Map of the world
Agriculture
  Sustainable grazing
  An overview of the parcels
  Radar and crops
  Subsidies for agriculture
  Harvest prediction
  Results
  Team
  Info
Ecology and land management
Humanitarian interventions
Applied Meteorology
Aquatic environment
Regional and city planning
Environmental risks
Global monitoring
Hyperspectral applications
Context
HARVEST PREDICTION
Predicting fat years and lean years

At European level, agriculture experts issue a monthly report setting out the expected harvest results. This is known as the MARS bulletin and is based on two distinct methods: remote sensing and crop growth modelling.

Remote sensing
Within the European context, the development of vegetation during the agricultural season is studied by means of NOAA-AVHRR images. These satellite data provide us with two indicators: surface temperature and NDVI (Normalised Difference Vegetation Index). These indicators directly express the crop situation, thereby permitting comparison between different harvests. Zones with a healthy vegetation have a higher NDVI. This makes it possible to identify any lateness in crop growth or ripening. The figure illustrates the percentage variation in NVDI between 2 years (for July). Zones with a higher NDVI than during the previous year (positive variation, green) and zones where crop growth is less good (negative variation, reddish orange) can be recognised at a glance.


July NDVI: percentage variation of the 1996 value from the 1995 value. Zones where the NDVI is higher than for the previous year are shown in green (positive variation); zones where the crop growth is less good are shown in reddish orange (negative variation).

At European level, agricultural harvests are estimated using two distinct methods: the crop growth model and satellite data. The Belgian project seeks to include these satellite data in the crop growth model.

Crop growth modelling
The Crop Growth Monitoring System, or CGMS, is a spatial crop growth model. The core principle of this model is diagrammatically represented in the figure. The driving force behind plant growth is photosynthesis. This is a process by which sunlight is used to produce essential substances for plant growth and health. The quantity of light a plant can capture is closely linked to the number of leaves (and thus to the LAI or Leaf Area Index). Some of the products produced by photosynthesis are immediately used by the plant for healthy growth. Depending on the growth stage the plant has reached, the remaining products are converted into biomass (leaf, stem, root, storage organs). The CGMS makes it possible to calculate, per crop, the quantity of biomass produced in relation to, among other things, the quantity of sunlight received.
At European level the satellite data and the CGMA are used as separate and independent tools.


The figure herunder diagrammatically represents this general European approach. All EU Member States are included. In the future, the EU plans to test this methodology on non-European countries, in Africa for example. In which case, apart from its economic utility, it could also serve as a detection system for food shortages (comparable to the FAO Early Warning System, but specific to a particular crop and in kg yield per hectare).

Simplified diagram of how a crop growth model functions

We want numbers

The Belgian project is based on the existing European harvest forecasting system, namely the CGMS (Crop Growth Monitoring System), but the databanks are supplemented and refined by inputting typical Belgian physical (such as soil data) and technical (e.g. temperature sum required for flowering, seed formation, etc.) parameters for the crops in question.

The Belgian project is also seeking to incorporate the satellite data in the CGMS as an aid to arriving at a quantitative estimate of production. In the European example, satellite data are used in the sense of "relatively more crop growth in a particular zone as opposed to another" or "relatively less yield in one year compared to another" (qualitative). This project, however, will look at whether these satellite data can be used to arrive at forecasts of the nature of "this region will produce about 2000 more tons of wheat than it did last year" (quantitative). Such forecasts are based on fAPAR time series and LAI estimates from NOAA AVHRR and SPOT VEGETATION images.

MARS Project

Over the past decade every EU Member State has collected and analysed its agricultural statistics in a different way, making it difficult to compare the results at European level. To implement its agricultural policy, the European Community needs timely, accurate and standardised data on agricultural land use, especially for economically important crops for which farmers receive subsidies on the basis of the cultivated surface area.

It is against this background that the Regional Inventory (MARS: Monitoring Agriculture with Remote Sensing) project was set up with the aim of obtaining accurate statistics on agricultural land use within the harvest year, using a standardised method with a known accuracy.
The Regional Inventory includes two parts: field work and remote sensing. The use of satellite pictures increases the efficiency of the surface area estimations.

Another important activity within the MARS project is crop monitoring. CGMS (Crop Growth Monitoring System) is used in combination with remote sensing to compile the monthly MARS bulletins.