In recent years, an increasing amount of research has been devoted to the
delineation of management zones. There have been quite a number of approaches
towards using small-scale data for subdividing the field into a small number
of zones, usually three or four. However, these zones are usually static,
often require multi-year data sets and are based on low-resolution sampling
methods for data acquisition.

Furthermore, existing research into the semi-automatic generation of
management zones usually uses a variety of clustering methods to subdivide the
field into homogeneous parts. However, most of the existing approaches
use only parts of the available data for the clustering process.

Therefore, we propose a novel approach which tackles the above issues. We base
our work on a site-year of high-resolution spatial data from selected wheat
fields in Germany. The available attributes are the following: previous year's
yield, three givings of nitrogen N1, N2, N3, apparent electrical conductivity
EC, remote sensing information REIP32, REIP49 and the current year's yield,
each at a real spatial resolution of 100 data samples per hectare. Given these
data, we present an improved meta-clustering approach to the delineation of
management zones. We propose to investigate how the zones change due to
external influences, especially weather, throughout the growing season as new
data from remote sensing instruments is made available. We investigate the
intercorrelation between data attributes in and in between the different
management zones. Our approach is likely to be implemented on fields in
Germany throughout 2011.