Crop Radar successfully measured maize harvest 2019 in Germany

low overall losses, but strong regional differences

Kleffmann CropRadar passed the next challenge: The system measured that the total area under maize cultivation in Germany was 2.598 million ha at harvest time this year in Germany. The difference to the cultivated area is therefore only 2.04% - but there were clearly visible regional differences.

The detected CropRadar area for the total maize acreage in Germany at harvest time amounts to 2.598 M ha. It varies only slightly from the predicted sown area despite arid conditions. Already in spring, the Federal Statistical Office published the value of 2.652 M ha cultivated maize for Germany. This represents a loss rate to the CropRadar Data of only 2 percent.

However, the striking difference between East Germany and the rest of the country is easily seen.

Figure 1: Marked in red: Federal States with the highest percentage variance between the sowing area and the remaining maize area at the time of harvest. For detailed information concerning exact numbers please visit Kleffmann homepage – products & services – CropRadar.

In particular, the federal states of Saxony, Saxony-Anhalt, Brandenburg and Thuringia were affected. In this context, the most significant differences can be observed in Saxony-Anhalt. Here, the farmers have to cope with a damage of approximately 13% of the cultivated area. The massive losses can first of all be traced back to considerably smaller plant height as well as large-scale gaps in the overall vegetation cover. Visual aids for poor plant development can be seen in the following figures.

Figure 2: Poor condition of many maize fields in the East of Germany. Date: 01.08.2019

Figure 3: Usual condition of maize fields in Germany at this time of the year.

As already forecasted in the previous newsletter, poor maize plant growth essentially complicated the precise CropRadar assessment of the maize cultivation area for this season. This is mainly attributable to the delayed growth of the plants and, consequently, to the poorly developed visual characteristics which normally help Kleffmann experts to distinguish between the varied crop types. For this reason, the whole classification process took a comparatively long time and required a very fine and detailed way of working. Nevertheless, a thorough validation enabled Kleffmann Group to release a verified and realistic outcome.

The next challenge for Kleffmann CropRadar team will be the identification and classification of oilseed rape fields in the whole of Europe by the end of December. One of the focus areas will be the implementation of different new and enhanced processes which are being developed and implemented by the in-house remote sensing development team. Consequently, the application of improved machine learning algorithms became a higher priority.

Author:

André Hinsenhofe, Junior Project Manager GeoSystems