1 
i 
4 
Front view 
measurements 
d signal is 
/er the instan- 
ie two signals 
sat frequency 
be shown (Krul 
the FM-CW 
taking advan- 
pearing when 
/e velocity, 
lar with a con- 
i frequency 
1 and the re- 
lce is called 
spier scattero- 
requires spe- 
Lon cells, 
ss used for 
sargets, we 
1 for the re- 
sroduction it 
spment of a 
system was 
>le and was 
sasic idea 
irriage was 
plots (fig. 8) . 
signal was re- 
i recording is 
the different 
snt crop types. 
rferent inci- 
s were chosen, 
sties of which 
.th slight modi- 
and including 
is to be expec- 
Table 1. Specifications of X-band ground-based 
scatterometer 
Frequency 
Frequency sweep 
Modulation waveform 
Modulation frequency 
Output power 
polarization 
Incidence angle 
Range 
Antenna 
Recording 
10 GHz 
400 MHz (till 1979) 
750 MHz (since 1980) 
triangle 
variable, used at 100 Hz 
10 mW 
W, HH, VH (HV) 
10 - 80° 
5 - 15m 
pencil beam, 4.4 
digital 
Table 2. Specifications of Q-band ground- 
based scatterometer 
Frequency 
34,4 GHz 
Frequency sweep 
800 MHz 
Modulation waveform 
triangle 
Modulation frequency 
variable, used at 100 Hz 
Output power 
10 mW 
Polarization 
VV, HH, VH (HV) 
Incidence angle 
10 - 80° 
Range 
5 - 15m 
Antenna 
pencil beam, 5.6 
Recording 
digital 
Table 3. Specifications of 6-frequency 
airborne scatterometer 
Frequencies 
1.2, 3.2, 5.3, 9.65, 13.7 
and 17.25 GHz 
Modulation type 
coherent pulse-modulation 
Pulse rep. freq. 
78.125 kHz 
Pulse - "width 
100 ns 
Output power 
250 mW, peak value 
Polarization 
W or HH 
Incidence angle 
10 - 80° 
Range 
50 - 1920m 
Antenna 
pencil beam, 0.9m para 
bolic dish 
Data acquisition 
I & Q, 8 bits, 20 M samples/s 
ted that the radar return for a given crop type not 
only depends on the polarization and incidence angle 
of the electromagnetic wave but also on its frequen 
cy it seemed interesting to introduce a second fre 
quency. For a number of reasons this frequency was 
chosen in the 8 mm band. 
The specifications of the instrument are given in 
table 2. Only during the growing seasons of 1980 and 
1981 this 8mm scatterometer was used in the systema 
tic. measurement program. 
Although the groundbased scatterometry has proven 
to be very useful in studying the scattering of 
microwaves by vegetation it has some limitations. In 
the first place the possible number of testfields'is 
very limited, therefore the statistical spread of 
the a°-values for different fields with the same crop 
type (ecological noise) cannot be investigated. Sec 
ondly the illuminated area is relatively small for 
groundbased measurements and in some cases even too 
small, which sometimes leads to differences in back- 
scatter values obtained by groundbased and airborne 
systems. In the third place it turns out to be 
difficult to build good instruments for low freq 
uencies mainly in consequence of antenna design con 
straints. Finally groundbased scatterometry is, al 
most by definition, not appropriate for forestry 
Figure 10. Backscattering coefficient Y =cfO /cos0 
for various bare soils. Ploughed fields 1 and 2 have 
rms roughness values of 5.25 and 5.13 cm. Seedbeds 
3 and 4 have rms roughness values of 0.763 and 1.17 
cm. 
measurements. The mentioned limitations can be over 
come by using airborne instruments like DUTSCAT. 
This Delft University of Technology Scatterometer 
is a multiband instrument based on the principle of 
pulse modulation. The system is installed in the 
Beechcraft Queen Air research airplane of the Natio 
nal Aerospace Laboratory (NLR). The system specifi 
cations are presented in table 3. Incidence angle 
and polarization are chosen (Attema e.a. 1984) by 
the operator inside the aircraft. He also can select 
one up to six frequencies to operate virtually at 
the same time. In order to facilitate the interpre 
tation of the measurements a video recording of the 
observed track is made. It is to be expected that 
the system will become operational during the 1986 
season. 
4 SOME IMPORTANT RESULTS 
In this section a number of results, selected from 
the large amount of available ones, will be discussed 
especially with a view on practical applications. 
During the various campaigns the a°-values of a large 
number of crop types, including their varieties were 
measured as a function of polarization state and in 
cidence angle. One exception was made however in 1978 
for which year a bare soil program was arranged with*- 
in the framework of Earsel. 
Since bare soil is a relatively simple remote sen 
sing object first a few results of the 1978 campaign 
will be discussed. In fig. 10 the back scattering 
coefficient Y=0°/cos0 was chosen for the vertical 
axis whereas the horizontal axis presents the grazing 
angle 90 -9. The different fields are characterised by 
different roughness values, plots 1 and 2 are ploug 
hed fields with rms roughness values of 5.25 and 5.13 
cm respectively whereas plots 3 and 4are seedbeds 
with rms roughness values of 0.763 and 1,17 cm. 
1037