Comr,lission VII, W. G. VII/2
H.-J . BOEHNEL, W. FISCHER, G. KNOLL,
Fraunhofer - Institut fur Physikalische Messtechnik,
(former Institut fur Physikalische Weltraumforschung)
Freiburg i . Br . , Federal Republic of Germany
A. KADRO,
Abt . Luftbildmessung und - Interpretation der Univ. Fr e iburg ,
Freiburg i . Br . , Federal Republic of Germany
DIFFERENCES IN THE SPECTRAL CHARACTERIST I CS BETWEEN HEALTHY AND
DISEASED CROPS DETERM I NED FOR SUGAR BEETS AND WIN TER BARLEY .
PAR T B: IN SITU MEASUREMENTS WITH SPECTRORADIOMETERS ,
Abstract
Radiometric measurements from 0 . 5 to 1.1 urn and from 1. 5 to
2 . 4 urn were conducted on sugar beets during two campaigns in
July and Sept ember 1979 at three controlled test sites in Italy
showing infestations b y nematodes , virus (rizomania) , and
fungus (cercospora leaf spot) , respectively . Additional me asure ments could be performed at a common farmland in Germany at
which the sugar beets were partly stressed by water deficiency
and nematodes . Measurements of barley from 0 . 5 to 1. 1 ,.urn were
carried out in 1979 at a German test site consisting of a field
plot treated against mildew, fungus , etc . and a non - treated
reference plot at intervals of two weeks during the last two
months before harvest . The spectral reflectance factors were
calculated and compared with the corresponding values det e rmined
from me asurements in the laboratory and with a multisp e ctral
scanner (Parts A and C of the joint paper) .
Introduction
Measur e ments with ground - based spectroradiome ters on sugar be e ts
and b ar ley as d e scribed in the abstract were carried out in the
frame of a research project on t he detection of vegetation
stress by changed reflection characteristics . The selection of
the sugar beet test sites in Rovigo, San Pietro and Anzola
(near Bologna) and the field wo r k in Ita l y were performe d in
cooperation with the Joint Research Centre / Ispra , the University
of Milan and an agricultural institute (Istituto sperimentale
per le colture industriali) of Bologna . The measurements on
2
barley could be conducted at selected test plots of 10 x 10 m
within a large field of the Landwirtschaftliche Versuchsslation
Limburgerhof of the BASF . One of these plots was growth - regulated and treated against mildew , fungus , etc . whereas at the
refer e nce plot only some treatme nt against weeds has been
appli e d . The instrumentation , the methodology and the calcula tion of the spectral reflectanc e factor R( A) as a quantity for
the characte rization of the r e fl e ction properties of vegetated
surfaces are described in another paper and will not be pre sented here /1/ .
092.
Results of the measurements on sugar beet3
The diagrams of R(\) versus \ of healthy sugar beets show the
characteristic response of vital vegetation (Figures 1 and 2) .
Observation at different zenith angles ~r leads generally to a
splitting of the R(\) - curves corresponding to~r which depends
on the azimuth difference between the incident solar radiation
and the observation direction (and also the zenith angles of
the sun and the instrument) . This splitting is higher for front
lighting/back lighting conditions than for side lighting .
Figure 10 illustrates the effect of front lighting (i . e . sun at
the back of the observer) . The highest values of R ( \ ) were
obtained for~ ~ ~ .. This is true for the whole spectral range
from 0 . 5 to
rurn: Back liqhting
·
o results in decreased values
compared to the result for~r = 0 . A further advantage of side
lighting with respect to remote sensing applications is the
symmetry of the resulting reflectance factors for both azimuth
directions perpendicular to the solar azimuth . The following
discussions are related to side lighting conditions and to com parable geometrical parameters for those measurements which
have been directly compared .
2.S
The spectral differences between healthy and diseased sugar
beets are now discussed in detail . The reflection properties of
diseased sugar beets are determined by changes of morphology and
leaf positions as well as reduced soil coverage . Examples of the
reflectance factor curves of soil are shown in Figures 1 and 7 .
The spectral response of diseased sugar beets compared to the
healthy test samples is different in the visible and the middle
infrared,where R(\) is higher for the diseased beets , and also
in the near infrared with higher values of R( \ ) for the healthy
beets . This corresponds to different types of processes which
determine the reflection properties of plants in the pigment absorbing region (0 . 4- 0 . 7 pm), the internal - leaf - structure
region (0 . 7 - 1 . 3 pm) , and the leaf - moisture - content region (1 . 3 2 . 5 pro) /2/ . In all investigated cases a clear distinction is
possible between both of them . The highestsensitivity for the
spectral change occurs at the 675 nm minimum and at the 2 . 120 pm
maximum . The increased splitting of the curves corresponding to
the variation of~ for the diseased sugar beets can be seen
very clearly in Fi~ures 3,7 and 9.
Infestation by nematodes
(ROVIGO and ARMSHEIM) :
As illustrated in Figures 1,2 and 3 for Rovigo there occurs a
remarkable difference between diseased and healthy sugar beets .
R(\) of the diseased canopy is about 100 % higher at 675 nm and
2 . 120 pm than the corresponding values of healthy beets (for
vertical observation) . This could be explained by the super position of two effects . Firstly the soil reflectance causes an
intensification in the visible and the middle infrared but a
decrease in the near infrared . Secondly this type of disease
affects the turgor of the leaves which become slacky so that
large petiole areas are exposed . Petiole reflectance is higher
(about 100 %) in the visible and lower (about 30 %) in the near
infrared than that of leaves /3/ . The loss of turgor is obviousl y
also associated with an increased reflectance factor in the
093.
leaf- moisture - content region . Sugar beets at Armsheim also
stressed by water deficiency because of nematodes and a subsur face gravel ground layer showed a similar behavior .
Infestation by rizomania and cercospora leaf spot (SAN PIETRO)
Figure 4 illustrates the difference in the spectral response
between the Rovigo and the San P i etro test site in the wave length range 0 . 5 to 1. 1 prn . In San Pietro the soil coverage by
the plants was somewhat higher . The lower reflectance of the
beets at this site at 675 nm is probably due to the smaller
portion of soil in the field of view and the lower contribution
of the petioles to the signal since the slacking of the leaves
seemed to be not so strong as in Rovigo . The higher values at
the green maximum could be caused by the beginning discoloration
of the leaves by light- brown leaf spots (a similar effect as for
yellow mosaic virus /3/) .
In September the ratio of R(A) between the diseased and the
healthy canopy amounted to a factor of about 3 at 675 nm and
2 . 120pm (as shown in Figures 5 , 6 and 7) due to the strongly
reduced soil coverage of the diseased plants .
Infestation by cercospora leaf spot (ANZOLA)
In the whole investigated wavelength range lower values of R(A)
of healthy sugar beets were obtained in Anzola compared to
San Pietro . This is shown in Figure 5 for the visible and the
near infrared . The same Figure indicates differences in the
spectral response of the diseased beets at both test sites only
in the infrared with higher values for San Pietro (see also
Figures 7 and 8) . This is probably due to different soil prop erties since the soil coverage by the plants was less than 50 %
in both cases .
Results of the measurements on winter barley
Similar to the measurements on sugar beets the R(A) curves are
split for different observational zenith angles~r · Also the
influence of the azimuth difference between the sun and the
instrument has the same effect as described above . Contrary to
the sugar beets the splitting of the curves (for side lighting)
is clearly stronger for the healthy field and only significant
in the near infrared as illustrated in Fig . 11 .
Measurements in the middle of May , shortly before the appearance
of the ears , showed no remarkable spectral difference between
the treated and the untreated field although the plant height
was not equal but 75 and 90 em , respectively .
Two weeks later after the appearance of the ears the spectral
reflectance factor of the diseased barley was about 20 % higher
in the green at 555 nm and in the near infrared at 860 nm , but
practically equal in the red minimum (675 nm) as shown in
Fig . 1 2 . The Figure represents the average of all the measure ments carried out with~r = 0° between the late morning and the
early afternoon , because there were only slight differences due
to the variation of the solar position .
In the middle of June R( A) of the healthy barley was slightly
lower than R( A) of the diseased barley in the visible , but
slightly higher in the near infrared . Visually no difference in
the color of both targets could be observed .
At the end of June , shortly before
response has strongly changed (see
reflectance factor of the diseased
(about 7 %) and 860 nm (about 18 %)
field , but lower at 675 nm (11 %) .
the harvest , the spectral
Fig . 12) . The spectral
canopy was higher at 555 nm
, than R(A) of the healthy
The assistance of Dr . C . MUNTHER in the preparation of computer
programs and the making available of barley test sites by the
BASF Landwirtschaftliche Versuchsstation Limburgerhof are grate fully acknowledged . The investigations in Italy were carried out
on behalf of the JRC Ispra in cooperation with the University of
Milan and the Istituto sperimentale per le colture industriali
of Bologna . Financial support was given by the Deutsche For schungsgemeinschaft and the JRC .
Sr
1-
z
60
w
u
=
0°
healthy
July 18, 1979
. . ··.· ...·····.·
Si=32°
IPj=52°
.'
0::
w
a_
......
'·
50
....··· ·· ..
0::
0
1-
u
<(
lL
..
......
. ..
.. ,•
40
July 15, 1979
diseased
w
u
z
<(
1-
u
w
30
_j
lL
w
0::
_j
soil ( S. PIETRO)
July 20, 1979
20
<(
0::
1-
u
w
a_
(J')
10
healthy
600
SUGAR BEETS
700
aoo
ROVIGO
Fig.1
095.
900
1000
'WAVELENGTH I NM
f-
z
w
u
0::
w
30
SUGAR BEETS. HEALTHY
ROVIGO I JULY 11. 1979
SIiJE LIGHT I:iG
s, = 24°
0...
0::
~
u
25
~
w
u
z
20
<!
o
Sr
•
Sr =
=
f-
u
w
_J
15
+ Sr=
LL
w
- Sr=
0::
oo
15°
30°
45°
..J
<!
0::
f-
u
w
a...
Vl
SUGAR BEETS
FIG. 2
f-
z
w
u
0::
w
0...
a::
0
f-
u
<!
LL
w
u
z
<!
f-
u
w
_J
LL
w
a::
..J
<!
a::
fu
w
0...
Vl
I
I
I
SUGAR BEETS. DISEASED
ROVIGO I JULY 16. 1979
SIDE LIGHTING
.Ji = 24°
]
J
.::0
I
;;·/~... ·
; ·
rr' /
I
~......
~
15 '---
J
/
•
Sr=
15°
30°
45°
~~,
:.------. --.~
c::::-~~
.r'
..!/
jiill
li
..
~
1.500
oo
- Sr=
1 .:'
J
Sr =
+ Sr=
/-~~
/
o
1.625
1750
1875
2.000
--'-o-¥
2.125
2.250
2.375
WAVELENGTH I f.J.m
SUGAR BEETS I DISEASED
FrG. 3
096.
I
l
!
l
. ..... RO VI GO
t-
z
w
60
July 15, 19 79
- - S. PI E TRO July 20 , 1979
u
a:::
w
0...
50
a:::
0
t-
u
<(
u...
w
. .• ..
40
u
z
<(
t-
u
w
JO
_J
u...
w
a:::
_J
20
<(
~~ =
25°
(/lj =230
a:::
t-
u
w
0...
(f)
10
600
SUGA R BEETS
700
aoo
900
1000
IJAV ELE NGTH I NM
DI SEASED
F ig L.
······ S. PI ET RO
t-
z
w
a:::
w
.......
~ ~ = 39 °
(/l j~ 2 1 °
A N ZOLA
Sep! B /10, 197 9
~~
(/l j= 20
60
u
0...
Sept. 3 /L. , 19 79
= 39 °
... ··
50
a:::
0
t-
u
<(
u...
w
40
u
z<(
t-
u
JO
_J
20
w
_J
u...
w
a:::
<(
a:::
t-
u
w
0...
(f)
10
600
700
aoo
9 00
1000
IJAVELENGTH I NM
SUGAR BEE TS
F1g 5
097.
SUGAR BEtTS, HEALTHY
SA~ PIETRO I SEPT. 2, 1979
SIDE LIGH TING
Si= 37°
a
.~~
oo
j
+ Sr= 30°
~:_:->-,~\\
;
Sr =
• Sr = 15°
- Sr = 45°
'
-
//
r.:··
lJ
1.500
l525
l750
l875
2.000
2.125
2.250
2.375
WAVELENGTH IJ-1m
SUGAR BEETS
FIG'
5
1-
z
I
l1J
u
5CL
a::
~
u
11_
SOIL
30
I
L~
-<
SOIL
I
I
25
l1J
u
z
20
<{
1-
u
l1J
_.J
lL
l1J
a::
SUGAR BEETS,
SAN PIETRO I SEPT. 6, 1979
SI DE LIGHTING
...J
<{
a::
1-
u
l1J
CL
si=
Vl
1.500
l525
l750
l875
2.125
2.250
2.375
WAVELENGTH I J-lm
SUGAR BEETS I DISEASED
F!G' 7
098.
i
1
38°
2.000
l
1-
z"
w
u
a::
w
a..
BEETS. DISEASED
I SEPT, 7, 1979
SIDE LI GHTL~G
SUG~R
a::
Af~Z O LA
0
1-
u
l1.
_s., ~ 39°
w
u
..:
z
<{
1-
u
w
_J
lL.
w
a::
....)
<{
a:::
1-
u
w
a..
U1
1.500
1.625
1750
1.875
2.000
2.125
2.250
2.375
WAVELENGTH
IJ.i.m
SUGAR BEETS I DISEASED
FIG'
1-
z
w
u
a:::
w
a..
8
60
_s.i" 1.0°
healthy
IPi""-10° (""South)
50
:i> :
.. .··
,.·'
a:::
0
1-
u
<(
lL.
40
w
u
z
oo
<(
1-
u
30
.:;;::~:::::::::::::
w
....)
lL.
w
a:::
....)
<(
0
• •
~::
: : : :
~
: : : : : :
f. 50
~
1
: I '; : : : I ' 1 :: •' ' l
:
~
0
0
0
I
' ' o 1 • • I
, : •
t
.
•,;:·
diseased
20
a:::
1-
u
w
a..
(J)
10
600
SUGAR BEETS
700
aoo
ANZOLA
Fig. 9
099.
900
1000
VAVELENGTH I NM
---,----·
t-
z
LU
u
0::
LU
a..
3o L
I
0::
0
t-
u
<(
LL
LU
u
z
<(
t-
u
LU
_J
LL
LU
0::
_j
<(
0::
t-
u
a..
LU
Vl
SUGAR BEETS, HEALTHY
SAN PIETRO I SEPT. '!..
FRDiH L1 GHTI NG
::1
.3-1 = 55°
/~\
.3-r= ao
• .3- r = 15°
+ .3-r= 30°
- .3-r = 45°
'
J
o
~----""\
I /~
J
j/~
//1
~"'
/t
t
' , ..
I t//l
5r !l.·"
_;
1979
~--,'-- ~/
~~~~---'e~~~
/ · ~'i
~
I
1.500
1.625
1.750
1.875
2.000
2.12 5
2.250
2.375
WAVELENGTH
lfJ.m
SUGAR BEETS
FIG .
t-
z
10
60
..... . May 29, 1979 diseased
- -May 30, 1979
.3-r =t.5 °
~
. .. : ....
. . ..
LU
u
0::
LU
a..
'
50
0::
0
t-
u
<(
LL
40
LU
u
z<(
t-
u
30
LU
_j
LL
LU
rpr = -90° (East!
0::
_j
<(
20
0::
t-
.3-i = 28 °
rp i ~ oo (South l
u
LU
a..
(/)
10
600
Fig 11
VINTER BARLEY
700
800
LIMBURG ERH OF
100.
900
1000
VAV ELENG TH I NM
r-
z
w
60
u
0::
w
0....
.......
G) May
29, 1979 diseased
0
May
30, 1979
CD
©
June
26, 1979 diseased
CD
June 27, 1979
...
50
.........
LL
·'
0
r-
<t:
.· .
. .· ·
0::
0
u
....
... .. ·
..
40
w
u
z
<t:
ru
w
JO
_J
LL
w
0::
_J
20
<t:
0::
t-,-
Sr=Oo
u
w
0....
(J)
10
600
VINTER BARLEY
700
500
LIMBURGERHOF
900
1000
WAVELENGTH
I
NM
Fig 12
References
/1/ BOEHNEL, H.-J . , W. FISCHER, G. KNOLL : The dependence of the
spectral signature of sugar beets on the observation level
and the reflection geometry. ISP congress Hamburg 1980,
Commission VII/9.
/2/ PLESSEY RADAR : Multisp e ctral scanning systems and their
potential application to Earth Resources. Reports to ESRO
und Contract No. 1673/72 EL, Vol . 2, Dec . 1972.
/3/ SANWALD, E.: Private communication .
101..