Agriculture Reference
In-Depth Information
Table 9.7
Correction of nitrogen rate depending on the water situation
Nitrogen supply
High
High
Low
Low
Water supply
High
Low
High
Low
Correction needed
No
No
No
Ye s
curves by using first or second derivatives, Fourier- or wavelet analyses should not
be excluded (see Sect.
9.3.1
).
The use of the red edge inflection point or of red edge ratios near this point for
nitrogen sensing (Table
9.5
) is an example for this trend of
targeted decomposing
of the reflectance curve. Several authors as
e.g.
Filella and Penuelas (
1994
), Yang
and Su (
2000
) as well as Shiratsuchi et al. (
2011
) found that the red edge inflection
points or red edge ratios were not much influenced by the water supply. Among the
various spectral indices that were used for nitrogen sensing, the red edge inflection
points and especially the red edge ratios were the least affected by crop water stress.
However, even if the red edge inflection points and particularly the red edge
ratios are less influenced by water stress, this does not completely remove the prob-
lem. These indices allow only for a look at the supply of the crop in the past days or
weeks. The actual sensing is directed at providing the nitrogen for coming weeks.
And it can be taken for granted that when the crop cannot transpire because of a
starting or ongoing drought, it will not be supplied with nitrogen. Hence the current
supply of nitrogen as well as of water should be sensed separately and simultane-
ously by a
dual sensing strategy
. The prerequisites for doing this are good.
Fortunately, the best wavelengths for sensing nitrogen on the one hand and water on
the other hand are well separated within the spectrum (see Sects.
6.5
and
9.4.3.1
).
How should the site-specific nitrogen rate be adapted to the water situation?
Only the extreme cases can be outlined here. These extreme cases are described by
respectively high or low supply of either nitrogen or water (Table
9.7
).
In places of high nitrogen supply, the crop needs no fertilizer; therefore the water
situation is of no avail for the fertilizing strategy. This also applies, when the sens-
ing signals indicate a low nitrogen supply, while the water situation is good. The
alert situation is, when both the nitrogen as well as the water supply are signaled as
low. Without any information about the water situation, a high site-specific nitrogen
dose would be applied. The dual sensing of nitrogen and water would prevent waste-
ful and perhaps environmentally harmful nitrogen application in this case. However,
this concept of dual sensing is not yet state of the art.
Since it probably can be assumed that the water supply of crops varies temporar-
ily as much as the nitrogen supply does and hence calls for short-termed control
reactions, the best concept would be tractor based
on-the-go dual sensing
. An alter-
native would be a dual control via a combination of mapping the water supply of the
crop by signals from satellites and tractor based on-the-go nitrogen sensing. This
could be a reasonable solution provided the transmission of the radiation is neither
impeded by the atmosphere nor by clouds and the frequency of remote recording
suffices. Unfortunately, the prime candidate for water sensing, namely infrared
