RS-1D LISS- 
r backscatter 
mum Oo? and 
channels of 
; the same 
be extracted 
T SAR and 
followed to 
satellite data 
re and radar 
relationships 
icient at 16° 
' and optical 
OrS. 
n a simple 
nomograms 
at different 
uired at 16° 
proach was 
ivala (1976) 
vity of radar 
)bservations 
tion and C- 
ighness and 
1 developed 
jl moisture 
Iding radar 
nfluence of 
> observed 
ions were, 
depth and 
rresponding 
icient at 16? 
nparison of 
nated using 
nd 0-10cm 
2/g and 8.1 
ound to be 
rop or with 
difficult to 
rop cover 
e and radar 
guration of 
on studies 
backscatter 
ned above, 
have been 
gression by 
ly between 
(i) each of 
moisture at 
ih, initially 
tions at 16? 
indancy. It 
X data sets. 
ook angle 
significant 
ostly to the 
IAPRS & SIS, Vol.34, Part 7, “Resource and Environmental Monitoring", Hyderabad, India, 2002 
The correlation study showed that the Extended Low Beam 
SAR data acquired at 16° look angle has linear correlation with 
the soil moisture in 0-5cm depth for bare fields with R’=0.72, 
which was reduced to 0.69 for 0-10cm soil depth. This indicates 
that RADARSAT SAR responds mainly to the soil moisture in 
the 0-5cm surface layer only. It was observed that the 
sensitivity of 16° look angle RADARSAT SAR backscatter 
coefficient, though exhibited a good linear relationship, is only 
0.09db/g/g and 0.104db/g/g, respectively with the moisture 
content in the two depths (Figure 4). The observed sensitivity is 
very low compared to the sensitivity of 0.25db/0.01g/cm? soil 
moisture reported by Dobson and Ulaby (1981), Ulaby and 
Batlivala (1976) with field scatterometer observations at C- 
band. No significant relationship could be observed with the 
soil moisture in the crop covered fields using the 16? look angle 
SAR data. Similarly, no significant relationship between soil 
moisture and backscatter coefficient at 45? look angle could be 
seen either with the bare fields or crop cover fields. However, it 
could be noticed that the backscatter coefficient at the oblique 
look angle of 45? correlates (r= 0.82, N=38) well with the 
Normalised Difference Vegetation Index, a measure of crop 
vigour, which was derived using the red and near infra red 
channels of IRS-1D LISS-III data used in the study. Based on 
this observation, combined data of IRS-LISS-III and 
RADARSAT SAR in two look angles were used in multiple 
regressions for soil moisture estimation, which was discussed in 
the following section. 
2.3.3c. Synergistic Approach: Multiple linear regression has 
been carried for soil moisture estimation in the 0-5 and 0-10cm 
depths with the radar backscatter coefficients at 160, 450 and 
the Normalized Difference Vegetation Index of the cultivated 
and uncultivated fields. Two additional parameters, viz., 
difference of the backscatter coefficient values and their ratio 
were also used in the regression. The regression has been 
carried out using different combinations of these five 
observables. Best estimates of soil moisture in the two depths 
could be seen when radar backscatter at 16? and 45? and NDVI 
data were used. The standard errors of estimates for soil 
moisture in the 0-5cm and 0-10cm depths were 4.24 g/g and 
4.58 g/g respectively with such combination. The SEE were 
marginally higher with 4.7 g/g for 0-5cm depth soil moisture 
and 4.8g/g for 0-10cm depth for the combination of radar 
backscatter at 16° and NDVI. Figure 5 shows a comparison of 
retrieved soil moisture using two radar channel and NDVI data 
with the observed values at two depths. It indicates that 
RADARSAT EXL (16 Deg) 10-12-2001 
Use/Cove 
  
synergistic use of optical data and C-band SAR data at near 
nadir angle would yield a solution for soil moisture estimation. 
It needs, however, to be thoroughly verified with more ground 
observations under varying soi! surface roughness and crop 
cover conditions. 
4 CONCLUSIONS 
In the present study, an attempt has been made to estimate soil 
moisture using RADARSAT SAR data acquired at near and off 
nadir look angles and NDVI derived from IRS-1D LISS-III 
data. Near nadir look SAR data were observed to have a good 
relation with bare field soil moisture conditions. The study also 
indicates that synergistic use of optical data and C-band SAR 
data at near nadir angle would yield a solution for soil moisture 
estimation. 
ACKNOWLEDGEMENTS 
Authors are grateful to Dr. R.R.Navalgund, Director, NRSA for 
his encouragement and support of the work. Authors are 
thankful to Sri M. V. Krishna Rao, Head, CI&DA Division, 
NRSA, Hyderabad for providing the IRS-LISS-III data of 
December 2001 used in the study. 
REFERENCES 
Dobson, M.C., and F.T., Ulaby, 1981, Microwave backscatter 
dependence on surface roughness, soil moisture and soil 
texture: Part-III — Soil Tension, IEEE Trans. Geoscience 
Remote Sensing, Vol. GE-19, No.1, pp. 51-61 
Dubois, P.C., J. van Zyl, and T. Engman. 1995. Measuring soil 
moisture with imaging radars. IEEE Trans. Geosci. Remote 
Sensing 33: 915-926. 
Oh, Y., K. Sarabandi, and F.T. Ulaby. 1992. An empirical 
model and an inversion technique for radar scattering from bare 
soil surfaces. IEEE Trans. Geosci. Remote Sensing 30: 370- 
382. 
Ulaby, F.T., and P.P.Batlivala, 1976, Optimum radar 
parameters for mapping soil moisture, IEEE Transactions on 
Geoscience Electronics, Vol. GE-14, No.2, pp. 81-92 
IRS-ID LISS-II of 09-12-2001 
A 
Figure 2. Part of the study area as seen in RADARSAT SAR and IRS-1D LISS-III imagery with major classes marked 
713