OIL
refers to the
varies with
water in the
in that soil
en made to
JAR
d specially
rticles in a
portant in
Since the
>s With the
ven soil by
cific size
nsiderable
, the total
is largely
Ir and air
e divided
' refers to
lar layers
ghtly held
d osmotic
rgy of the
molecules
molecules
from soil
h relative
> Water is
n the soil
nd bound
nines the
fact that
incident
se of ice.
n that is
(1986a)
avior of
e in soil
constant)
tant) are
ne value
IAPRS & SIS, Vol.34, Part 7, “Resource and Environmental Monitoring”, Hyäerabad, India, 2002
3. BASIS OF THE METHODOLOGY
The dielectric constant of mixture of soil and water that can
vary from 3 (for very dry soil) to 35 (for very wet soils)
depends upon the total surface area of the soil particles present
in a unit volume. The soil texture gives information of the
surface area of soil particles in a unit volume, which in turn
indicates the amount of
Figure
1.
dx
c
(Field capacity)
Soil moisture at 1/3 bar
C
ta
ta
Ba
ry
e
Soil moisture at 15 bar
(Wilting point)
oo
Unavailable water
Volumetric Soil Moisture (94)
©
Sand Sandy Loam Sit Clay Clay
loam loam loam
Fineness of texture >
Variation of available water with soil texture (modified drawing
from Brady, 1995)
free water molecules in the soil. This calls for defining a
parameter that represent that part of soil moisture which relate
to the physical interaction process that takes place between
microwave signal and the soil water mixture. Thus one can not
overlook the effect of soil texture, which governs the surface
area of soil particles and thus the amount of free water in a
given soil water mixture while developing the models to
retrieve soil moisture information.
Since the transition point between bound water and free water
is based on an arbitrary criterion, it is thought to represent soil
moisture in terms of percentage of available water for different
soil textures as shown in Figure 1. In this paper an attempt has
been made to include the effect of soil texture in the soil
moisture retrieval model by representing the soil moisture in a
more realistic term that would relate to the possible amount of
free water in a given soil water mixture. The difference
between the amount of soil moisture at 15 bar and 1/3 bar
pressure represent the water that is available to the plant
(Brady, 1995). Thus soil moisture in terms of percentage of
available water appears to be an appropriate representative of
the amount of free water molecule in a given soil water
mixture. Equation 1 gives the soil moisture represented in terms
of percentage of available water.
SM % 4.w. — (SM oss * 100) / (SM 1/3bar— SM 15bar) (1)
where SM Aw. — Soil moisture in percentage of available
water
SM ops = Observed soil moisture from field
SM jp, = Soil moisture at field capacity
SM 15bar = Soil moisture at wilting point
721
(SM i54, — SM 15bar) = available water
4. DATA SET AND STUDY AREA
In order to validate the above concept, Extended Low-1 beam
mode RADARSAT-1 SAR data has been acquired on 13-
March-1999 over parts of Agra, Mathura (U.P.) and Bharatpur
(Rajasthan) districts. Large coverage of RADARSAT-1 EL1
image (170 X 170 Km?) provided data over a variety of soil
texture write from very fine to vary course. The RADARSAT-1
has a C-band SAR sensor with HH polarization. It operates
under a variety of viewing mode with varying spatial and
radiometric resolutions (RSI, 2000). Details of RADARSAT-1
EL1 beam mode digital data is given in Table 1.
Beam mode & position Extended Low -1
Date of Pass 13-March-1999
Incidence angle 10° - 23°
Nominal resolution 35 meters
Nominal coverage 170 X 170 Km?
Number of looks 1X4
Look direction Ascending
Orbit Number 17507
Product type Path image
Data format CEOS
Number of lines 12837
Number of pixels 13556
Number of bits per pixel 16 bits unsigned
Table 1. Details of RADARSAT-1 ELI digital data
In addition to SAR data, two SAT (Shift along track) scenes of
IRS L-III with P/R 96/52,53 (dated 08-March-99) and P/R
97/52,53 (dated 05-March-99) were also used. Spectral bands
for IRS L-III are Green (0.52-0.59u), Red (0.62-0.68u), Infra
red (0.77-0.86u) and Short wave Infrared (1.55-1.70p). Besides
the optical and microwave digital data, 1:50,000 scale SOI
topographic maps were also used for ground truth planning,
distric/block boundaries and identification of ground control
points on to the image.
The study area is mainly a flat level terrain dominated by
agricultural land, covering irrigated as well as un-irrigated land.
Major crops during Rabi season were Mustard and wheat.
During data acquisition in middle of the March 1999, most of
the mustard fields were harvested and rest fields were under
ripened stage whereas wheat was under grain filling stage.
Other land covers categories like habitation, water bodies and
forest plantation were also present in this study area.
5. DATA ANALYSIS / METHODOLOGY
Following are the major steps involved in the data analysis /
methodology:
5.1 Reconnaissance survey and sampling location selection
Selection of sampling locations for ground truth data collection
including in-situ soil moisture measurements. Sampling
locations were selected carefully after reconnaissance field
survey. Sampling locations were selected near the ground
control points. Enough care was taken regarding homogeneity
