STORAGE - Fundamentals of Hydrology - page 70

on two factors: the soil wetness and the surface
roughness. Where soil roughness is well known and
there is little or no vegetation cover the radar
backscatter has been well correlated with surface
moisture (Griffiths and Wooding, 1996; Kelly et al .,
2003). This technique offers hope for future
estimation of soil moisture from satellites, but there
is still much research to be done in understanding
the role vegetation plays in affecting radar back-
scatter.
The great advantage of any remote sensing
technique is that it automatically samples over a
wide spatial area. The satellite measures the electro-
magnetic radiation within each pixel; this is an
average value for the whole area, rather than a point
measurement that might be expected from normal
soil moisture measurements. The question that
needs to be answered before satellite remote sens-
ing is widely accepted in hydrology is whether
the enhanced spatial distribution of measurement
is sufficient to overcome the undoubted limitations
of direct soil moisture measurement.

Passive microwave

The earth surface emits microwaves at a very low
level that can be detected by satellites. These are
referred to as passive microwaves, in the sense that
the earth emits them regardless of whether the
satellite is present or not. The amount of water
present on or above the surface of the earth affects
the passive microwave signal emitted (it is a lower
signal the more water there is present). The SSM/I
satellite platform is able to measure passive micro-
waves, but only at a very coarse spatial resolution
( ≈ 100 100km). This is the major drawback of this
technique at present.

Active microwave

Active microwaves are emitted from a satellite and
the strength of returning signal measured. This is a
complex radar system and has only been available
on satellites since the early 1990s. The strength
of microwave backscatter is primarily dependent

Case study

REMOTE SENSING OF SOIL MOISTURE AS A REPLACEMENT FOR
FIELDWORK

Remote sensing of soil moisture may offer a way
of deriving important hydrological information
without intensive, and costly, fieldwork pro-
grammes. Grayson et al . (1992) suggest that this
could be used to set the initial conditions for
hydrological modelling, normally a huge logistical
task. One of the major difficulties in this is that
the accuracy of information derived from satellite
remote sensing is not high enough for use in
hydrological modelling. As a counter to this it
can be argued that the spatial discretisation offered
by remote sensing measurements is far better
than that available through traditional field
measurement techniques.

In an attempt to reconcile these differences
Davie et al . (2001) intensively monitored a
15-hectare field in eastern England and then
analysed the satellite-derived, active microwave
backscatter for the same period. The field pro-
gramme consisted of measuring surface soil
moisture (gravimetric method) at three different
scales in an attempt to spatially characterise the
soil moisture. The three scales consisted of: (a)
thirteen samples taken 1 m apart; (b) lines of
samples 30 m apart; and (c) the lines were
approximately 100 m apart. The satellite data was
from the European Remote Sensing Satellite (ERS)
using Synthetic Aperture Radar (SAR).