Revision as of 13:03, 15 June 2006; view current revision
←Older revision | Newer revision→
Two basic modelling options
A crop monitoring program has to be based on data from different locations inside the country of region that needs to be monitored.
Technically these monitoring locations are points. Points can be pinpointed at the earth’s surface with coordinates. The coordinate system that is used in crop forecasting is called a geographic coordinate system whereby the coordinates are expressed in Longitude - Latitude pairs.
Example : the geographic coordinates of Pnom Penn (Cambodia) are:
- Longitude: 104.917445
- Latitude : 11.558831
Normally expressed as 104.917445, 11.558831 (longitude always first). Read more: Geographic Coordinates
Depending on the circumstances, the crop forecasting network can be based on:
Modelling type 1. Rainfall stations in Africa
- An irregularly spaced network based on real-world stations (Modelling type 1). As an example below part of the rainfall recording stations in Africa. This type of network has a preference over the second type whenever station data (to run the water balance) are available. The main advantages of this type of monitoring network are:
- Input and output of the water balance model can be checked against the real situation in the field.
- Station weather data are more accurate than gridded data in the vicinity of the stations themselves.
- Usually requires less computer power due to lower number of monitoring points.
- Although the water balance is calculated on a irregularly spaced network, output data can be transformed into grids or images using interpolation techniques. This enables the estimation of water balance parameters for all corners of the monitoring area.
- A regularly spaced grid (Modelling type 2). In this case, grid points do not coincide with meteorological stations. This type of monitoring network is particularly useful if ground data are scarce or inaccurate and weather data with sufficient accuracy are available on a grid or image basis.
Modelling type 2. Grid example for Afghanistan (0.5 degree distance)
Within the crop forecasting tools these locations areWhen setting up a crop monitoring program two basic approaches exist
gridding before modelling, and modelling before gridding. Advantages and disadvantages in terms of errors, labour and accuracy of forecasts.
