20d7b390-4063-11dd-a6a0-0014385046ad Budzanani Tacheba Consorzio ITA Regional Coordinator +267 365 1865 ext 5067

116 Millenium Office Park, Kgale Mews, Gaborone South East BW Botswana btacheba@yahoo.co.uk

2010-03-22T17:23:06 ISO 19115:2003/19139 1.0 Geographic, WGS 1984 2008-06-22 4.0 The yield forecast is realized at (sub-) national level up to the departmental level (or 3rd administrative boundary level e.g. RDP). The yield calculation is based on a yield model obtained from a statistical analysis (stepwise regression) which relates yield to explanatory agrometeorological variables. There are three kinds of explanatory variables. (i) The phenological variables. They are obtained by using the FAO AgroMetShell, a model based on the water balance which gives information on phenological stages of the crop during the growing season. (ii) The meteorological variables which consist of simple and cumulated actual rainfall. (iii) The remote sensed variables which are extracted from NOAA GAC images using WINDISP, an easy FAO image processing software. The models are calibrated on a long term average (for the case of Malawi this was 24 years from 1984 to 2007) and validated by leave-one-out cross validation. The yield forecast at departmental level (RDP) is compared with the average yield for each department on the long term average (in the case of Malawi from 1984 to 2007). The yield forecast map shows those relative yields. The yield forecast at national level which is the required final product, is obtained by aggregating the yield forecast at departmental level. It is indicated on the yield forecast map. The forecasted production is compared with production of the official statistics. Note that the method might slightly differ from one country to another, based upon the local availbility of data and requirements. The GMFS crop yield forecasts are generated once during the growing season. This product is covering Malawi countrywide. The purpose is to forecast crop yield at national level as early as possible in the growing season in order to take appropriate action in case of food crisis. Yield forecasts are an important element for estimating final production of major crops in a given country. These are needed to assess the total for availability in a given country. The yield forecast presented here is an additional method for estimating yields as early in the season as possible, and provides the additional advantage of combining remote sensing data and classical agrometeorological methods. It has the adidtional advantage that it builds upon the longer term effort of the FAO agrometshell software, widely used throughout Africa. Bernard Tychon University of Liège +32 63 23 08 29 +32 63 23 08 00

Bâtiment FUL, 185 Avenue de Longwy Arlon B-6700 Belgium bernard.tychon@ulg.ac.be

Damien Rosillon University of Liège +32 63 23 09 75 +32 63 23 08 00

Bâtiment FUL, 185 Avenue de Longwy Arlon B-6700 Belgium Damien.Rosillon@ulg.ac.be

MWICYSMALL2008.gif thumbnail gif MWICY2008.jpg large_thumbnail jpg Crop Yield Assessments CY Malawi SADC region No Use constraints provided appropriate acknowledgement is given. 1000000 en fr farming 2007-10-01T02:00:00 2008-04-30T02:00:00 32.68187 35.92097 -17.13528 -9.37667 Crop Yield estimates are generated following the procedure described in the document "GMFS_S5 Productsheet Crop Yield" WWW:LINK-1.0-http--link Available through the GMFS ftp site. Request login to Point of Contact http://gmfsgeonetwork.gim.eu/productsheets/GMFS_S5_ProductSheet_CropYield.pdf WWW:LINK-1.0-http--link Product Sheet Crop Yield Assessments