(Difference between revisions)
| Revision as of 13:52, 10 June 2006 Peter (Talk | contribs) ← Previous diff |
Current revision Peter (Talk | contribs) |
||
| Line 1: | Line 1: | ||
| - | Introduction | ||
| - | 1) 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 | ||
| - | |||
| - | 2) 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. | ||
| - | |||
| - | 3) Presentation of Potential Evapotranspiration (PET) and its role in the calculation of crop water budgets and crop forecasting | ||
| - | |||
| - | 4) 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 | ||