key diagnostic, variables and predictands

NOTE:  indexes and key variables have been computed at the
Centre de Recherches de Climatologie (UMR 5120 CNRS / University of Burgundy, Dijon, FRANCE). Any questions and remarks regarding the time-series or figures provided here should be adressed to Nicolas Fauchereau (



  1. precipitation indexes limits
  2. precipitation indexes time-series
  3. time-series plots
  4. hoevmuller plots


The atmospheric monsoon indexes were computed from
NCEP / DOE II dataset. using the following softwares: GRADS, the python language and its CDAT, spherepack, pyclimate, Scientific packages, and some of the plots provided here were done with R. All are "free" softwares, and all the scripts and programs that were used can be obtained by mail (

You can take a look at the
README file for dynamic indexes to know how the things were computed and what does what ...

NOTE: "Derived" daily fields (i.e. specific humidity, relative vorticity, velocity potential, Moist-Static-Energy fluxes and their convergence, atmospheric moisture fluxes and their convergence) at all standard pressure levels can be made available on demand, at either global or "African" scale (defined here as 30W - 30E / 35S - 35N). The fields are provided in NetCDF (COARDS compliant) format. Requests should be adressed to Nicolas Fauchereau (Nicolas.Fauchereau

  1. low-level (i.e. 1000 hPa) descriptors
    1. Moist Static Energy (MSE) gradients
    2. Mean Sea Level Pressure (MSLP) gradients


  1. WAM dynamical indexes
    1. meridional wind (v) at 925 hPa
    2. zonal wind (u) at 925 hPa
    3. zonal wind (u) at 600 hPa
    4. zonal wind (u) at 200 hPa
    5. vertical velocity (omega) at 850 hPa
    6. vertical velocity (omega) at 400 hPa
    7. relative vorticity at 925 hPa
    8. precipitable water
    9. velocity potential at 200hPa
    10. West African Monsoon Index (wind modulus at 925 hPa minus zonal wind at 200 hPa)
    11. convergence of Moist Static Energy, integrated from the surface to 850 hPa
    12. convergence of atmospheric moisture, integrated from the surface to 850 hPa
    13. potential vorticity at 700 hPa ... to be computed soon




Precipitation indexes limits

Rainfall indexes are computed from CMAP and GPCP datasets from 1979 to 2004  for 2 regional and 6 sub-regional domains, the figure below delineates the rainfall boxes:

Precipitation indexes time-series

Time-series are computed as the non-weighted average of precipitation values (mm/day) reported for grid-points inside the box limits. the format is ASCII, each file contains 5 columns, with:

column 1: a real with integer part being the year and decimal part being 1/73 increment (for plotting convenience, e.g. in many languages, plot(file(:,1),file(:,5)) or equivalent should place correct labels on the x-axis)
column 2: integer = number of the pentad of the year (from 1 to 73)
column 3: integer = day of the beginning of the pentad
column 4: integer = month of the beginning of the pentad
column 5: real = mean pentad precipitation rate (mm/day)

You can take a look at the
README file for precipitation indexes to know how the things were computed and what does what ...

regional scale indexes

    sub-regional scale indexes

Time-series plots

These plots show the time-evolution of CMAP (black) and GPCP (red) precipitation (raw data, including the annual cycle, in mm/day).

The bottom plot on each figure is the difference (CMAP minus GPCP), allowing to identify disprecancys and outliers in the two datasets.


Hoevmuller plots

Hereafter are presented hoevmuller time-latitude diagrams averaged over 10W - 10E for each year