AMIP II Diagnostic Subproject 20

Model Evaluation on the West African Monsoon
Project coordinator:
Serge Janicot and Jan Polcher1
Chris Thorncroft2
Henri Laurent3
Thierry Lebel4
1Laboratoire de Meteorologie Dynamique du CNRS
2Reading University
3ORSTOM, Montpellier
4ORSTOM, Grenoble

Data Requirements
Validation Data


The general large-scale features and their seasonal evolution are reasonably well known. They are characterized during the rainy season in northern summer by an inflow of moisture from the ocean (Cadet and Nnoli, 1987). At higher levels, 600/700 hPa, a strong easterly jet is present. The AEJ has been shown to be barotropically and baroclinically unstable and supports the growth of African easterly waves with characteristic frequencies of 3 to 5 days (Reed et al., 1977; Thorncroft and Hoskins, 1994a; Thorncroft and Hoskins, 1994b; Thorncroft, 1995). Waves with characteristic frequencies between 6 to 9 days have been also observed (de Felice et al., 1990). The evolution of convection during the West African monsoon is much less known. Large convective systems were found to travel from east to west but it is not yet clear how they are linked to the instabilities of the easterly jet (Duvel, 1990). The link between these large scale systems and the more mesoscale squall line has still to be established (Barnes and Sieckman, 1984). These large convective systems are an important component of the West African monsoon as they bring about 70 percents of the total annual rainfall. The West African region is one of the regions of the world most exposed to the interannual variability of climate. It has experienced a severe drought over the last 30 years. Two main explanations have been put forward to explain this sensitivity. First Charney (1977) suggested that land surface changes might have affected the evolution of the monsoon. The second explanation attributes these rainfall anomalies to variations of the sea surface temperatures (Rowell et al., 1995).

New studies have shown that the interannual variability of rainfall in the West African region is characterized by a strong reduction in the frequency of intense convective systems while the number of small and medium systems remains largely unchanged (Le Barb? and Lebel, 1997). Interestingly this sensitivity is reproduced by AGCMs in experiments where land-surface conditions have changed (Polcher, 1995). These new results and the already well known features of the climate of West Africa will be used to evaluate the ability of current AGCMs to model the monsoon and its interannual variability. Special attention will be paid to the characteristics of the variability of convection with the aim of verifying the processes modeled by AGCMs. This analysis subproject will be part of the work proposed to the European Commission in the project West African Monsoon Project (WAMP) coordinated by Chris Thorncroft.


The aim of this subproject is to evaluate the ability of the atmospheric general circulation models participating in AMIP to simulate various aspects of the African monsoon. This will include an examination of intraseasonal, seasonal and interannual variability. Special focus will be given to precipitation, easterly waves, the African easterly jet (AEJ) and convective systems. A number of mechanisms have been proposed to describe the evolution of the African monsoon but very few consider the interactions between the dynamics and convection. The interannual fluctuations of rainfall in this region have been in turn attributed to land-surface processes and variations in the sea surface temperatures. With this comparison of AGCMs, an attempt will be made to answer the following questions :

  • How well do models simulate the West African monsoon climatology ?
  • Which African waves are represented by the models ?
  • Do the dominant frequencies found in rainfall vary from one model to the other ?
  • Are convection and dynamical waves linked in similar ways in all models ?
  • Is the interannual variability of these processes represented by all models ?

An effort will be made to develop indices or other simple diagnostics which can easily be applied to all models. Then the analysis will be refined on those models which are representative of the largest number or show special features.

3.1 Climatological fields
The climatological fields of rainfall and dynamical/thermodynamical variables will be examined over West Africa, especially in July-September, and compared with observational data and re-analyses. Rainfall indices will be computed for different areas (Western and Eastern Sahel, Soudan, Guinea Coast) as well as their seasonal cycle. These indexes will be used to define rainfall patterns over West Africa for each summer (Janicot, 1992). A West African Monsoon Index WAMI, as difference between wind at 200 hPa and at 925 hPa (Fontaine et al., 1995) will be used to characterize monsoon dynamics in the different GCM. This index will be correlated with SST and West African rainfall patterns. Several horizontal and vertical cross-section mean fields in July-September will be inter-compared, describing the main features of the wind fields and humidity fields (monsoon ux, AEJ, TEJ, moist static energy). Potential vorticity will be computed for the analysis of barotropic and baroclinic instability conditions linked to the development of easterly waves.

3.2 African easterly waves
The easterly wave activity in the AGCMs will be investigated and compared. This will include a study of the nature of the waves including their period, wavelength and track, as well as their relationship with rainfall. It is likely for example that the relationship with model rainfall will vary from model to model depending on the convection scheme used. The relevance of such differences to simulations of the WAM will be investigated. One important focus of this work will be to identify the relationship between interannual variability of the WAM rainfall and easterly wave activity. The African easterly jet will also be an important focus in this study.

3.3 Frequency of convective systems
As it was done with in-situ observations (Le Barb? and Lebel, 1997) convective systems will be described and binned according to the precipitation they produce. The frequency distribution can then be compared to the observational record in order to validate the mean climate and it's interannual variations. This measure
 will also be used to compare the models. It is expected that the frequency distribution will be related to the convection scheme used in the AGCM but it will also depend on the quality of the simulated African monsoon.

3.4 Synthesis : the interannual variability
All these diagnostics will then be compared to determine if models behave similarly in each of them. These intercomparisons will indicate if relations exist between these diagnostics or if on the contrary models failing one might do well in another. At this point the interannual variability simulated by the models for the three topics described above will be compared in relation to SST anomalies. This will yield information on the relation which might exist between these three aspects of the African monsoon. For instance is the interannual variability in the frequency of convection related to that of the easterly waves or is it correlated in similar ways to the SST anomalies as a mean precipitation index ?

Data Requirements

The data is only required over a limited area of the globe. This is the box 50W to 90E and 30N to 15S.

Table 1a : northward wind, eastward wind, vertical motion, air temperature, geopotential height and specific humidity, from 1000 hPa to 100hPa.

Table 1c : temperature tendencies due to total diabatic heating, SW radiation, LW radiation, moist convective processes, and dry convective processes, moisture tendencies due to total diabatic processes and convective processes, cloud fraction, from 1000 hPa to 100 hPa.

Table 2 : ground temperature, surface (2m) air temperature, mean sea level pressure, total precipitation rate, convective precipitation rate, surface specific humidity (2m), surface sensible heat flux, surface latent heat flux, eastward surface wind stress, northward surface wind stress, surface incidence SW radiation, surface reflected SW radiation, surface downwelling LW radiation, surface upwelling LW radiation, OLR, total cloud amount.

Table 3 : northward wind at 850 hPa, eastward wind at 850 hPa, OLR, total precipitation rate.

Table 6 : air temperature at 850 hPa and 500 hPa, specific humidity at 850 hPa and 500 hPa, vertical motion at 500 hPa, perceptible water. The data will be archived at LMD and from there redistributed to all participants.

Validation Data

NCEP reanalysis data are available at LMD, and ECMWF reanalysis data at IPSL (France). Daily raingauge data covering West Africa are available at ORSTOM on the AMIP period. Finally satellite data like ISCCP are available at LMD.


Barnes, G. and Sieckman, K. (1984). The environment of fast and slow-moving tropical mesoscale convective cloud lines. Mon. Weather. Rev., 112,1782{ 1794.

Cadet, D. and Nnoli, N. (1987). Water vapor transport over Africa and the Atlantic ocean during summer 1979. Quart. J. Roy. Meteor. Soc., 113, 581{ 602.

Charney, J., Quirk, W., Chow, S., and Korn_eld, J. (1977). A comparative study of the effects of albedo change on drought in semi-arid regions. J. Atmos. Sci., 34, 1366{1385.

de Felice, P., Viltard, A., Monkam, D., and Ouss, C. (1990). Characteristics of north African 6-9 day waves during summer 1981. Mon. Weather. Rev., 118, 2624{2633.

Duvel, J. P. (1990). Convection over tropical Africa and the Atlantic ocean during northern summer. Part II : Modulation by easterly waves. Mon. Weather. Rev., 118:1855{1868.

Fontaine, B., Janicot, S., Moron, V. (1995). Rainfall anomaly patterns and wind _eld signals over West Africa in August (1958-1989). J. Climate, 8:1503{ 1510.

Janicot, S. (1992). Spatio-temporal variability of West African rainfall. Part I : Regionalization and typings. J. Climate, 5, 489{497.

Le Barb, L. and Lebel, T. (1997). Rainfall climatology of the HAPEX-Sahel region during the years 1950-1990. J. Hydrol., in press.

Polcher, J. (1995). Sensitivity of tropical convection to land surface processes. J. Atmos. Sci., 52(17):3143{3161.

Reed, J., Norquist, D., and Recker, E. (1977). The structure and propagation of African wave disturbances as observed during phase III of GATE. Mon. Weather. Rev., 105:317{333.

Rowell, D. P., Folland, C. K., Maskell, K., and Ward, N. (1995). Variability of summer rainfall over tropical North Africa (1906-92) : Observations and modelling. Quart. J. Roy. Meteor. Soc., 121:669{704.

Thorncroft, C. and Hoskins, B. (1994a). An idealized study of African easterly waves. Part I : Alinear view Quart. J. Roy. Meteor. Soc., 120:953{982.

Thorncroft, C. and Hoskins, B. (1994b). An idealized study of African easterly waves. Part II : A non linear view Quart. J. Roy. Meteor. Soc., 120:983{1015.

Thorncroft, C. (1995). An idealized study of African easterly waves. Part III : More realistic basic states Quart. J. Roy. Meteor. Soc., 121:1589{1614.

For further information, contact the AMIP Project Office (
Last update: 21 June 1999.  This page is maintained by

LLNL Disclaimers