Crop Type Mapping and Yield Estimates

Remote sensing technology can be used to prepare maps of crop type and acreage. The use of satellites is advantageous as it can generate a systematic and repetitive coverage of a large area and provide information about the health of the vegetation.

Through scaling procedures it is possible to compare less frequent high resolution data with daily deliveries of inexpensive data with a lower resolution. Interpretations from remotely sensed data can be input into a Geographic Information System (GIS) and combined with ancillary data to provide information about ownership, management practices, crop rotation systems etc.

Optical Data
When using optical data the spectral reflection of a field will vary with respect to changes in the phenology (growth), stage type, and crop health, and these paramters can thus be measured and monitored by multispectral sensors. Water stress indicators and vegetations greenness provide information about the status of vegetation. The biomass of crops (crop yield) can be estimated based on light-use efficiency relations.

Radar data
In contrast, radar data are sensitive to the structure, alignment, and moisture content of the crop and can thus provide complementary information to the optical data. Biomass can also be assessed. Combining the information from radar and optical sensors increases the information available for distinguishing each target class and its respective signature, and thus there is a better chance of performing a more accurate classification.

In countries where agricultural statistics are reliable, the main use of remote sensing is for monitoring purposes. During the growing season, the status of crops can be inspected using satellite images and the need for irrigation and fetilisation can be monitored on an operational basis.

In countries with less reliable agricultural statistics remote sensing is both relevant for monitoring and for updating agricultural database statistics. Through classifications of crop types, the acreage of each crop can be computed with high accuracies. Crop yield can in most cases be estimated with app. 85% accuracy which is adequate in many applications.