Projected changes in African easterly wave intensity and track in response to greenhouse forcing

Edited by Kerry A. Emanuel, Massachusetts Institute of Technology, Cambridge, MA, and approved April 1, 2014 (received for review October 17, 2013)
April 28, 2014
111 (19) 6882-6887

Significance

African easterly waves (AEWs) are ubiquitous weather disturbances over the tropical Atlantic basin during the summer season. However, despite the critical influence of AEWs on Sahel rainfall, transport of Saharan dust, and initiation of Atlantic hurricanes, little is known about how global warming influences AEW activity. We use an ensemble of general circulation models to investigate the response of AEWs to elevated greenhouse gas concentrations. We find that increases in regional temperature gradients and the strength of convergence and uplift along the Intertropical Front of Africa lead to increases in the strength of AEWs. Given the relationships that are observed in the current climate, changes in AEW strength could influence climate over both West Africa and the larger Atlantic basin.

Abstract

Synoptic-scale African easterly waves (AEWs) impact weather throughout the greater Atlantic basin. Over the African continent, AEWs are instrumental in initiating and organizing precipitation in the drought-vulnerable Sahel region. AEWs also serve as the precursors to the most intense Atlantic hurricanes, and contribute to the global transport of Saharan dust. Given the relevance of AEWs for the climate of the greater Atlantic basin, we investigate the response of AEWs to increasing greenhouse gas concentrations. Using an ensemble of general circulation models, we find a robust increase in the strength of the winds associated with AEWs along the Intertropical Front in West Africa by the late 21st century of the representative concentration pathway 8.5. AEW energy increases directly due to an increase in baroclinicity associated with an enhanced meridional temperature gradient between the Sahara and Guinea Coast. Further, the pattern of low-level warming supports AEW development by enhancing monsoon flow, resulting in greater convergence and uplift along the Intertropical Front. These changes in energetics result in robust increases in the occurrence of conditions that currently produce AEWs. Given relationships observed in the current climate, such changes in the location of AEW tracks and the magnitude of AEW winds carry implications for the relationship between AEWs and precipitation in the Sahel, the mobilization of Saharan dust, and the likelihood of cyclogenesis in the Atlantic. Our results therefore suggest that changes in AEW characteristics could play a critical role in shaping the response of Atlantic basin climate to future increases in greenhouse gas concentrations.
African easterly waves (AEWs) are westward propagating synoptic-scale disturbances that develop over northern Africa during the boreal summer season. AEWs are instrumental in organizing and initiating convective systems in the semiarid Sahel region of Africa (1, 2). AEWs have also been found to play a role in mobilizing and transporting Saharan dust within Africa and globally (3, 4), and in initiating tropical cyclones in the Atlantic basin, including the most intense Atlantic hurricanes (57). However, despite the crucial influence of AEWs on the climate of West Africa and the greater Atlantic basin, relatively little work has focused on the potential response of AEW tracks to enhanced radiative forcing (8). We therefore investigate the response of AEW vortices to increasing greenhouse gas concentrations, with the goal of providing insight into the atmospheric processes that are likely to influence the response of West African precipitation (9, 10), Saharan dust transport (11), and Atlantic tropical cyclones (12, 13).
AEWs are convectively triggered in East and Central Africa, and grow through a mixed barotropic−baroclinic mechanism associated with the midtroposphere [600 mbar (mb)] African Easterly Jet (AEJ) (1416), which largely controls the structure and wavelength of the AEWs (17, 18). AEWs typically have periodicities of 3–5 d, wavelengths between 2,000 km and 4,000 km, and two circulation centers—one to the north of the AEJ and one to the south (19). Despite often being part of the same AEW structure (19, 20), the north and south circulation centers are commonly treated as separate AEW events (6).
Consistent with the Charney and Stern (21) instability criterion, the circulation south of the AEJ develops in association with the reversed potential vorticity gradient near the core of the AEJ (14). The systems south of the AEJ are driven by barotropic and baroclinic energy conversions, and exhibit strongest circulation over West Africa near 700 mb and ∼11°N. North of the AEJ, the circulation center grows as a result of the interaction between the negative potential vorticity anomaly in the AEJ core and the positive surface potential temperature gradients below and north of the AEJ (19, 22). The northern systems grow primarily through baroclinic energy conversions, and reach maximum amplitude over West Africa around 850 mb in the low static stability atmosphere along the Intertropical Front, near ∼20°N (23, 24). Latent heating associated with convection also provides an important energy source for the growth and maintenance of AEWs, and can modify the structure of both the northern and southern vortices (20, 25).
Because the northern and southern AEW vortices track westward at different latitudes, they impact climate over the greater Atlantic basin in different ways. The southern circulations propagate near the Intertropical Convergence Zone (ITCZ) and are often associated with moist convection (20). They therefore exert a greater influence on precipitation variability in West Africa, and are more efficient at undergoing tropical cyclogenesis in the Main Development Region (MDR) (26) of the Atlantic (2729). Although the northern vortices are primarily characterized by dry convective processes (30), they too are capable of influencing precipitation variability in the Sahel through the advection of moisture within the southerly flow of the vortex (31). In addition, because the northern vortices track near the southern border of the Sahara, they are an important source of dust mobilization and transport over West Africa and the Atlantic (4). However, compared with the southern vortices, the northern vortices are less efficient at undergoing tropical cyclogenesis. As the northern vortices propagate westward into the Atlantic, they often advect relatively dry Saharan air, known as the Saharan Air Layer, downstream of the wave trough (7). As a result of the dry air advection, northern-track vortices that do transition to tropical cyclones often do not undergo tropical cyclogenesis until they reach the Western Atlantic, where environmental conditions are more favorable for development (28).
It is important to note that variability in AEW activity is not necessarily associated with variability in Atlantic tropical cyclones. For example, at the interannual timescale, variability in the number of AEWs leaving the West African coast is uncorrelated with the variability in the number of tropical cyclones in the Atlantic basin (32). The number and characteristics of AEWs comprise only a subset of the many factors that influence the likelihood of tropical cyclogenesis, including vertical wind shear, sea surface temperatures, and tropospheric humidity (33, 34). However, tracking AEW vortices and their characteristics provides valuable information for tropical cyclone forecasting, and understanding the response of AEW activity to the north and south of the AEJ could likewise indicate whether AEW changes are likely to be relevant for the tropical cyclone response to global warming.
Although the roles that AEWs play in shaping precipitation, dust emissions, and tropical cyclone tracks are complex, improved understanding of the response of AEWs to elevated greenhouse forcing has the potential to provide valuable insight into the mechanisms by which continued global warming could influence the climate of West Africa and the greater Atlantic basin.

Data and Methods

We compare climate simulations of the recent past (1980−2005) and the late 21st century (2075−2100) using a single, transient realization from each of 17 general circulation models (GCMs) archived in the Coupled Model Intercomparison Project Phase 5 (CMIP5) dataset (35) (Table S1). (Analysis of AEWs requires daily 3D atmospheric fields, the availability of which limits the number of models that can be included in the ensemble analysis.) Observed changes in atmospheric composition that are consistent with natural and anthropogenic sources are prescribed during the “historical” period of each simulation (1850−2005). Beginning in the simulation year 2006, each model is prescribed the time-evolving concentrations of atmospheric constituents in the representative concentration pathway 8.5 (RCP8.5). RCP8.5 exhibits prescribed radiative forcing of 8.5 W·m−2 and concentrations near 1370 ppm CO2-equivalent by the year 2100 (36). Global carbon dioxide emissions over the past decade have tracked near the RCP8.5 values (37).
We use the European Centre for Medium-Range Weather Forecasts (ECMWF) Interim Reanalysis (ERA-Interim) (38) to evaluate the simulation of AEWs in the CMIP5 models. ERA-Interim data extend from 1979 to the present day. Our analysis makes use of the daily-scale 3D atmospheric variables archived at 1.5° resolution on the ECMWF data server. To enable comparison between the different CMIP5 models and ERA-Interim, we interpolate each of the datasets to a common 1° × 1° grid using a patch recovery method (39).
We visualize the multimodel data based on the methodology of Tebaldi et al. (40). This technique identifies model agreement based on both the sign of the differences between the present and future periods and the statistical significance of the differences. For a given variable, we first use Student t test to identify the locations of statistically significant differences (at the 95% confidence level) between the future and present time periods for each model. We then map the multimodel mean difference between the future and present periods. Grid boxes are shaded with the color corresponding to the value of the multimodel mean change, and are overlain with black stippling, where (i) greater than 50% of the models project a statistically significant change and (ii) greater than 80% of those models agree on the sign of the change. Grid boxes are shaded white where greater than 50% of the models exhibit statistically significant changes but less than 80% of those models agree on the sign of change. Grid boxes are shaded only with the color corresponding to the value of the multimodel mean difference (with no stippling added) where less than 50% of the models project a statistically significant change. (In addition to Student t test, we test the robustness of the results using the Permutation test, which is a nonparametric method of significance testing. The Permutation test assesses the significance of mean changes between the future and present time periods for each model without making any assumptions about the underlying distribution of the data.)
We detect AEWs in the reanalysis and CMIP5 models using the meridional component of the wind at 850 mb and 700 mb (16, 19). To isolate variations in the flow associated with the passing of AEWs, we filter the daily time series of meridional wind between June and September (JJAS) of each year using a 2- to 6-d band-pass filter (41). We then calculate the variance of the filtered time series at each grid point for each season to attain a proxy of mean seasonal AEW activity. This approach is advantageous for detecting AEWs within long-term climate time series because it is computationally efficient, and yields a clear, coherent spatial pattern of mean AEW activity over Africa (16). (We test the robustness of the results to the width of the band-pass filter by comparing the results from the 2- to 6-d band-pass filter with those from a 2- to 10-d band-pass filter.)
To identify the number of AEWs in each JJAS season, we choose a reference point that is north of the ensemble mean AEJ core in both the historical and RCP8.5 periods (18°N, 0°E; Fig. 1). We count the number of instances in which the filtered daily mean meridional wind at the 850-mb location exceeds 1 m⋅s−1 (31, 42). Each of these instances is considered a separate AEW event. When the filtered meridional wind exceeds 1 m⋅s−1 on consecutive days, we only count the day with the largest meridional wind value to ensure that we do not double-count single AEW events. To calculate changes in the average number of intense AEWs, we calculate a distribution of AEW event intensity based on the magnitude of the 850-mb meridional wind associated with each AEW at (18°N, 0°E) during the 1980–2005 period. For each model, we define an AEW as being intense if the meridional wind exceeds the mean-plus-one-standard-deviation threshold from the historical period distribution. Of these, we define an AEW as being extremely intense if the meridional wind exceeds the mean-plus-two-standard-deviations threshold from the historical period distribution. We then calculate the difference in the number of AEWs that exceed the intensity thresholds in each model’s historical and future periods.
Fig. 1.
CMIP5 multimodel mean change in the JJAS variability of the 2- to 6-d filtered meridional wind at (A) 850 mb and (B) 700 mb. Mean change calculated between 2075−2100 and 1980−2005. The white contour in each plot marks the 10 m·s−1 zonal wind contour at 600 mb during 2075−2100 and is indicative of the location of the AEJ core. Black dots indicate areas of robust change. See Data and Methods for explanation of the multimodel agreement methodology. Areas where topography intersects the 850 mb pressure level are masked with gray shading in A.
For consistency with previous work on AEWs, we refer to the northern and southern circulations as northern-track AEWs and southern-track AEWs, respectively.

Results and Discussion

Nearly all models agree that AEWs along the Sahel−Sahara border intensify in response to enhanced radiative forcing in the RCP8.5 pathway (Fig. 1 and Figs. S1, S2C, and S3). At 850 mb, the maximum change is located between 15°N to 25°N and 15°E to 25°W, where, on average, the models exhibit a 25% increase in the JJAS mean variability of the filtered meridional wind (Fig. 1A). Similar, but weaker, increases in meridional wind variance are found in the same region at 700 mb (Fig. 1B). This region of increased AEW activity is located within and just north of the region of maximum AEW activity seen in the historical period multimodel mean (Fig. 2B), and north of the projected future (2075−2100) location of the AEJ core.
Fig. 2.
JJAS mean variability of the 2- to 6-d filtered meridional wind at (A and B) 850 mb and (C and D) 700 mb (shaded). Mean averaged over 1980−2005 from (A and C) ERA-Interim Reanalysis and (B and D) the CMIP5 multimodel ensemble. The white contour line in A and B indicates the 10 m·s−1 zonal wind contour at 600 mb during 1980−2005 from (A) ERA-Interim, and (B) the CMIP5 multimodel mean, and is indicative of the location of the AEJ core. Contours of the variability in the meridional wind variance at 850 mb from ERA-Interim are overlain on top of C. Contours of the variability in the meridional wind variance at 850 mb from CMIP5 are overlain on top of D. The white contours in C and D are plotted at 1 m2⋅s−2 intervals beginning at 5 m2⋅s−2. Areas where topography intersects the 850-mb and 700-mb pressure levels are masked with gray shading.
South of 15°N, on the cyclonic shear side of the AEJ, the CMIP5 multimodel ensemble does not exhibit robust changes in AEW activity at 850 mb or 700 mb (Fig. 1). The response of AEWs in this southern region is likely complicated by the response of moist convective processes to enhanced greenhouse forcing in the RCP8.5 pathway. Indeed, across the CMIP5 ensemble, some models simulate increases in filtered meridional wind variance in this southern region, while other models simulate decreases (Fig. 1 and Fig. S1). The lack of agreement on the sign or statistical significance of changes in AEW activity beneath and south of the AEJ at both 850 mb and 700 mb suggests a lower likelihood of change in AEWs along the southern Sahel and Guinea Coast region of West Africa.
One question in evaluating the projected changes in wind variance is whether the 2- to 6-d filter is exclusively capturing AEW activity. Specifically, given the relatively high latitude (15°N−25°N) of the projected changes in filtered meridional wind variance, it is possible that an increase in midlatitude trough activity could be contributing to the robust signal (43). However, given the projected decrease in storm activity along the equatorward side of the midlatitude storm track in the Northern Hemisphere, and the projected poleward shift in the location of storm tracks in the Northern Hemisphere (44), we expect the contribution from an increase in summertime midlatitude troughs over West Africa to be small. Indeed, our analysis shows that the variance in JJAS 2- to 6-d filtered meridional wind at 250 mb, a proxy for midlatitude storm track activity (44), is reduced over the subtropical Atlantic, southern Europe, and far northern Africa, and increased north of 50°N, suggestive of a northward shift in the mean location of midlatitude troughs (Fig. S4A). A similar pattern of robust changes in wind variance is seen at 850 mb (Fig. S4B). While these results do not completely rule out a possible contribution from midlatitude troughs, they do require that increases in the contribution of midlatitude troughs to the wind variability over West Africa would have to occur within the context of a northward shift in the midlatitude storm tracks. We therefore conclude that the majority of the increased variability in meridional winds seen in CMIP5 over West Africa is most likely driven by changes in easterly wave activity.
A second question in evaluating the CMIP5 results is whether the models are able to accurately simulate the processes that are responsible for AEW activity in the current climate. The CMIP5 models exhibit considerable spread in the simulation of meridional wind variability at both 850 mb and 700 mb during the 1980–2005 period (Figs. S5 and S6). The location of the northern track of AEWs is clearly visible at 850 mb between 15°N and 25°N in the ERA-Interim Reanalysis, and there is also evidence of the southern track of AEWs at 850 mb near 10°N between 5°W and 15°W (Fig. 2A). The CMIP5 mean correctly captures the location and magnitude of the center of maximum 850-mb wind variance near 15°N at the West African coast, and the eastern edge of the maximum 850-mb wind variance near 10°E (Fig. 2B). In addition, the spatial pattern of 700-mb AEW activity in the CMIP5 multimodel mean is very similar to that of the ERA-Interim Reanalysis, with highest meridional wind variance in a zonal band that stretches west from the Ethiopian Highlands (∼30°E) to the West African coast along ∼10°N, and peaks near the coast between 10°N and 15°N (Fig. 2 C and D).
Of greatest concern are the observations that (i) there is not clear separation between the northern and southern track AEW activity at 850 mb in many of the models (Fig. S5) nor in the multimodel mean (Fig. 2B), and (ii) the distinction between northern track activity at 850 mb and southern track activity at 700 mb is less pronounced in the multimodel mean than in the reanalysis (Fig. 2 C and D). This lack of distinction between the northern and southern AEW tracks is likely the result of the relatively coarse horizontal resolution of the models, and the fact that some of the models simulate an excessively strong connection between AEW circulation and convection, which results in excessively strong AEW activity within the ITCZ over West Africa (16). However, some models correctly simulate the majority of 850-mb activity to be north of the AEJ (Fig. S5), and some correctly simulate the maximum in 700-mb activity to be south of the maximum in 850-mb activity, suggestive of vertically deeper AEWs in the southern Sahel and Guinea Coast region of West Africa (Fig. 2D and Figs. S5 and S6). Overall, although some models fail to fully simulate the distinction between the northern and southern AEW tracks in the present climate, the projected increase in AEW activity is clearly weighted toward higher latitudes over West Africa (including increases north of the AEJ in 15 of the 17 models) (Fig. S1). Although the placement of enhanced meridional wind variance north of the AEJ suggests that the increased AEW activity is associated with changes in northern track AEWs, we refrain from explicitly calling the robust change an increase in northern-track AEW activity due to the fact that some of the models do not simulate distinct southern- and northern-track AEW circulations.
Given the robustness of the response in AEW activity north of the AEJ, and lack of a robust response south of the AEJ, we focus our attention on understanding the causes of the changes north of the AEJ. Because AEWs north of the AEJ derive the majority of their energy from baroclinic processes associated with the interaction of the low-level meridional temperature gradients in the Sahel−Sahara region and the potential vorticity gradients in the AEJ core, the energetics of AEWs should in part reflect the local changes to surface air temperature that occur in response to enhanced radiative forcing. Stronger gradients in potential temperature result in greater zonal available potential energy for AEW growth (17, 24, 30). AEWs access this source of zonal available potential energy by fluxing temperature down the meridional temperature gradient, creating eddy available potential energy. AEWs then grow by converting the eddy available potential energy to eddy kinetic energy through baroclinic overturning. To enable sustained AEW growth, the eddy available potential energy must be replenished by the zonal available potential energy (17, 30).
Warming in subtropical northern Africa exceeds warming in the moist equatorial regions of Africa during the JJAS season by the end of the 21st century, with mean surface potential temperatures increasing by 5.5 °C to 6.5 °C north of 20°N in the Sahara, and 3 °C to 4 °C over the Guinea Coast region south of 10°N (Fig. 3A). Both remote and local processes drive this pattern of surface warming. By the end of the 21st century, surface temperatures over southern Europe and the Mediterranean, including far northern Africa, are projected to warm in response to enhanced midtroposphere anticyclonic circulation, large-scale subsidence, and reduced precipitation (45). These changes result in robust increases in surface shortwave downwelling radiation over much of Europe, the Mediterranean Sea, and far northern Africa (Fig. 3C). Additionally, the Sahara warms preferentially in response to an anomalous increase in surface longwave downwelling radiation (Fig. 3D). Robust increases in precipitation also serve to minimize low-level warming throughout the central Sahel (as is expected through increased cloud clover and surface evaporative cooling), further enhancing the contrast in temperature between the Sahara and regions to the south (Fig. 3B).
Fig. 3.
JJAS CMIP5 multimodel mean change in (A) 2-m potential temperature (dashed contours), (B) precipitation, (C) surface (Sfc) downwelling shortwave radiation, and (D) surface downwelling longwave radiation. Mean change calculated between 2075−2100 and 1980−2005. Colored contours in A represent the 1980−2005 JJAS mean 2-m potential temperature from CMIP5. Multimodel agreement methodology is the same as in Fig. 1.
It should be noted that many of the CMIP5 models exhibit a warm bias in the coastal upwelling region of the Gulf of Guinea in the historical period (46, 47), which may influence the projected future temperature response to greenhouse forcing. Specifically, in many of the models, the anomalously warm waters in the Gulf of Guinea result in a southward-displaced ITCZ, and anomalously low precipitation over the Sahel (48). It is possible that model biases in the Gulf of Guinea influence the projected robust increase in precipitation and corresponding reduced surface warming over the central Sahel in the RCP8.5 simulations.
The resulting pattern of surface warming enhances the regional poleward meridional temperature gradient, particularly over the Sahel between 10°N and 25°N, thereby increasing the zonal available potential energy (Fig. 3A). We find that the conversion of zonal available potential energy to eddy available potential energy associated with AEW development also increases as a result of this enhanced regional poleward meridional temperature gradient (Fig. 4A). The CMIP5 models correctly simulate the region of negative mean covariance of the 2- to 6-d filtered 850-mb meridional wind and temperature (“t-v covariance”) over West Africa in the historical period (Fig. S7). This negative t-v covariance implies a conversion of zonal available potential energy to eddy available potential energy. By the end of the 21st century, t-v covariance becomes more negative in the region of increased AEW activity, indicating a stronger flux of temperature down the temperature gradient, and greater eddy available potential energy for AEW growth (Fig. 4A). We therefore conclude that the increased meridional temperature gradient over West Africa is at least partially responsible for providing the necessary energy for the stronger AEW development that occurs in response to increasing greenhouse forcing (Fig. 1).
Fig. 4.
CMIP5 multimodel mean change between 2075−2100 and 1980−2005 in (A) 850-mb 2- to 6-d filtered temperature and meridional wind variance (t-v covariance), (B) sea level pressure and 925-mb winds, and (C) omega. Areas where topography intersects the 850-mb pressure level are masked with gray shading in A. Wind vectors with magnitudes below 0.25 m·s−1 are masked out in B. The mean omega change in the latitude−height cross section in C is averaged between 15°W and 15°E. Multimodel agreement methodology is the same as in Fig. 1. White crosses in B are equivalent to white grid box shading in the mapview plots.
In addition to increased zonal available potential energy, the projected low-level heating over Africa results in a deepening of the Saharan Heat Low (SHL), strengthened monsoon flow, and increased convergence along the Intertropical Front (Fig. 4B). The strengthened SHL and warmer surface air temperatures maintain the low static stability that is characteristic of the region. Low static stability allows for greater interaction between the surface temperature gradients and the AEJ-level potential vorticity gradients, resulting in stronger low-level AEWs north of the AEJ (19, 23, 24). The pressure-driven increase in low-level westerly and southwesterly flow (Fig. 4B) yields greater convergence between the monsoon and Harmattan winds, and results in stronger vertical ascent along the Intertropical Front near 20°N over West Africa (15°W to 15°E) (Fig. 4C). The CMIP5 models correctly simulate a vertically deep region of ascent associated with the ITCZ near 10°N and a shallow region of ascent associated with the Intertropical Front near 17°N during the historical period, which suggests that the vertically deep increase in ascent near 20°N is not simply an enhancement of model biases in the simulation of the large-scale overturning circulations in West Africa (Fig. S8). Large-scale vertical ascent along the Intertropical Front contributes to the projected increases in AEW strength through vortex stretching (6, 30).
Additionally, the projected robust increase in deep vertical ascent along the Intertropical Front suggests the presence of increased latent heating (Fig. 4C). An increase in latent heating is consistent with the robust increases in moist monsoon flow and precipitation throughout the Sahel (with the exception of the drying over the far western Sahel) (Figs. 3B and 4B). Although AEWs north of the AEJ are generally characterized by dry dynamics in the current climate (30), the projected increase in moist convective processes in the Sahel suggests an increased role of moist convective processes in AEW development along the Sahel-Sahara border in RCP8.5. Given the importance of convectively driven diabatic heating in generating eddy available potential energy for AEW growth (20, 25), it is possible that the increased moisture in the region within and north of the AEJ also plays a critical role in enhancing AEW energy.
Lastly, consistent with increased AEW activity, many CMIP5 models project a slight strengthening of the AEJ over West Africa (Fig. S9). Although there is no consistent movement in the location of the jet across the CMIP5 models, nearly all models project increased low-level westerly flow beneath the AEJ, particularly on the jet’s poleward side (Fig. S9). The resultant enhancement in vertical zonal wind shear may support the growth of AEWs (49), helping to explain the robust increases in AEW activity that occur north of the AEJ core in response to elevated forcing (Fig. 1).
Together, the dynamic and thermodynamic responses to the pattern of low-level warming provide an environment conducive to stronger AEWs north of the AEJ core over West Africa. Indeed, nearly all models simulate an increase in the occurrence of intense and extremely intense AEWs along the Sahel−Sahara border (Fig. 5). This includes median changes of +2.0 events per season (and +39%) for intense AEWs, and +0.8 events per season (and +72%) for extremely intense AEWs (Fig. 5). The increased AEW activity along the Sahel−Sahara border (Fig. 1) has important implications for future Saharan dust transport and Atlantic tropical cyclone development. For example, the region of greatest projected increase in AEW activity lies over the prolific dust sources within the western Sahara (Fig. 1) (4). In addition, AEWs characterized by stronger low-level (850 mb) circulations and high relative humidity at the West African coast are more likely than weaker AEWs to undergo tropical cyclogenesis (7). However, strong winds associated with southern- and northern-track AEW vortices can advect dry Saharan air downstream of the AEW and impede development into a tropical cyclone entirely, or can delay tropical cyclogenesis until the wave reaches favorable environmental conditions farther west in the Atlantic basin (7, 28). Further, although there is much evidence that northern-track waves can influence tropical cyclogenesis, in the present climate, many of the northern-track waves weaken as they reach the eastern Atlantic, or track north of the MDR (7, 27). The impact of increased frequency of intense AEWs north of the AEJ on tropical cyclone development is therefore currently unclear (7).
Fig. 5.
Boxplots of the change between 2075−2100 and 1980−2005 in the number (A) and percentage (B) of AEWs per JJAS that exceed the historical period (1980−2005) one-SD and two-SD intensity thresholds at a reference location of (18°N, 0). Intensity thresholds are defined based on the magnitude of the AEW meridional wind (see Data and Methods).

Summary and Conclusions

We use the CMIP5 ensemble of GCMs to quantify the response of AEWs to increased radiative forcing. Despite considerable spread in the simulation of AEWs between CMIP5 models during the historical period (1980−2005), nearly all of the models project statistically significant increases in AEW strength north of the AEJ along the Sahel−Sahara border (15°N to 25°N) by the late 21st century in RCP8.5 (2075−2100) (Fig. 1 and Fig. S1). The energy associated with AEWs increases in response to differential heating between the Guinea Coast and the Sahara, which enhances baroclinicity and drives stronger convergence, convection, and vertical ascent along the Intertropical Front. AEWs that track south of the AEJ in the rainy zone of the ITCZ do not exhibit a robust response to enhanced greenhouse forcing across the CMIP5 ensemble. The lack of a robust signal in AEW activity south of the AEJ reflects the lack of model agreement in the sign of projected precipitation change in this region of West Africa.
The projected changes in AEWs have important implications for the climate of the Atlantic basin. Stronger AEWs along the Sahel−Sahara border have the potential to increase the transport of moisture northward into the semiarid Sahel region (31). The prospect of increased precipitation over the Sahel has important implications for the drought-vulnerable human populations in the region. In addition, increased precipitation combined with the stronger low-level winds and stronger vertical motion associated with the AEWs can increase dust mobilization and transport over Africa and the broader Atlantic basin (50), influencing local radiative balance, cloud and convective processes, and biogeochemical cycles (51, 52). Relationships between AEWs and tropical cyclones observed in the current climate also carry implications for how tropical cyclone tracks could respond to elevated greenhouse forcing. The robust response of AEWs seen in our results thereby offers potential to substantially improve understanding of how continued global warming is likely to affect the climate of Africa and the greater Atlantic basin.

Acknowledgments

We thank two anonymous reviewers and the Member Editor for their insightful and constructive comments. We wish to thank Dr. Jen-Shan Hsieh and Deepti Singh for helpful discussions. We acknowledge the World Climate Research Programme and the climate modeling groups for making available their model output, and the U.S. Department of Energy’s Program for Climate Model Diagnosis and Intercomparison for coordinating and supporting database development. Our work was supported by National Science Foundation Award 0955283. Computational resources for data processing and analysis were provided by the School of Earth Sciences’ Center for Computational Earth and Environmental Science at Stanford University.

Supporting Information

Supporting Information (PDF)
Supporting Information

References

1
TN Carlson, Some remarks on African disturbances and their progress over the Tropical Atlantic. Mon Weather Rev 97, 716–726 (1969).
2
A Mekonnen, CD Thorncroft, AR Aiyyer, Analysis of convection and its association with African easterly waves. J Clim 19, 5405–5421 (2006).
3
C Jones, N Mahowald, C Luo, The role of easterly waves on African desert dust transport. J Clim 16, 3617–3628 (2003).
4
P Knippertz, MC Todd, The central west Saharan dust hot spot and its relation to African easterly waves and extratropical disturbances. J Geophys Res 115, D12117 (2010).
5
CW Landsea, A climatology of intense (or major) Atlantic hurricanes. Mon Weather Rev 121, 1703–1713 (1993).
6
T-C Chen, Characteristics of African easterly waves depicted by ECMWF reanalyses for 1991–2000. Mon Weather Rev 134, 3539–3566 (2006).
7
SB Hopsch, CD Thorncroft, KR Tyle, Analysis of African easterly wave structures and their role in influencing tropical cyclogenesis. Mon Weather Rev 138, 1399–1419 (2010).
8
LM Druyan, P Lonergan, T Eichler, A GCM investigation of global warming impacts relevant to tropical cyclone genesis. Int J Climatol 19, 607–617 (1999).
9
IM Held, TL Delworth, J Lu, KL Findell, TR Knutson, Simulation of Sahel drought in the 20th and 21st centuries. Proc Natl Acad Sci USA 102, 17891–17896 (2005).
10
M Biasutti, A Giannini, Robust Sahel drying in response to late 20th century forcings. Geophys Res Lett 33, L11706 (2006).
11
NM Mahowald, C Luo, A less dusty future? Geophys Res Lett 30, 1903 (2003).
12
GA Vecchi, BJ Soden, Effect of remote sea surface temperature change on tropical cyclone potential intensity. Nature 450, 1066–1070 (2007).
13
KA Emanuel, Downscaling CMIP5 climate models shows increased tropical cyclone activity over the 21st century. Proc Natl Acad Sci USA 110, 12219–12224 (2013).
14
RW Burpee, The origin and structure of easterly waves in the lower troposphere of North Africa. J Atmos Sci 29, 77–90 (1972).
15
NMJ Hall, GN Kiladis, CD Thorncroft, Three-dimensional structure and dynamics of African easterly waves. Part II: Dynamical modes. J Atmos Sci 63, 2231–2245 (2006).
16
CB Skinner, NS Diffenbaugh, The contribution of African easterly waves to monsoon precipitation in the CMIP3 ensemble. J Geophys Res Atmos 118, 3590–3609 (2013).
17
DC Norquist, EE Recker, RJ Reed, The energetics of African wave disturbances as observed during Phase III of GATE. Mon Weather Rev 105, 334–342 (1977).
18
J-S Hsieh, KH Cook, On the instability of the African easterly jet and the generation of African waves: Reversals of the potential vorticity gradient. J Atmos Sci 65, 2130–2151 (2008).
19
I Pytharoulis, C Thorncroft, The low-level structure of African easterly waves in 1995. Mon Weather Rev 127, 2266–2280 (1999).
20
RJ Cornforth, BJ Hoskins, CD Thorncroft, The impact of moist processes on the African easterly jet-African easterly wave system. Q J R Meteorol Soc 135, 894–913 (2009).
21
JG Charney, ME Stern, On the stability of internal baroclinic jets in a rotating atmosphere. J Atmos Sci 19, 159–172 (1962).
22
CD Thorncroft, BJ Hoskins, An idealized study of African easterly waves. I: A linear view. Q J R Meteorol Soc 120, 953–982 (1994).
23
C-B Chang, Impact of desert environment on the genesis of African wave disturbances. J Atmos Sci 50, 2137–2145 (1993).
24
CD Thorncroft, An idealized study of African easterly waves. III: More realistic basic states. Q J R Meteorol Soc 121, 1589–1614 (1995).
25
GJ Berry, CD Thorncroft, African easterly wave dynamics in a mesoscale numerical model: The upscale role of convection. J Atmos Sci 69, 1267–1283 (2012).
26
SB Goldenberg, LJ Shapiro, Physical mechanisms for the association of El Niño and West African rainfall with Atlantic major hurricane activity. J Clim 9, 1169–1187 (1996).
27
C Thorncroft, K Hodges, African easterly wave variability and its relationship to Atlantic tropical cyclone activity. J Clim 14, 1166–1179 (2001).
28
T-C Chen, S-Y Wang, AJ Clark, North Atlantic hurricanes contributed by African easterly waves north and south of the African easterly jet. J Clim 21, 6767–6776 (2008).
29
AS Daloz, et al., The ability of general circulation models to simulate tropical cyclones and their precursors over the North Atlantic main development region. Clim Dyn 39, 1559–1576 (2012).
30
K-H Lau, N-C Lau, The energetics and propagation dynamics of tropical summertime synoptic-scale disturbances. Mon Weather Rev 120, 2523–2539 (1992).
31
G Gu, African easterly waves and their association with precipitation. J Geophys Res 109, D04101 (2004).
32
SB Hopsch, CD Thorncroft, K Hodges, A Aiyyer, West African storm tracks and their relationship to Atlantic tropical cyclones. J Clim 20, 2468–2483 (2007).
33
WM Gray, Hurricanes: Their formation, structure, and likely role in the tropical circulation. Meteorology Over the Tropical Oceans, ed DB Shaw (Royal Meteorol Soc, Reading, UK), pp. 155–218 (1979).
34
SB Goldenberg, CW Landsea, AM Mestas-Nunez, WM Gray, The recent increase in Atlantic hurricane activity: Causes and implications. Science 293, 474–479 (2001).
35
KE Taylor, RJ Stouffer, GA Meehl, An overview of CMIP5 and the experiment design. Bull Am Meteorol Soc 93, 485–498 (2012).
36
RH Moss, et al., The next generation of scenarios for climate change research and assessment. Nature 463, 747–756 (2010).
37
GP Peters, et al., The challenge to keep global warming below 2 °C. Nat Clim Chang 3, 4–6 (2012).
38
DP Dee, et al., The ERA-Interim reanalysis: Configuration and performance of the data assimilation system. Q J R Meteorol Soc 137, 553–597 (2011).
39
H Gu, Z Zong, K Hung, A modified superconvergent patch recovery method and its application to large deformation problems. Finite Elem Anal Des 40, 665–687 (2004).
40
C Tebaldi, JM Arblaster, R Knutti, Mapping model agreement on future climate projections. Geophys Res Lett 38, L23701 (2011).
41
CE Duchon, Lanczos filtering in one and two dimensions. J Appl Meteorol 18, 1016–1022 (1979).
42
A Diedhiou, S Janicot, A Viltard, P De Felice, H Laurent, Easterly wave regimes and associated convection over West Africa and tropical Atlantic: Results from the NCEP/NCAR and ECMWF reanalyses. Clim Dyn 15, 795–822 (1999).
43
P Knippertz, AH Fink, A Reiner, P Speth, Three late summer/early autumn cases of tropical–extratropical interactions causing precipitation in Northwest Africa. Mon Weather Rev 131, 116–135 (2003).
44
EKM Chang, Y Guo, X Xia, CMIP5 multimodel ensemble projection of storm track change under global warming. J Geophys Res 117, D23118 (2012).
45
F Giorgi, P Lionello, Climate change projections for the Mediterranean region. Global Planet Change 63, 90–104 (2008).
46
R Roehrig, D Bouniol, F Guichard, F Hourdin, J-L Redelsperger, The present and future of the West African monsoon: A process-oriented assessment of CMIP5 simulations along the AMMA transect. J Clim 26, 6471–6505 (2013).
47
ER Martin, C Thorncroft, BBB Booth, The multidecadal Atlantic SST - Sahel rainfall teleconnection in CMIP5 simulations. J Clim 27, 784–806 (2013).
48
KH Cook, EK Vizy, Coupled model simulations of the West African monsoon system: Twentieth- and twenty-first-century simulations. J Clim 19, 3681–3703 (2006).
49
S Leroux, NMJ Hall, On the relationship between African easterly waves and the African easterly jet. J Atmos Sci 66, 2303–2316 (2009).
50
S Engelstaedter, I Tegen, R Washington, North African dust emissions and transport. Earth Sci Rev 79, 73–100 (2006).
51
TN Carlson, SG Benjamin, Radiative heating rates for Saharan dust. J Atmos Sci 37, 193–213 (1980).
52
NM Mahowald, et al., Observed 20th century desert dust variability: impact on climate and biogeochemistry. Atmos Chem Phys 10, 10875–10893 (2010).

Information & Authors

Information

Published in

Go to Proceedings of the National Academy of Sciences
Proceedings of the National Academy of Sciences
Vol. 111 | No. 19
May 13, 2014
PubMed: 24778244

Classifications

Submission history

Published online: April 28, 2014
Published in issue: May 13, 2014

Keywords

  1. greenhouse gas forcing
  2. CMIP5

Acknowledgments

We thank two anonymous reviewers and the Member Editor for their insightful and constructive comments. We wish to thank Dr. Jen-Shan Hsieh and Deepti Singh for helpful discussions. We acknowledge the World Climate Research Programme and the climate modeling groups for making available their model output, and the U.S. Department of Energy’s Program for Climate Model Diagnosis and Intercomparison for coordinating and supporting database development. Our work was supported by National Science Foundation Award 0955283. Computational resources for data processing and analysis were provided by the School of Earth Sciences’ Center for Computational Earth and Environmental Science at Stanford University.

Notes

This article is a PNAS Direct Submission.

Authors

Affiliations

Christopher Bryan Skinner1 [email protected]
Department of Environmental Earth System Science and
Noah S. Diffenbaugh
Department of Environmental Earth System Science and
Woods Institute for the Environment, Stanford University, Stanford, CA 94305

Notes

1
To whom correspondence should be addressed. E-mail: [email protected].
Author contributions: C.B.S. and N.S.D. designed research; C.B.S. performed research; C.B.S. contributed new reagents/analytic tools; C.B.S. analyzed data; and C.B.S. and N.S.D. wrote the paper.

Competing Interests

The authors declare no conflict of interest.

Metrics & Citations

Metrics

Note: The article usage is presented with a three- to four-day delay and will update daily once available. Due to ths delay, usage data will not appear immediately following publication. Citation information is sourced from Crossref Cited-by service.


Citation statements




Altmetrics

Citations

If you have the appropriate software installed, you can download article citation data to the citation manager of your choice. Simply select your manager software from the list below and click Download.

Cited by

    Loading...

    View Options

    View options

    PDF format

    Download this article as a PDF file

    DOWNLOAD PDF

    Get Access

    Login options

    Check if you have access through your login credentials or your institution to get full access on this article.

    Personal login Institutional Login

    Recommend to a librarian

    Recommend PNAS to a Librarian

    Purchase options

    Purchase this article to access the full text.

    Single Article Purchase

    Projected changes in African easterly wave intensity and track in response to greenhouse forcing
    Proceedings of the National Academy of Sciences
    • Vol. 111
    • No. 19
    • pp. 6853-7161

    Media

    Figures

    Tables

    Other

    Share

    Share

    Share article link

    Share on social media