Agriculture Reference
In-Depth Information
scale for
crop yield
of both
maps
in t per ha
block kriging
60
40
2
20
3
0
4
0
20
40
60
80
100
120
easting in m
5
6
60
point kriging
7
40
8
20
0
0
20
40
60
80
100
120
easting in m
Fig. 2.7
Maps of wheat yields from site-specific harvesting after block- and point-kriging. The
smoothing effect of block-kriging is obvious (From Whelan et al.
2002
, rearranged and altered)
site-specific farming, since it artificially erases local differences and thus eliminates
possibilities to react on them. Yet as long as the kriged blocks are smaller than the
cells of the machinery, any leveling effects of averaging are unavoidable - just
because the size of the machinery does this anyway.
With
point-kriging
, there is no processing error at the actual sampling- or sensing
point. However, this theoretical advantage of point-kriging can only improve the
overall precision, if the farm machinery too can do punctual work. As long as this is
not the case, block kriging will primarily be the choice for creating maps.
The objective is that maps should only show the spatial distribution of the respective
soil- or crop property. However, maps always too reflect the influence of the sampling-
or sensing techniques and in addition of the kriging method that was used (Fig.
2.7
).
Maps from the same basic data therefore can look very different.
The averaging and estimating that is connected with kriging should be ine-tuning
of the spatial resolution as it is needed for the respective site-specific operation.
A fine-tuning by averaging that is oriented at the cell-size of the machine operations
should be the objective. This adjusting of the resolution can take place either during
online and on-the-go control or when maps are processed.
