (2016). Seasonality of the observed amplified Sahara warming trend and implications for Sahel rainfall. Journal of Climate.Find this resource:
Vizy, E. K., Cook, K. H., Chimphamba, J., & McCusker, B. (2015). Projected changes in Malawi’s growing season. Climate Dynamics, 45, 1673–1698.Find this resource:
Vizy, E. K., Cook, K. H., Crétat, J., & Neupane, N. (2013). Projections of a wetter Sahel in the 21st century from global and regional models. Journal of Climate, 26, 4664–4687.Find this resource:
Washington, R., & Preston, A. (2006). Extreme wet years over southern Africa: Role of Indian Ocean sea surface temperatures. Journal of Geophysical Research—Atmospheres, 111, Art. No. D15104.Find this resource:
Yang, W. C., Seager, R., Cane, M. A., & Lyon, B. (2015a). The annual cycle of East African precipitation. Journal of Climate, 28, 2385–2404.Find this resource:
Yang, W. C., Seager, R., Cane, M. A., & Lyon, B. (2015b). The rainfall annual cycle bias over East Africa in CMIP5 coupled climate models. Journal of Climate, 28, 9789–9802.Find this resource: