West African Monsoon 

Modeling and Evaluation

  Forcing and   Boundary Data
  Output Data
  Data/Model Output Access
  Models GCM RCM



1) NCEP Reanalysis 2:


Upper fields and some surface fields necessary for model runs can be found on:


Alternative mirror site:




2) UK Met Office:






UK Met Office SST/Sea-ICE data:



1) Download and unzip datasets :

                                        HadISST1_ICE_1971-2006.txt.gz      (4.5MB)

                                        HadISST1_SST_1971-2006.txt.gz     (30.0MB)

                                        HadISST1_ICE_2007.txt.gz                (0.2MB)

                                        HadISST1_SST_2007.txt.gz               (0.8MB)


2) Read datasets (example) separately both SST and Sea-ICE


3) Use your interpolation software to interpolate from 1x1deg to your models grid (all 36 years)


4) Convert interpolated datasets to 'pp' format (pp format example) both SST and Sea-ICE


5) Use Karl Taylor variance correction method to correct interpolated datasets: karl_taylor.f

More about Karl Taylor procedure at: http://grads.iges.org/c20c/c20c_forcing/karling_instruct.html

In karl_taylor.f code you should change following parameters:

      parameter (nlon=144,
     & nlat=90,
     & mlat=90,
     & abbr='144x90',
     & outpath='./Karl',
     & insst='interp_sst.pp',
     & insic='interp_ice.pp')
nlon, nlat - number of interpolated longitude/latitude points.

If your computer cannot handle larger amount memory needed for this run, try smaller number for mlat, in this case something like 45 or 30 instead of 90 (this will cut domain into 2, or 3 chunks respectively and require less memory to run). Abbr shall just appear as part of filename. Be sure to make

outpath directory before running code. Insst & insic are your interpolated data saved in pp format - input data for this code.


Output file names (among other files) are going to be like:





6) Convert from 'pp' format to your models native format, and data is ready to use, expect of if you need daily SST & Sea-ICE data, you should interpolate this monthly dataset to daily (usually using plain linear interpolation assuming that each month is represented by 15th day of the current month).