Geoscience Reference
In-Depth Information
USA, in conjunction with hydrological modelling and ground surveys, for water resources assessment,
flood forecasting and dam regulation. The utility of the technique remains subject to the limitations of the
spatial and temporal resolution of the satellites used, the requirement for generally cloud-free conditions
in the image, and the fact that only cover can be easily assessed, rather than snow water equivalent,
although attempts have been made to use passive microwave remote sensing to estimate the spatial
patterns of water equivalent of snow packs. It is, however, the only method for getting the widespread
large-scale coverage that is required for major river basins. Ground-based photographic measurements
have also been used in snowmelt modelling (e.g. Bloschl
et al.
, 1991).
There are some other remote-sensing techniques and interpretation models that may become more
useful in the future. Work in this area has become increasingly important with the development of work
in “macroscale” hydrology, which has been encouraged by the needs of global atmospheric circulation
modelling for hydrological predictions at large scale. Active and passive microwave techniques for the
measurement of soil moisture have been studied for some time. Active sensors transmit a signal to the
ground and measure the return signal; passive systems measure only the natural microwave transmission
from the surface. The first satellites with active microwave sensors that can be used for soil moisture
estimation (ERS1, ERS2, JERS1 and Radarsat) have been orbiting since the early 1990s. They provide
good spatial resolution of 30-100 m but with repeat coverage periods of the order of 30 days. Active
microwave systems have also been used from ground- and aircraft-based platforms. The wavelengths
normally used penetrate, at most, only the first few centimetres of the soil surface. The return signal then
depends on the dialectric constant of the surface soil layer. The dialectric constant varies with moisture
content, most strongly for soils that are neither too wet or too dry. Sensitivity of the measurement is
much reduced at both high and low moisture contents and it is difficult to differentiate, for example,
between a wet and saturated soil surface, particularly when the vegetation cover may also be wet. Passive
microwave satellite sensors (SSM/I, AMSR and SMOS) provide much higher frequency measurements
(1-3) days but much lower spatial resolution (5-50 km) than would be considered useful in rainfall-
runoff modelling.
The problem is that, for both active and passive microwave sensors, the signal also depends on the
water content of the vegetation, the roughness of the surface and the state of the atmosphere. In most
images from active radar systems, for example, the most obvious features are those associated with
the topography and roughness of the surface, such as different vegetation covers. Extracting the soil
moisture content signal means extracting a second-order effect and often relies on the availability of
ground measurements that can be used to calibrate the interpretation of the radar image. The approach
will work best, therefore, where there is a uniform surface, particularly if there is little vegetation cover.
There have, however, been some very interesting results presented from airborne and satellite microwave
sensors, using both active and passive microwave systems. The techniques may well be improved in the
future. The wavelength of the signal remains a limitation, in that only surface soil moisture can currently
be detected in this way. There is, then, a problem in relating that surface soil moisture content to the
moisture profile and flow processes or to the results of a hydrological model. One of the few studies that
has attempted such a comparison at a large scale is that of Wood
et al.
(1993).
There is a remote-sensing technique for estimating the changes of water storage in the landscape over
time. Changes in storage will change the gravity anomaly and data from the GRACE satellite, launched
in 2002, are processed to produce monthly estimates of patterns of water storage. The spatial scale of
the estimates is, however, 60 km so that the information will be most useful in constraining macroscale
models of large catchments or global hydrology (e.g. Guntner, 2008) or changes in large scale groundwater
bodies (e.g. Rodell
et al.
, 2007).
Satellite information might also be useful in the estimation of river stage and discharges. At present,
the various satellite altimetry methods do not have a sufficiently high spatial or time resolution to be
useful except on the very largest rivers (Birkett
et al.
2002; Trigg
et al.
, 2009). This might change in the
future, however, when the planned SWOT satellite is launched some time in the next decade. This will
