Home > CGMS 9.2 database > Database compilation - Data collecting

• Introduction
• Thematic description of CGMS level 1
• Thematic description of CGMS level 2
• Additional functionality in the CGMS software for yield forecast
• Overview of different CGMS versions
• CGMS 9.2 database
• Table by table
• Relation between tables
• Database compilation
  • Data collecting
  • Recommendations for building tables
• CGMS User Interface
• References
• Annex A Using CGMS in batch mode
• Contacts
• Legal Notice

Database compilation

Data collecting

The first step to the CGMS database compilation is to collect necessary data.

Weather

The length of weather time series should be a minimum of 15 years weather data, as a basis for quantifying the annual deviations from long term averages. Technically the minimum length of time series is three years. The following data are required:

• Station number
• Calendar day (dd-mm-yyyy)
• Daily sunshine duration (2) (hours)
• Daily global radiation at surface (1) (KJ.m^-2.d^-1)
• Minimum air temperature (4) degrees Celcius
• Maximum air temperature (4) degrees Celcius
• Daily mean vapor pressure (hPa)
• Daily mean wind speed at 10 m height (m.s^-1)
• Daily rainfall (mm.d^-1)
• Daily mean of total cloud cover (3) (oktas)

To derive the global radiation four methods are available, in preferred order (the numbers that occur in the list of data above relate to one of these 4 methods):

1. Measured global radiation
2. Derive global radiation from daily sunshine duration and Angstrom formula
3. Derive global radiation from daily mean of total cloud cover and Supit formula
4. Derive global radiation from temperature and Hargreaves formula

One of the three data sources (measured, sunshine, cloud cover) should be available, otherwise temperature is used. Thus the minimum basic set is temperature, vapor pressure, wind speed and rainfall.

When data are not available, they could be estimated from other data. For example there exist formulas to convert wind speed from one height to another height. Also formulas to derive vapor pressure from dew point temperature, relative humidity or wet bulb temperature (Drift and Van Diepen, 1992).

It is also necessary to collect meta information about the weather stations:

• Latitude (decimal degrees)
• Longitude (decimal degrees)
• Altitude (m)

GIS data

The following GIS-data should be available:

• Vector coverage or grid with meteorological grid boundaries and the altitude.
• Vector coverage  of soil map.
• Vector coverage with administrative boundaries.
• Not necessary but recommended: a vector coverage of the spatial distribution of the crops.

Soil data

The soil map is used to determine the maximum available water which a soil can supply to a plant during a period of drought. Further, if crop masks are missing the soil map can be used as a "land use probability map" to define which crops have to be included in the crop growth simulation for a given soil unit.

Crop data

For each crop (and each crop variety) for each agricultural zone for each year (as many years as possible) the following data is needed:

• Sowing date (ddd)
• Emergence date (ddd)
• Leaf area index: maximum value and when reached
• Flowering date (ddd)
• Maturity date (ddd)
• Grain yield (kg.ha^-1.yr^-1)

The minimum set of data is the start of the crop calendar thus sowing or emergence data. But if calibration of the crop parameters is required more data are needed like the list given above.

For the linkage of simulated crop yields to regionally observed crop yields it is necessary to have a time series yield statistics (at least 10 years). It may be helpful to have soil data, irrigation data and fertiliser use data on the regions as well. These data are used for the crop file parameters calibration.