Revision as of 13:55, 10 June 2006; view current revision
←Older revision | Newer revision→

Introduction

Quick overview of FAO food security work, the raison d'être of the crop forecasting and presentation of the overall crop forecasting philosophy adopted by FAO

Introduction into the principles of crop modelling (including basic crop model overview) and their implementation in AgroMetShell and the CMBox. The principle of indicators like ETa and the WSI index.

Presentation of Potential Evapotranspiration (PET) and its role in the calculation of crop water budgets and crop forecasting

Introduction to Remote Sensing (CCD and NDVI) and its role in crop forecasting

5) Introduction to data formats and GIS.

Gathering data and getting them right.

6) The two basic modelling options: gridding before modelling, and modelling before gridding. Advantages and disadvantages in terms of errors, labour and accuracy of forecasts.

7) Selection of reference periods: a compromise between statistical significance and agronomic significance.

8) Practical introduction to Geostatistics and the spatial interpolation of agroclimatic and other variables. This contains a description of LocClim and SEDI

9) Development of practical and simplified PET and radiation computation procedure.

10) Preparation of ten-daily PET maps (36 dekads per calibration year)

11) Preparation of ten-daily rainfall maps (36 dekads/year). If necessary, develop a technique to derive/interpolate rainfall based on Global Telecommunications System (GTS of WMO) and Japanese meteorological satellite images

12) Analysis of time series of climate and crops to identify trends, if they are present. Construction of detrended crop yield time series

13) Preparation of polygons for main crop growing areas in the country and define cropping practices and conditions (planting/transplanting, soil features, irrigation water amounts...)

Using satellite imagery

14) Development of a standard procedure to define actual phenology (in particular crop planting date), based on local practice and satellite imagery

15) Extract Normalised Difference Vegetation Index (NDVI) images for the country from the global data

Running the FAO water balance model

16) Read all data prepared above into the AgroMetShell crop simulation software (AMS)

17) Run AMS for the historical time period, extract average water balance parameters over main crop growing areas 18) Prcatical introduction to multiple regression techniques and the selection of variables through a principal components analysis 19) Calibrate crop yields against water balance outputs and other variables against and validate the coefficients.

Forecasting Yield

20) Using equations derived under 19) above, compute crop yield maps and derive tables of agricultural statistics from the maps (the forecasts)

21) Prepare write-up of the products above as inputs to national crop monitoring bulletins