8.2 1996 experiment 
8.2.1 Crop cover estimation by LAI using ERS SAR for two 
beet fields 
Ground truth data were collected at weekly intervals and not on 
the satellite overpass dates. LAI and soil moisture were 
calculated for satellite overpass dates by linear interpolation 
between the measurement dates and used for radar backscatter 
prediction and LAI comparison (Table 1). Correlation 
coefficient of 0.92 and 0.96 were found for fields X and Y, 
respectively between measured and predicted LAI (Figure 1). 
Table 1: LAI and soil moisture used in the water cloud model 
for 1996 
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
Field X Field Y 
Date LAI |m,(9?v))| LAI m, (96) 
07.6.96 0.16 10.78 0.31 10.75 
13.6.96 0.34 9.40 0.54 9.81 
16.6.96 0.52 6.80 0.77 2.21 
22.6.96 0.73 7.45 1.07 6.36 
23.6.96 0.77 7.45 1.12 6.36 
11.7.96 1.33 13.80 2.00 12-13 
18.7.96 1.55 7.11 2.59 6.00 
21.7.96 1.54 5.68 3.06 4.96 
27.7.96 1.73 5.65 3.19 4.99 
16.8.96 1.55 25.65 2.17 ‘21.59 
N es (AS 
^ 
= 
93 : O Field X 
u A Field Y 
B pH A —— 1:1 Line 
38 ZA 
3 
© iB 
317 
a OA A 
" A 
0 + i + i 
0 1 2 3 4 
| Measured LAI 
  
  
Figure 1: Comparison between measured LAI and LAI 
predicted by ERS SAR for two sugar beet fields: 1996 
Experiment 
Figure 2 shows predicted crop cover (96) using ERS-1 and 
ERS-2 SAR data for the two sugar beet fields on ten dates 
from June to August. Correlation coefficients of 0.95 and 0.92 
were achieved for fields X and Y, respectively. 
High correlation between measured crop cover and cover 
values predicted by ERS SAR in this study can be related to 
Dockter and Kuhbauch's (1990) observation that radar 
backscatter from beet fields increases up to July and afterwards 
becomes stable in a similar manner to the f profile. Ground 
truth - 
data was collected in this study upto canopy closure (July end). 
IAPRS & SIS, Vol.34, Part 7, “Resource and Environmental Monitoring”, Hyderabad, India,2002 
  
  
  
  
  
| 
100 re, 
€ 807 oc /A 
PTE LRY 
= B A O FieldX | 
o ; 
e A Field Y 
ul 60:7 d — 1:1 Line 
a 
$ 40 a 
9 + 
8 Oa A 
8 A 
o 
5 ni H A 
0 + + + + 
0 20 40 60 80 100 
Crop cover by measured LAI 
  
  
  
Figure 2: Comparison between crop cover (%) from measured 
LAI and LAI by ERS SAR for two sugar beet fields: 1996 
Experiment 
8.2.2 Extension of radar ground-data comparison to nine 
sugar beet fields 
Nine sugar beet fields near Brooms Barn were identified on the 
1996 imagery with the help of a field survey. Backscatter 
values were calculated for each of the fields and these were 
used to estimate crop canopy cover by equation 3. August 12- 
13 experiment data were used as reference for august date 
comparison with measured. Correlation coefficient 0.71 was 
found between crop cover estimates from ERS SAR and 
measured LAI in August. Weather data can be used to adjust 
and calculate m, values as advised by previous studies (Xu et 
al, 1995, 1996). It shows that 10% error in m, makes a 
difference of less than 0.5 in derived LAI at smaller LAI. 
8.3 Sugar yield estimation 
8.3.1 Sugar yield prediction for 1995 and 1996 experimental 
fields 
Sugar yield (t/ha) was predicted by using equation 4 for two 
sugar beet fields in 1995 and 1996. Crop cover were used in 
above equation from ERS SAR, measured LAI and SRM 
(Table 2). In 1996 crop cover was also estimated from vertical 
and off-nadir photographs. 
Table 2: Two sugar beet fields yield prediction for 1995 and 
1996 using Werker and Jaggard (1997) model 
  
  
  
  
  
  
1995|Observed|ERS |SPOT |LAI |SRM 
Yield 
(t/ha) 
A 6.320777 89T 
B 6.8| 11.8 1131 121 11 
1996 Observed |ERS |SPOT |LAI |SRM |Vertical [Angle 
Yield photo  |photo 
(t/ha) 
X 8.64| 9.2 11.4] 8| 9.3 9.8| 10.4 
Y 10.24| 10.8 12.3| 11.2| 10.6 11.1] :-11.2 
  
  
  
  
  
  
  
  
  
  
Although there was a large bias in the yield prediction, there 
was a correlation coefficient of 0.85 between observed sugar 
yield (t/ha) and yield predicted by SPOT data (Figure 3).