Description of CPC/FEWS Africa Rainfall Estimation Archive, RFE Version 1, June 1995 to December 2000

Original documentation can be found in the 10day_readme.txt file at

Also see: CPC/FEWS NET and the document Objectively Determined 10 Day African Rainfall Estimates Created for Famine Early Warning Systems

Please note that the estimated precipitation values for Africa are given in terms of 10-day (dekadal) accumulations, and in the Data Library dekads are designated by the central day of the dekad (e.g., 6 Aug 2000 = 1st dekad of August 2000).

The following text is excerpted from the "readme" file referred to above:

This file contains readme information for archived 10-day Africa precipitation estimate data from the
CPC rainfall algorithm (version 1), 1995-2000.  The algorithm used to create these estimates was replaced
on January 1, 2001 and is no longer operational.

The files in this archive are available in .bil and .dat format with a Fortran90 program included to
convert .dat files to comma delimited ascii.  From 1995-2000, each 10-day rainfall estimate covers the 
domain from 20W-55E and 40S-20N with a resolution of 0.1 degree.  The file naming convention is as follows:

'afpfinalcalYYMMD.bil' or 'afpfinalcalYYMMD.dat'
    where YY is the last 2 digits of the year
          MM is the month (1..12)
          D is the dekad (10-day period) of the month (1=day 1..10; 2=day 11..20; 3=day 21..end)

Each file contains rainfall estimates over the domain in a (751x601) array in 16-bit integer (.bil)
and floating point (.dat) binary format.

These rainfall estimates were created by the Climate Prediction Center for the United States Agency for
International Development (USAID) Famine Early Warning System (FEWS) project to assist in drought
monitoring and flood forecasting efforts throughout Africa.  This rainfall algorithm (RFE1.0) uses infrared
temperature satellite data, rainguage data, and modeled wind and relative humidity data to compute ten day
rainfall estimates.  Meteosat-5 IR temperature data is first used to compute estimated rainfall via the 
Goes Precipitation Algorithm (GPI).  MOdeled relative humidity and wind data is then compared to 
topographical data to estimate cross-terrain flow as orographic precipitation.  These two estimates are
then compared to Global Telecommunications System (GTS) rain guage measurements, and ground truthing is
performed to remove bias and create the final rainfall estimate.  See: for further explanation.

For additional information contact:

Tim Love
Climate Prediction Center
301-763-8000 x7549