SEARCH

SEARCH BY CITATION

Keywords:

  • African monsoon;
  • AMMA;
  • AFLAM;
  • inverted-V trough;
  • African Easterly Wave

Abstract

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Overview of AEWs in July 2006
  5. 3. Model evaluation
  6. 4. AEW initiation
  7. 5. Mature AEW
  8. 6. Impacts of the AEW on rainfall and the surface
  9. 7. Discussion
  10. 8. Conclusion
  11. Acknowledgements
  12. References

The detailed structure of an African easterly wave (AEW) observed during the AMMA field campaign is analysed. The wave was present from 25 to 29 July 2006. A complex circulation pattern was observed: the overall structure of convection and the positive vorticity of the trough region had an elongated inverted-V appearance, wrapped around an area of low winds and clear skies. Satellite imagery showed that the AEW was a significant influence on the modulation of convection on the large scale.

The wave was identified initially through its strong signature on soil moisture and convection. The AEW structure observed was not anticipated and has not been discussed in previous literature. In addition, wave tracking using a Hovmöller diagram of meridional winds did not detect the wave, and a Hovmöller of vorticity showed the wave moved at a slower speed than other AEWs in July.

New schematics explaining the structure are presented, describing the case as observed by satellites and analysed by a limited-area version of the Met Office Unified Model. It is proposed that the positive vorticity branches of the inverted-V can be regarded as analogous to atmospheric fronts, with characteristic gradients in winds and thermodynamic properties, acting as locations for enhanced convection. The implications of the new case are discussed in relation to previous theory and it is suggested that the accepted model of an idealised AEW is incomplete and should be extended to include more complex structures. Copyright © 2011 Royal Meteorological Society and British Crown Copyright, the Met Office


1. Introduction

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Overview of AEWs in July 2006
  5. 3. Model evaluation
  6. 4. AEW initiation
  7. 5. Mature AEW
  8. 6. Impacts of the AEW on rainfall and the surface
  9. 7. Discussion
  10. 8. Conclusion
  11. Acknowledgements
  12. References

African easterly waves (AEWs) are present over the West African region from May to October. These waves are tropical disturbances on synoptic time-scales. AEW amplitude is maximised around the African easterly jet (AEJ) level, at a height of around 600–700 hPa. Composite studies have shown AEWs have a complex structure which varies depending on location (e.g. Kiladis et al., 2006). In general, the waves are cold-core, with cold air beneath the 700 hPa vortex (Reed et al., 1977). In the northern Sahel, where rain is less frequent, AEWs can play an important role in the modulation of convection. Here, AEW amplitude peaks at lower levels, at approximately 850 hPa (Reed et al., 1977; Pytharoulis and Thorncroft 1999), and waves may exhibit warm-core features.

The growth of AEWs has been described in terms of the barotropic and baroclinic instability of the AEJ since Burpee (1974). However, it has been realised that the growth rates inferred from instability models (e.g. Thorncroft and Hoskins, 1994) are rather small (e-folding time of about 3 days) in comparison with the propagation time of waves along the AEJ. More recently, Cornforth et al. (2009) showed that including moist processes in a similar model led to an increased wave growth rate, demonstrating that wave development theory is by no means complete. Many AEWs may also be initiated by a large-amplitude stimulus rather than local barotropic instability. Indeed, Hall et al. (2006) showed, using a dry idealised model, that the AEJ may not even be dynamically unstable at all if frictional effects are taken into account, but it still may support wave-like responses to a large-amplitude initial forcing. Waves may be initially forced by deep convection towards the east, over the Chad region or the Nigerian Highlands for example (Berry and Thorncroft 2005; Thorncroft et al., 2008), and move westwards at a speed of 6–8 m s−1.

Over the years, the main obstacle to better understanding AEWs has been the sparsity of the observing network. A composite of observed waves during the 1974 Global Atmospheric Research Program Atlantic Tropical Experiment (GATE) field campaign was presented by Reed et al. (1977), and remains one of the most comprehensive and cited AEW descriptions available. Reed et al. (1977) presented diagrams of streamlines and vorticity showing a series of oval-shaped positive vorticity troughs and negative vorticity ridges, orientated southsouthwest to northnortheast. The waves were identified in streamline analysis, with an approximate wavelength of 3000 km, and were strongest at 700 hPa. Even in this case, the upper-air data were patchy, with only 30% of soundings received at the forecasting centres (Kuettner and Parker, 1976), and with insufficient data in the northern Sahel to accurately describe the northern AEW structure.

There is considerable intraseasonal variability in the occurrence of AEWs, and forecasts for more than a few days are typically poor, probably related to the systematic errors which models tend to exhibit in this region (Thorncroft et al., 2003; Tompkins et al., 2004; Agusti-Panareda et al., 2010a;). AEWs play a significant role in the modulation of rainfall (e.g. Fink and Reiner, 2003) over West Africa and also act as a precursor for tropical storms and hurricanes in the Atlantic (Landsea and Gray, 1992; Avila and Pasch, 1992; Thorncroft and Hodges, 2001; Zawislak and Zipser, 2010, Hopsch et al., 2010). Hence, the current lack of forecast ability has direct consequences on the forecasting of these events also.

In recent years, while model analyses and reanalyses have become more reliable in many parts of the world, their quality for West Africa has remained problematic (Tompkins et al., 2004; Faccani et al., 2009; Agusti-Panareda et al., 2010a). Our understanding of AEWs remains founded on climatological studies which are limited by the upper-air network, with only a handful of detailed case-studies available (e.g. Redelsperger et al., 2002; Berry and Thorncroft, 2005). In reality, there are wide differences in the structure of individual AEWs within a given season which do not adhere to the climatological composites. Hence, there remains an important need for detailed case-study analyses of different AEWs.

This article documents a case-study of an AEW, present from 25 to 29 July 2006 and describes the atmospheric structure in detail as well as the impact on rainfall and the land surface. This event occurred during one of the Special Observing Periods, SOP2, of the African Monsoon Multidisciplinary Analysis (AMMA) field campaign in West Africa (Redelsperger et al., 2006). As part of the SOPs, the upper-air network was greatly enhanced over the region, with unprecedented spatial coverage across West Africa, and much better transmission of data to the international weather prediction centres (Parker et al., 2008). Therefore, there has been no better-observed period for which to document AEW behaviour using upper-air measurements and the model analyses which make use of these measurements.

The AEW was initially identified by its strong impact on rainfall and, rather unusually, was well-captured by operational numerical weather prediction systems several days in advance. Barthe et al. (2010) gave an overview of the AEW and described the rainfall associated with the event in the Niamey region. Cuesta et al. (2009) described a rainfall event in the Hoggar massif immediately prior to, and during, the case-study period. They suggest that the Hoggar rainfall event was a result of increased southerly flow pulling moisture northwards. The anomalous flow was caused by the AEW location combined with the orientation and location of the Saharan heat low. Here we investigate the southerly structure of the AEW trough, and explain how it influenced convection and surface meteorology in the southern region of West Africa. The overall structure of convection and positive vorticity associated with the wave was an ‘inverted-V’ shape. Frank (1969) described similar weather system structure in the Atlantic as an anomaly in the Intertropical Convergence Zone. Subsequent workers (e.g. Ferreira and Schubert, 1997; Bracken and Bosart, 2000) have referred to inverted-V systems in the Tropics but mostly concentrated over ocean regions. This is the first detailed documentation of such a structure over the West Africa region.

Section 2 gives an overview of AEWs in 2006 over West Africa to give context to the case-study. Section 3 introduces the data and evaluates the numerical model which is used in the analysis stage. Section 4 details the initiation of the wave and section 5 discusses the mature phase structure of the wave. Section 6 is discussion and section 7 the conclusion.

2. Overview of AEWs in July 2006

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Overview of AEWs in July 2006
  5. 3. Model evaluation
  6. 4. AEW initiation
  7. 5. Mature AEW
  8. 6. Impacts of the AEW on rainfall and the surface
  9. 7. Discussion
  10. 8. Conclusion
  11. Acknowledgements
  12. References

Janicot et al. (2008) give an overview of the summer monsoon conditions during the AMMA SOP in 2006. The 2006 season was relatively unusual as the dynamical monsoon onset was in line with climatology but the convective monsoon onset (defined by the first rains in the north and the shift in convection patterns; Sultan and Janicot, 2003) was a month later. There was therefore a dry period after the wind shift had initiated the monsoon circulation.

Janicot et al. (2008) showed that in June and the beginning of July there were fewer identifiable AEWs than the climatological average, and those that were identified were often weaker in their dynamical structure than had been observed from previous years. By mid-July to early August, wave activity had increased, but the initiation of waves occurred further to the west than in the mean climatology (Thorncroft et al., 2007). The 25 to 29 July case was one of the stronger wave events during this early monsoon period; the AEW could be detected in the dynamical fields and convection was forecast rather well by all the models (discussed in section 3). However, despite these strong features, the wave structure differed somewhat from typical AEW troughs.

Figure 1(a) shows the meridional (v) wind Hovmöller diagrams for July 2006 as in Fink and Reiner (2003). These are often used to track the longitudinal location of AEWs. The locations of AEW troughs (bold lines) are shown where winds are switching from southerlies to northerlies. These locations were confirmed by manual tracking of cyclonic signals in streamline plots (not shown) using the same methodology as Fink and Reiner (2003). Eight wave events were identified using this method.

thumbnail image

Figure 1. (a) Hovmöller plot of meridional (v) winds in July 2006; shading represents negative v-winds at 850 hPa (17–23°N), contours represent v-winds at 700 hPa (7–13°N) every 2 m s−1, with positive v-winds (southerlies) being solid lines, and negative dotted lines. Bold black lines show the tracks of AEWs. (b) Hovmöller plot of 2–6 day band-passed filtered relative vorticity at 700 hPa (11–17°N), contoured every 6 ×10−6s−1. Shading represents positive vorticity, and dashed contours negative vorticity. AEWs identified from (a) are overplotted as bold black lines, and the case-study wave is shown by the bold dashed line. Composed from ECMWF AMMA re-analysis (details are given in Agusti-Panareda et al., 2010b).

Download figure to PowerPoint

Figure 1(b) shows Hovmöllers of vorticity at 700 hPa in the southern part of the domain (11–17°N). Hovmöllers of vorticity tend to be noisier than v-winds, so band-pass filtering is used to highlight the AEW signal. The eight events shown from Figure 1(a) are overlaid directly onto the plot and a mismatch can be seen between the location of the AEWs and the positive vorticity, particularly towards the end of the month. Some mismatch may be expected as northern and southern centres of an AEW may be offset in the longitude. However, the second to last trough is significantly offset and moved westward at a faster speed than the vorticity signature. It is this slower-moving vorticity signal (indicated by the dashed line) which we attribute to the case-study described in subsequent sections. It is clear from this figure that this trough was moving at a slower speed than the other AEWs.

Figure 1 therefore demonstrates that the wave described in our case-study was not identifiable in the v-wind Hovmöller but could be seen in the vorticity. The signal was not found in the v-winds because the trough was narrow and orientated at an angle to the coast, with a significant east–west extension, leading to diminished amplitude in the meridional averaging in Figure 1(a).

3. Model evaluation

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Overview of AEWs in July 2006
  5. 3. Model evaluation
  6. 4. AEW initiation
  7. 5. Mature AEW
  8. 6. Impacts of the AEW on rainfall and the surface
  9. 7. Discussion
  10. 8. Conclusion
  11. Acknowledgements
  12. References

Output from a limited-area model (LAM) version of the UK Met Office Unified Model (MetUM), the Africa LAM (hereafter AFLAM), is used in this paper to describe the AEW structure on the large scale. The AFLAM was initialised each day at 0600 UTC from the MetUM global analysis. It then underwent two 6-hourly 3D-Var assimilations before the model was run from 1800 onwards. All the data used in the figures are between the T+6 h (0000 UTC) to T+27 h (2100 UTC the following day) forecasts from these runs. During 2006, MetUM had a resolution of approximately 60 km over Africa, and the AFLAM had a horizontal resolution of approximately 20 km, covering 20°W to 55°E, 40°N to 40°S. Both AFLAM and the global MetUM had physical formulation similar to the HadGEM1 version of the Met Office Hadley Centre climate model (Martin et al., 2006). More details of the global MetUM version can be found in Allan et al. (2007). The AFLAM was used due to its high availability of time sampling and its high spatial resolution.

The description in subsequent sections will focus on AEW structure in the period 26 to 28 July when the wave organisation was consistent in model and observations. Inspection of data from other general circulation models (not shown; available at AMMA operational centre web pages*) confirmed that there was good consistency between different models in their representation of the large-scale dynamics at this time.

A comparison between the AFLAM and Bamako airport radiosonde observations is shown in Figure 2. The times when radiosonde data are available have been marked by black triangles; the remainder of the plot is interpolated in time. Both model and observations showed changes in temperature and wind regimes between 26 and 28 July at the time of trough passage.

thumbnail image

Figure 2. Time slice vertical cross-sections using sondes from Bamako airport and AFLAM output for 25 to 29 July 2006: (a) temperature anomaly interpolated from radiosondes (indicated by black triangles), (b) temperature anomaly from AFLAM, (c) meridional wind anomaly from radiosondes, and (d) meridional wind anomaly from AFLAM. Hatching denotes missing data.

Download figure to PowerPoint

The radiosonde observations in Figure 2(a) showed a tight temperature gradient near the surface on 26 July, changing from warmer to cooler anomalies between 0000 and 1200 UTC. Thermal infrared imagery from the Meteosat Second Generation (MSG) satellite indicated the rapid local development of deep convection (cloud tops colder than −80°C) in the early hours of 26 July. There was also evidence of significant wetting of the surface in the vicinity of the airport from MSG land surface temperature and AMSRE soil moisture data on 26 July, though no rain was reported at the airport itself. Where the surface was wet, one would expect a cooling of the boundary layer, as observed. An alternative (or additional) influence on the temperature at the larger scale may also have been the arrival of the head of the cool monsoon, which penetrates inland at night-time (Parker et al., 2005a). The radiosonde at 0000 UTC on 28 July shows the surface as warmer, but whether this was due to gradual warming on the 27 July (as linear interpolation suggests) is unknown due to a missing sonde launch at 1200 UTC on 27 July.

An anomalously cooler region was also present at the 600 hPa level on 26 July. This appeared to move downwards with time to join with the surface cool anomaly on 27 July, which could have acted to maintain low temperatures first introduced by any nocturnal monsoon surge or rainfall event. This period coincided with the initial trough passage. The temperature cross-sections could therefore indicate that the trough was sloping westward with increasing height.

The model successfully captured many thermal features identified by the radiosonde interpolation. However, the vertical complexity of the wave was not fully represented by the model. Additionally, the arrival of the near-surface cool anomaly occurred a few hours earlier (0000 UTC on 26 July) in the model than in reality (1200 UTC on 26 July). This is likely due to the presence of a large rain storm in the model, present over Bamako between 2100 UTC on 25 July and 0000 UTC on 26 July. The storm concurs with observations from the satellite which showed a storm near Bamako, though the timing of the model storm is possibly a couple of hours too early.

In the meridional wind anomalies, the northerly component of the AEJ was present from 25 July at 0000 UTC to 26 July at 1200 UTC above 850 hPa. The winds became southerly (dark shading) by 27 July at 0000 UTC. In the model at mid-levels (approximately 700 hPa) the arrival of the southerlies was somewhat delayed (0000 UTC on 27 July) in comparison to the interpolated observations. However, the model may not be unrepresentative as the gradual change shown in the observations of Figure 2(c) could be due to the linear interpolation between radiosondes at 1200 UTC on 26 July, and 0000 UTC on 27 July. The southerlies persisted at 600 hPa until 28 July at 0000 UTC. The 800 hPa northerlies (light shading) observed at 0000 and 1200 UTC on 28 July were not apparent in the model.

To assess the model's ability to simulate moist features, the AFLAM rainfall is shown alongside the observed TRMM rainfall in Figure 3 for 27 July, at the peak of AEW maturity. The figure shows that the AFLAM overestimated the rainfall amount in comparison to TRMM and also enlarged the regions of strong precipitation around 10°W and 20°W. The TRMM showed rainfall east of 5°E, over the Nigerian highlands and in southern Chad and in western Mauritania (18°N, 15°W), details of which were not found in the model. However, the synoptic envelope of rainfall was qualitatively similar in model and data, with meridional differences not exceeding 5° latitude between the limits of convection in each case. The heavy rainfall in southwestern Mali (15°N, 15–5°W) was in the correct location (though too strong), and an additional band of rainfall along the Mali–Algerian border was also captured by the model. The envelopes of convection also agreed in the southern region of West Africa, over Ivory Coast, where there was sparse rainfall, though rainfall over Ghana and the eastern part of the Guinean coast was overestimated. It is a common tendency of models to smooth out rainfall and miss the extremes, therefore it is equally important to correctly model regions where rainfall is very low or absent.

thumbnail image

Figure 3. Mean rainfall (mm) on 27 July from (a) TRMM satellite and (b) AFLAM. Note the different shading for low values to include small TRMM rainfall amounts.

Download figure to PowerPoint

Overall this comparison has demonstrated that the model had a reasonable representation of the location of rainfall at this time in that in general the rainfall envelopes were in agreement, despite the model having too much rain. Both the rainfall comparison and the radiosonde comparison have shown differences from observations, demonstrating that the full complexity of AEWs may still be unrepresented in dynamical models. Despite the differences, this brief evaluation has indicated that the AFLAM was able to capture many aspects of the vertical and horizontal structure of the trough which were confirmed by observation and this lends confidence to its use as a diagnostic tool.

4. AEW initiation

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Overview of AEWs in July 2006
  5. 3. Model evaluation
  6. 4. AEW initiation
  7. 5. Mature AEW
  8. 6. Impacts of the AEW on rainfall and the surface
  9. 7. Discussion
  10. 8. Conclusion
  11. Acknowledgements
  12. References

The 25 to 29 July wave was first seen in the vorticity fields to initiate (or become more organised) over central Niger on 25 July (Figure 1(b)). A number of factors contributed to its initial structure.

A previous trough had passed from 22 to 25 July, shown in the Hovmöllers in Figure 1. The northern vortex of this system was particularly strong and formed in the lee of the Hoggar mountains, in conjunction with an upper-level trough from midlatitudes. This previous AEW had caused great disturbance in the winds at 850 hPa in the Sahara (Cuesta et al., 2009, provide a full description).

The 22 to 25 July wave was later seen to interact with the initiating phase of the case-study wave: the strong southerly component of the trough on 24 July pushed the western axis of the AEJ northwards, while the eastern end of the AEJ remained further south. Figure 4 shows the initiation and development of the case study wave from 25 to 27 July at 850 hPa in the zonal winds. The first plot, at 0000 UTC on 25 July, shows the maximum zonal winds, or ‘jet core’, orientated southeast to northwest. A small kink is observed to the south of the jet in southern Nigeria (7°N, 7°E). Satellite images showed a strong convective event occurred on the southern edge of the jet in the same region as the kink. The perturbation grew in the meridional direction, creating a northward distortion in the jet on 26 July, seen in Figure 4(c).

thumbnail image

Figure 4. Initiation and development of the wave from AFLAM in the zonal (u) winds (shading) at 700 hPa. Streamlines are overlaid in grey and wind vectors are shown as black arrows: (a) 0000 UTC on 25 July (T+6 h), (b) 1200 UTC on 25 July (T+18 h), (c) 0000 UTC on 26 July (T+6 h), (d) 1200 UTC on 26 July (T+18 h), (e) 0000 UTC on 27 July (T+6 h), (f) 1200 UTC on 27 July (T+18 h). Forecasts were initiated at 1800 UTC on (a, b) 24 July, (c, d) 25 July and (e, f) 26 July.

Download figure to PowerPoint

The passage of a fast-moving vortex in the Sahara on 25 July (seen in the streamlines of Figure 4(b) at 22.N, 3.W and also in Cuesta et al., 2009) appeared to contribute/enhance the development of the kink on the zonal jet. The winds associated with the northern vortex advected cool, moist air northward, bringing the zonal jet northward also through thermal wind balance. By 0000 UTC on 26 July (Figure 4(c)), the bend in the jet was well established. Subsequent images will demonstrate this is the time when the positive vorticity to the south of the jet formed into an inverted-V shape. This was the mature phase of the AEW.

5. Mature AEW

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Overview of AEWs in July 2006
  5. 3. Model evaluation
  6. 4. AEW initiation
  7. 5. Mature AEW
  8. 6. Impacts of the AEW on rainfall and the surface
  9. 7. Discussion
  10. 8. Conclusion
  11. Acknowledgements
  12. References

5.1. Dynamical structure: Vorticity and winds at AEJ level

The wave reached maturity from 26 to 28 July 2006. At this time a significant vorticity signal could be distinguished and the large-scale winds were influenced by the feature.

Figure 5(a) is a summary of the main dynamical features, showing lines of constant geopotential height plotted over vorticity (shading) with wind vectors at 0000 UTC on 27 July 2006, from the T+6 h AFLAM forecast. This time was chosen as it is the peak of the mature phase and less turbulence exists in the nocturnal, rather than daytime, boundary layer (Parker et al., 2005a).

thumbnail image

Figure 5. AFLAM diagnostics of the wave at 0000 UTC on 27 July from the 6 h forecast: (a) vorticity (shading) where reds are positive, blues are negative vorticity. Overlaid are contours of constant geopotential height in green and wind vectors, all at 700 hPa. (b) Contoured vorticity at 700 hPa above 10−5s−1 (bold black line) overlaid on temperature anomaly from the zonal mean at 850 hPa (shading). Dark grey shading is anomalously cold, light grey shading is anomalously warm.

Download figure to PowerPoint

The geopotential height contours show a trough-like feature tilting south-west to north-east from Liberia (7°N, 12°W) to eastern Mali (15°N, 0°E). The trough is therefore tilted eastward in the horizontal plane against the westward direction of the AEJ. Along the trough in a narrow strip is high positive vorticity. For future reference this will be called the ‘western branch’ of the system.

A contiguous narrow branch of positive vorticity extends from the peak of the trough at 15°N, 0°E to the south-east at 7°N, 10°E. This corresponds to the wind changing from south-easterly to easterly in this narrow region, as it follows the lines of constant geopotential height. This will be referenced in the remainder of the paper as the ‘eastern branch’ of the system. The veering of the wind direction to more southerly further west (over northern Benin, for example) is consistent with there being negative vorticity in this area. The picture is therefore showing a sloped trough, with positive vorticity acting like branches spanning the whole system in an inverted-V shape, and weak negative vorticity to the east of the centre where the winds switched to become more southerly in the trough.

The shape of the trough also shows up strongly in the location and path of the AEJ, shown in Figure 4. The strongest wind associated with the jet is seen to curve to the north of the positive vorticity strip. Initially the shape of the vorticity strip was oriented east to west in accordance with the location of the AEJ. The jet was seen to break down during the decay of the wave on 28 July, with light westerly winds intruding over West Mali.

5.2. Thermal structure

Figure 5(b) shows the potential temperature anomaly at 850 hPa, below the jet level, overlaid by the vorticity structure: we anticipate from previous studies (Reed et al., 1977; Parker et al., 2005a; Cook 1999) that the AEW and AEJ structures resemble thermal wind balance, with the jet-level winds linked to the thermal patterns in the layer beneath. The temperature anomaly shows a cool region over the western branch of the system. This corresponds to the termination region of southerly low level winds, seen in Figure 5(a), which would have brought cooler, moist air northwards. There were uniformly cool temperatures to the south of the trough structure between 10°W and 5°E. This shows as a zero anomaly between 8°W and 3°E up to 13°N. At these latitudes, there was a large contrast from east to west, with a warm anomaly at the eastern branch of the 700 hPa vortex, a neutral centre and a cold western branch. Temperatures increased to the east of the eastern branch, up to the edge of the domain.

The cool anomaly extended north-eastwards of the main vortex structure. It is unclear whether the cool air in the north (i.e. over Algeria) was due to the influence of the AEW, or whether it was connected directly to the rain event (Cuesta et al., 2009) and increased cloud cover in the preceding days, cooling the land surface and lower atmosphere.

5.3. Vertical structure

Figure 6 shows zonal cross sections of the vorticity structure (shaded) with potential temperature, mixing ratio and meridional winds. The cross sections were taken through 12.5°N, to transect across the full bend in the jet and vorticity. The positive vorticity features of the west and east branches can be seen in light grey shading. The cross sections show a forward vertical slope of the positive vorticity feature to the west (western branch) and the backwards slope of the vorticity feature to the east (eastern branch). These have been marked on the figure by hand to join up the regions of positive vorticity.

thumbnail image

Figure 6. Cross-sections at 12.5°N from AFLAM at 0000 UTC on 27 July (T+6 h). Vorticity shaded overlaid with contours of (a) theta anomaly (K), (b) mixing ratio anomaly (g kg−1) and (c) meridional winds (m s−1). The slopes of the western and eastern branches have been marked as bold black lines.

Download figure to PowerPoint

Figure 6(a) shows cool potential temperature (theta) anomalies (dashed lines) were collocated with the western branch above 900 hPa and at the eastern branch above 600 hPa. This shows that both before and after trough passage, anomalously cool air (relative to its level) was present above anomalously warmer air at the surface.

Figure 6(b) shows changes in the mixing ratio occurred close to the branch regions. The western branch was coupled with a strong moistening (positive mixing ratio anomaly) at the same time as the cooling of the troposphere. To the east of the eastern branch, the mixing ratio anomalies are mostly negative (dashed) indicating drier air.

Figure 6(c) shows changes in the meridional winds. The meridional winds were out of phase with the western branch in contrast to the thermal and moist structures. There were southerlies to the east of the western branch and northerlies to the west at 700 hPa. This indicates there was significant horizontal shear across the western branch, accounting for the high vorticity. The shear across the eastern branch was less distinct due to dominating southerlies.

The northerlies at 700 hPa (dashed lines in Figure 6(c)) between 8 and 13°W are where the AEJ exited the cross-section. The AEJ entered the cross-section at the same level at 6°E. The winds near the surface throughout the domain were predominantly southerly due to influence of the monsoon at night.

The configuration of the meridional winds coupled with the thermal structure would have brought cooling to the east of the centre of the trough and warming to the west, causing the thermal pattern to propagate eastwards relative to the mean flow. This propagation shares the dynamics of Rossby waves, and would counteract the tendency of the AEJ to tilt the vortex strip and thermal structure downwind (i.e. westwards).

The overall vertical structure of the western branch may be described as similar to that of a split cold front in midlatitudes, in which cooler air over-runs the surface front. In a split cold front, changes in vertical wind shear can lead to portions of the front becoming tilted downshear (Parker, 1999). Alternatively, differential latent heating may set up an anomalous frontal structure (Locatelli et al., 2002). In the case of Figure 6, this could mean that the single tilt structure indicated by the bold hand-drawn lines is too simplified, and the front (or western branch) was tilting in the opposite direction below the cold anomaly at 900 hPa. This proposition may also be supported by the radiosondes at Bamako in Figure 2. The eastward tilt of the eastern branch was not observed in the radiosondes as the latter part of the wave had broken up by the time it reached Bamako.

To the east of the eastern branch, higher potential temperature, coupled with mostly drier mixing ratio values were present, signifying the final passing of the trough and a change towards AEW ridge conditions.

6. Impacts of the AEW on rainfall and the surface

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Overview of AEWs in July 2006
  5. 3. Model evaluation
  6. 4. AEW initiation
  7. 5. Mature AEW
  8. 6. Impacts of the AEW on rainfall and the surface
  9. 7. Discussion
  10. 8. Conclusion
  11. Acknowledgements
  12. References

6.1. Convergence and Convection

Figure 7(a) shows the model-generated convergence patterns plotted over equivalent potential temperature (theta-e) and wind vectors at 925 hPa on 27 July at 0000 UTC. The vorticity structure of the trough is overlaid in bold contour. The figure shows a well developed convergence line (hatching) north of the vorticity structure (e.g. east and west of 0°E, 20°N). There is a second set of convergence lines in Algeria, lying at the intertropical discontinuity (ITD, e.g. east of 0°E, 23°N). In addition, lines of convergence can be seen surrounding 22°N 7°W, which is most probably associated with the core of the Saharan heat low. The heat low is very dry, accounting for the low values of equivalent potential temperature, despite the high local temperatures. The southern part of the domain shows southerly low level winds and cool temperatures. The region with the highest equivalent potential temperature (indicated by light shading) is largely consistent with the location of the AEJ, the convection over Mauritania and also over Algeria. The arrows indicate large-scale southerlies at 925 hPa associated with the monsoon inflow. Winds at the eastern branch turn from southerlies to westerlies, indicating weak anticyclonic circulation.

thumbnail image

Figure 7. Convergence and convection at 0000 UTC on 27 July (a) AFLAM T+6 h forecast of theta-e (shading) at 925 hPa overlaid by wind vectors (arrows) and convergence (hatched contours). (b) Satellite IR image overlaid by AFLAM (solid line) vorticity structure at 700 hPa. Overlaid on both plots is vorticity at 700 hPa above 10−5s−1 (bold black line), to indicate the location of the trough.

Download figure to PowerPoint

Figure 7(b) shows the satellite image from 27 July at 0000 UTC, overlaid by the outline of the vorticity strip at 700 hPa taken from AFLAM. The figure suggests two main regions of cloudiness, one along the axis of the jet/trough, the second further to the north in Algeria. The northern cloud was coincident with the convergence near the ITD, identified from Figure 7(a). Inspection of further satellite images (not shown) indicated the convergence of cloud along the AEW trough axis was most notable on 27 July between 1800 and 0000 UTC.

There is therefore correspondence between the region of increased vorticity at 700 hPa and observed cloud cover in the satellite IR field. To the south and east of the inverted-V structure, convection is notably absent. It is true that convection will have a feedback on vorticity, but the positive vorticity structure was present before the convection, leading to the assertion that the structure was unlikely to be an effect of the increased cloud, but could have acted to stimulate or enhance convective development.

6.2. Impact on soil moisture

The soil moisture gradient in the Sahel region is particularly strong from north to south. Rainfall in the Sahel region (12–17°N) can have significant impact on land surface fluxes and change local gradients in soil moisture. Large scale rainfall anomalies therefore have the potential to make significant changes in surface fluxes which may impact the boundary layer.

Section 6.1 demonstrated that the passage of the AEW enhanced convection in the inverted-V trough region and inhibited it in the ridge region behind. The result of this was a wetting of the soil in the trough region with drying taking place afterwards, effectively leading to the westward propagation of soil moisture, or in effect, a westward propagating wave in the soil moisture. Figure 8 shows microwave satellite images of soil moisture and anomalous soil moisture during the period from the AMSRE satellite (Njoku et al., 2000; Gruhier et al., 2008). The location of the trough has been overplotted as a bold contour. The figures show a moist anomaly propagating from east to west along the Sahelian gradient region, consistent with the location of the trough. The anomalies from 26 to 28 July from the 25 July are shown to highlight the propagating moist feature. Taylor et al. (2005) observed this phenomenon in satellite remote sensing and global model analyses of the 2000 season. However, the case described here may be the first detailed case-study of such a soil moisture wave caused by AEW trough passage.

thumbnail image

Figure 8. (a) Daily maximum volume soil moisture (mm) from AMSRE microwave satellite on 25 July; wet soils are darker greys. (b) Soil moisture on 26 July, (c) soil moisture on 27 July, and (d) soil moisture on 28 July, all three as anomalies from 25 July. Positive vorticity (> 10−5s−1) at midnight each day is overplotted on (b)–(d) in bold to show the location of the trough.

Download figure to PowerPoint

The cooler, wetter land surface at 5°W–5°E in Figure 8(b) was a result of convection on 25 July. The cool land surface was coincident with the region where the atmospheric wave was growing in the meridional direction (Figure 4(d)). The apex of the inverted-V and the trough's developing moist, cool western branch passed over the wetter land surface on 26 July. The cool surface could have therefore acted to amplify the cool signal at the western branch seen in the mature phase of the wave (Figure 5(b) and Figure 6(a)).

6.3. Surface observations

SOnic Detection And Ranging (SODAR) data from Banizoumbou field site (13.03°N, 2.04°E) were used to assess the impact of the synoptic event at the surface. While larger scale synoptic events can be decoupled from the boundary layer, Bain et al. (2010) also demonstrated that the boundary layer winds could be impacted by AEW passage if conditions are favourable. Figure 9(a) shows the wind direction up to 800 m from 27 July from the SODAR. There was a significant change in the wind direction from westerly to easterly during the day with the most noticeable change occurring between 1400 and 1500 UTC. The wind speed increased at the same time as this change.

thumbnail image

Figure 9. (a) Wind direction (degrees) and (b) vertical wind speed (m s−1) from Banizoumbou SODAR on 27 July. The white areas denotes missing data.

Download figure to PowerPoint

This coincides with a time of sustained uplift, shown in the vertical velocity plot of part (b), with vertical wind speeds reaching 1 m s−1. Infra-red satellite pictures showed no storms nearby (within 100 km), giving confidence the uplift it is due to some synoptic or meso-scale atmospheric movement. The horizontal speed was maintained in the afternoon, providing further evidence that the feature which caused the directional change was not a gust front.

The passage of the eastern branch of the inverted-V shaped trough was occurring at the time of the wind changes. The SODAR data indicate a veering of the winds with time and also with height. This could be indicative of warm advection taking place. This concurs with Figure 5(b) and Figure 7(a) which showed warm air to the east of the system.

7. Discussion

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Overview of AEWs in July 2006
  5. 3. Model evaluation
  6. 4. AEW initiation
  7. 5. Mature AEW
  8. 6. Impacts of the AEW on rainfall and the surface
  9. 7. Discussion
  10. 8. Conclusion
  11. Acknowledgements
  12. References

The description of the wave present 25 to 28 July 2006 shows that this was a strong weather event. It had many features which were consistent with previous examples of AEWs, such as positive vorticity associated with increased convergence and convection. However, there were also some features which varied strongly from the well-known AEW composite figures from Reed et al. (1977). The most notable was the pattern of the winds and associated positive vorticity which characterised the system. The inverted-V banding of the vorticity was due to the distortion of the geopotential height diagonally across the normally zonal orientation of the AEJ. The AEJ was deformed northwards at this time, leading to a kink in the jet which accounted for the shape of the positive vorticity to the south of the jet.

Figure 10 is a schematic of the system in its mature phase, including winds at low levels. The western branch of the system was cool, moist and baroclinically unstable at this point. This would have led to further development of the southerly winds in the centre of the system, explaining the amplification and elongation of the wave in the north-south direction during maturity. The two vorticity branches spanning the system were characterised by very different features. The western branch was a confluence zone with strong horizontal shear in the v-winds. This was also an area of cool temperatures, high moisture and convergence, accounting for the initiation of convection on 27 July. The eastern branch had stronger thermal contrasts on the east side of it, where warmer, dry air was brought southwards by the northerly winds coming from the desert.

thumbnail image

Figure 10. Schematic of the AEW seen from (a) a horizontal perspective where the positive vorticity is shaded and the AEJ location is shown as a block arrow, lower level winds are represented by small black arrows and the orientation of the trough axis is shown by the dashed line, and (b) a vertical perspective with location of convection and winds represented as coming in and out of the diagram; the eastern and western branches are marked by black lines; the view is looking northwards, from the south.

Download figure to PowerPoint

The system shown is reminiscent of a midlatitude cyclone, with a cold core, baroclinic structure and two regions of frontal convergence. The branches of the positive vorticity anomaly act like fronts in that there is convergence and a change in air characteristics across them, and convection is initiated along them. The model and observations suggested a moist centred system with drier air brought south within the northerly flow behind the eastern trough branch.

The wave was slow moving in comparison to other AEWs identified in the Hovmöller of Figure 1(b) and was not identified by the v-winds in Figure 1(a). This case suggests that theory on African Easterly Waves is by no means complete and raises questions on whether this type of system may have gone previously unnoticed. A useful extension to this research would be to assess how often this type of inverted-V structure occurs, and whether there are certain climate conditions which favour development.

A further observation was that the atmospheric wave produced substantial large scale variations in soil moisture through its modulation of convection within the trough region. Here, we have described the impact as an eastward propagating soil moisture wave. Further work on the feedback of this propagating enhanced wet region on the atmosphere would provide new information on indirect impacts of AEWs on the boundary layer, the AEJ, and on the dynamics of the AEW itself.

8. Conclusion

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Overview of AEWs in July 2006
  5. 3. Model evaluation
  6. 4. AEW initiation
  7. 5. Mature AEW
  8. 6. Impacts of the AEW on rainfall and the surface
  9. 7. Discussion
  10. 8. Conclusion
  11. Acknowledgements
  12. References

This work has introduced a case-study of an AEW which had a strong influence on convection and had an inverted-V shaped vorticity signature. Model analyses of vertical and horizontal atmospheric fields have been shown together with radiosonde, SODAR and satellite observations. The AEW structure observed in this case-study has not been previously explained or observed in literature, and was not identifiable in wave tracking using a Hovmöller of meridional winds.

The wave was identified initially through its strong signature on soil moisture and convection. The schematic in Figure 10 describes the case as observed by satellites and analysed by the UM. The AEW had a cold, moist centre and convergence, convection and rainfall lined the vorticity branches. It was also discussed that the observed structure of the system shared some characteristics with midlatitude cyclones, with the positive vorticity branches of the AEW acting like fronts between different air masses and locations for enhanced convection.

This case has shown that there are alternative AEW structures to the accepted Reed et al. (1977) composite model of alternating vorticity maxima/minima. The case shown is particularly relevant for its influence of rainfall and its accurate depiction in the numerical forecast models, but it may have been missed by current methods of trough detection. Previous literature has focused primarily on composite studies of AEWs and as such the structure observed in this case has not been discussed before. This highlights the need for further research into observed wave structures.

Acknowledgements

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Overview of AEWs in July 2006
  5. 3. Model evaluation
  6. 4. AEW initiation
  7. 5. Mature AEW
  8. 6. Impacts of the AEW on rainfall and the surface
  9. 7. Discussion
  10. 8. Conclusion
  11. Acknowledgements
  12. References

Thanks go to Susan Pohle, Robert Schuster and Volker Ermet for their help with wave tracking. The SODARs were installed and operated through NCAS with help from Matthew Hobby. Ken Knapp provided the satellite data from the GridSat database. This work was initiated by a PhD project funded by NERC grant NER/S/A/2004/12332. The work was also supported by NERC grant NE/B505538/1.

Based on a French initiative, AMMA was built by an international scientific group and is currently funded by a large number of agencies, especially from France, the UK, the USA and Africa. It has been the beneficiary of a major financial contribution from the European Community's Sixth Framework Research Programme. Detailed information on scientific coordination and funding is available on the AMMA International web site http://www.amma-international.org

References

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Overview of AEWs in July 2006
  5. 3. Model evaluation
  6. 4. AEW initiation
  7. 5. Mature AEW
  8. 6. Impacts of the AEW on rainfall and the surface
  9. 7. Discussion
  10. 8. Conclusion
  11. Acknowledgements
  12. References
  • Agusti-Panareda A, Beljaars A, Cardinali C, Genkova I, Thorncroft CD. 2010a. Impact of assimilating AMMA soundings on ECMWF analyses and forecasts. Weather Forecasting 25: 11421160.
  • Agusti-Panareda A, Beljaars A, Ahlgrimm M, Balsamo G, Bock O, Forbes R, Ghelli A, Guichard F, Koehler M, Meynadier R, Morcrette J-J. 2010b. The ECMWF re-analysis for the AMMA observational campaign. Q. J. R. Meteorol. Soc. 136: 14571472.
  • Allan RP, Slingo A, Milton SF, Brooks ME. 2007. Evaluation of the Met Office global forecast model using geostationary Earth Radiation Budget (GERB) data. Q. J. R. Meteorol. Soc. 133: 19932010.
  • Avila LA, Pasch RJ. 1992. Atlantic tropical systems of 1991. Mon. Weather Rev. 133: 19932010.
  • Bain CL, Parker DJ, Taylor CM, Kergoat L, Guichard F. 2010. Observations of the nocturnal boundary layer associated with the West African monsoon. Mon. Weather Rev. 138: 31423156.
  • Barthe C, Asencio N, Lafore J-P, Chong M, Campistron B, Cazenave F. 2010. Muli-scale analysis of the 25–27 July 2006 convective period over Niamey: Comparison between Dopper radar observations and simulations. Q. J. R. Meteorol. Soc. 136: 190208.
  • Berry GJ, Thorncroft CD. 2005. Case study of an intense African easterly wave. Mon. Weather Rev. 133: 752766.
  • Bracken WE, Bosart LF. 2000. The role of synoptic-scale flow during tropical cyclogenesis over the North Atlantic Ocean. Mon. Weather Rev. 128: 353376.
  • Burpee RW. 1972. The origin and structure of easterly waves in the lower troposphere of North Africa. J. Atmos. Sci. 29: 7790.
  • Burpee RW. 1974. Characteristics of North African easterly waves during the summers of 1968 and 1969. J. Atmos. Sci. 31: 15561570.
  • Cook KH. 1999. Generation of the African easterly jet and its role in determining West African precipitation. Part 1 J. Climate 12: 11651184.
  • Cornforth RJ, Hoskins BJ, Thorncroft CD. 2009. The impact of moist processes on the African easterly jet–African easterly wave system. Q. J. R. Meteorol. Soc. 135: 894913.
  • Cuesta J, Lavaysee C, Flamant C, Mimouni M, Knippertz P. 2009. Northward burst of the West African monsoon leading to rainfall over the Hoggar Massif, Algeria. Q. J. R. Meteorol. Soc. 135: 3442.
  • Faccani C, Rabier F, Fourrié N, Agusti-Panareda A, Karbou F, Moll P, Lafore J-P, Nuret M. Hdidou FZ, Bock O. 2009. The impact of the AMMA radiosonde data on the French global assimilation and forecast system. Weather Forecasting 24: 12681286.
  • Ferreira RN, Schubert WH. 1997. Barotropic aspects of ITCZ breakdown. J. Atmos. Sci. 54: 261285.
  • Fink A, Reiner A. 2003. Spatio-temporal variability of the relation between African easterly waves and West African squall lines in 1998 and 1999. J. Geophys. Res. 108: 4332, DOI: 10.1029/2002JD002816.
  • Frank NL. 1969. The ‘inverted-v’ cloud pattern—An easterly wave? Mon. Weather Rev. 97: 130140.
  • Gruhier C, de Rosnay P, Kerr Y, Mougin E, Ceschia E, Calvet J-C. 2008. Evaluation of AMSR-E soil moisture product based on ground soil moisture measurements over temperate and semi-arid regions. Geophys. Res. Lett. 35: L10405. DOI: 10.1029/2008GL033330.
  • Hall NJM, Kiladis GN, Thorncroft CD. 2006. Three-dimensional structure and dynamics of African easterly waves. Part II: Dynamical modes. J. Atmos. Sci. 63: 22312245.
  • Hopsch SB, Thorncroft CD, Tyle KR. 2010. Analysis of African easterly wave structures and their role in influencing tropical cyclogenesis. Mon. Weather Rev. 138: 13991419.
  • Janicot S, Ali A, Asencio N, Berry G, Bock O, Bourles B, Caniaux G, Chauvin F, Deme A, Kergoat L, Lafore J-P, Lavaysse C, Lebel T, Marticorena B, Mounier F, Redelsperger J-L, Ravegnani F, Reeves C, Roca R, de Rosnay P, Sultan B, Thorncroft CD, Tomasini M, Ulanovsky A. 2008. Large-scale overview of the summer monsoon over West and Central Africa during the AMMA field experiment in 2006. Annales Geophys. 26: 25692595.
  • Kiladis GN, Thorncroft CD, Hall NMJ. 2006. Three-dimensional structure and dynamics of African easterly waves. Part I: Observations. J. Atmos. Sci. 63: 22312245.
  • Kuettner JP, Parker DE. 1976. GATE: Report on the field phase. Bull. Amer. Meteorol. Soc. 41: 1127.
  • Landsea CW, Gray WM. 1992. The strong association between Western Sahelian monsoon rainfall and intense Atlantic hurricanes. J. Climate 5: 435453.
  • Locatelli JD, Schwartz RD, Stoelinga MT, Hobbs PV. 2002. Norwegian-type and cold front aloft-type cyclones east of the Rocky Mountains. Weather Forecasting 17: 6682.
  • Martin GM, Ringer MA, Pope VD, Jones A, Dearden C, Hinton TJ. 2006. The physical properties of the atmosphere in the new Hadley Centre Global Environmental model, HadGEM1. Part I: Model description and global climatology. J. Climate 19: 12741301.
  • Njoku E, Koike T, Jackson T, Paloscia S. 2000. Retrieval of soil moisture from AMSR data. In Microwave Radiometry and Remote Sensing of Earth's Surface and Atmosphere. Pampaloni, Paloscia (eds.) 233525.
  • Parker DJ. 1999. Passage of a tracer through frontal zones: A model for the formation of forward-sloping cold fronts. Q. J. R. Meteorol. Soc. 124: 17851800.
  • Parker DJ, Thorncroft CD, Burton RR, Diongue-Niang A. 2005a. Analysis of the African easterly jet, using aircraft observations from the JET2000 experiment. Q. J. R. Meteorol. Soc. 131: 14611482.
  • Parker DJ, Fink A, Janicot S, Ngamini J-B, Douglas M, Afiesimama E, Agusti-Panereda A, Beljaars A, Dide F, Diedhiou A, Lebel T, Polcher J, Redelsperger J-L, Thorncroft CD, Wilson GA. 2008. The AMMA radiosonde program and its implications for the future of atmospheric monitoring over Africa. Bull. Amer. Meteorol. Soc. 89: 10151027.
  • Pytharoulis I, Thorncroft CD. 1999. The low-level structure of African Easterly Waves in 1995. Mon. Weather Rev. 127: 22662280.
  • Redelsperger JL, Diongue A, Diedhiou A, Ceron JP, Diop M, Gueremy JF, Lafore J-P. 2002. Multi-scale desciption of a Sahelain synoptic weather system representative of the West African monsoon. Q. J. R. Meteorol. Soc. 128: 12291257.
  • Redelsperger JL, Thorncroft CD, Diedhiou A, Lebel T, Parker DJ, Polcher J. 2006. African Monsoon Multidisciplinary Analysis—An international research project and field campaign. Bull. Amer. Meteorol. Soc. 87: 17391746.
  • Reed RJ, Norquist DC, Recker EE. 1977. The structure and properties of African wave disturbances as observed during phase III of GATE. Mon. Weather Rev. 105: 317333.
  • Sultan B, Janicot S. 2003. The West African monsoon dynamics. Part II: The ‘preset’ and ‘onset’ of the summer monsoon. J. Climate 16: 34073427.
  • Taylor CM, Parker DJ, Lloyd CR, Thorncroft CD. 2005. Observations of synoptic-scale land surface variability and its coupling with the atmosphere. Q. J. R. Meteorol. Soc. 131: 913937.
  • Thorncroft CD, Hodges K. 2001. African easterly wave variability and its relationship to Atlantic tropical cyclone activity. J. Climate 14: 11661179.
  • Thorncroft CD, Hoskins BJ. 1994. An idealized study of African easterly waves. I: A linear view. Q. J. R. Meteorol. Soc. 120: 953982.
  • Thorncroft CD, Parker DJ, Burton RR, Diop M, Ayers JH, Barjat H, Devereau S, Diongue A, Dumelow R, Kindred DR, Price NM, Saloum M, Taylor CM, Tompkins AM. 2003. The Jet 2000 project: Aircraft observations of the African easterly jet and African easterly waves. Bull. Am. Meteorol. Soc. 84: 337351.
  • Thorncroft CD, Lafore J-P, Berry1 G, Roca R, Guichard F, Tomasini M, Asencio N. 2007. ‘Overview of African weather systems during the summer 2006’. CLIVAR Exchanges 122 1820.
  • Thorncroft CD, Hall NMJ, Kiladis GN. 2008. Three-dimensional structure and dynamics of African easterly waves. Part III: Genesis. J. Atmos. Sci. 65: 35963607.
  • Tompkins AM, Diongue A, Parker DJ, Thorncroft CD. 2004. The African easterly jet in the ECMWF integrated forecast system: 4D-Var analysis. Q. J. R. Meteorol. Soc. 131: 28612885.
  • Zawislak J, Zipser EJ. 2010. Observations of seven African easterly waves in the East Atlantic during 2006. J. Atmos. Sci. 67: 2643.