Agriculture Reference
In-Depth Information
mainly because at the time of the first dressing, the crop canopy is not yet closed.
Even if an indication of biomass and chlorophyll content could be obtained at the
time of the first dressing, the very small plants often do not provide a reliable esti-
mation of the nutrient supply from the soil. Because initially the germinating and
emerging plants obtain their nitrogen from the seeds and not from the soil. Therefore,
differences in the supply from the soil might not well show up with small grains at
a developmental stage below EC or BBCH 25 (less than 4 tillers).
Apart from this, for
not closed crop canopies
it is important to exclude the influ-
ence of soil on the reflectance as much as possible in order to avoid mixed signals.
Two methods can be used for this, either
narrowing the view
of the vertically ori-
ented sensor to a closed canopy strip exactly above and along a row, or using an
oblique view
that is oriented perpendicular to the row-directions. The first method
- narrowing the view - can only be successful if the canopy strip above the row is
really closed. The second method - oblique viewing - relies on getting the signals
mainly from the upper part of the crop. The more oblique this field of view is, the
more the reflectance received comes from the crop instead of the soil.
For most crops, the
minimum growth stage
for avoiding mixed signals is lower
with oblique viewing than with vertical viewing. So for small grains, the minimum
growth stage with oblique viewing is about EC or BBCH 30, which is the beginning
of stem elongation (Meier
2001
). With vertical viewing it is about EC or BBCH 32,
which is stage node 2 (Schmid and Maidl
2005
). A similar effect of the viewing
direction on the minimum growth stage that is needed has been observed for maize
(Bausch and Diker
2001
).
However, oblique- instead of vertical viewing also means that the spectrum might
saturate faster with increasing leaf-area-indices of the crops (see Sects.
6.2
and
9.4.3.1
). Since this
saturation effect
reduces the sensitivity in measuring, reliable
oblique viewing depends particularly on reflectance indices that do not tend to satu-
rate fast. Consequently, accurate nitrogen sensing in lush crops with oblique viewing
requires avoiding wavelengths from the main absorption regions, hence from the
blue and especially from the red range. With proximal sensing, this practice presently
is state of the art. The poor results of the red Normalized Difference Vegetation Index
(NDVI) as defined in Tables
9.3
and
9.5
for nitrogen sensing might be even worse
when the viewing occurs from an oblique- and not from a vertical direction.
With natural illumination, oblique viewing also increases the noise due to the
varying solar azimuth angle. Yet an effective method of eliminating this noise -
apart from using artificial light - is averaging signals coming from several and
opposite directions (Fig.
9.23
). The original Kiel system (Figs.
9.1
and
9.25
) now is
operating in practice with oblique viewing by sensors positioned on the roof of the
tractor relying either on solar or artificial illumination.
9.4.5
Sensing Nitrogen by Fluorescence
