13. Measuring Water Availability and Uptake in Ecosystem Studies
207
two broad categories, passive and active techniques. Passive techniques measure the natural microwave radiation emitted by the ground surface
(plants and the soil), while active techniques generate a pulse of microwave radiation and measure
the backscatter to the sensor. The wavelengths most
commonly used to minimize interference from the
atmosphere are those greater than 5 cm (or a frequency of <6 GHz, where v = ciA and v is frequency, A is wavelength, and c is the speed oflight).
Data are generally taken with truck-mounted equipment or from airplanes, but satellite measurements
are becoming increasingly common (e.g., Owe et
al. 1988).
Soil brightness temperature (T B ) as measured by
a radiometer is related to soil temperature (Tsoi1) by
the following relationship for a bare surface and
isothermal soil (the Rayleigh-Jeans approximation
to Planck's law):
(13.9)
where r is reflectivity, c is emmisivity, and
r = 1 - c. The emissivity is dependent on surface
roughness and the dielectric constant of the soil
(Engman and Chauhan 1995). The ability to assess
soil moisture remotely is therefore based on the
same dielectric effect used by time domain
reflectometers-an increase in the soil dielectric
constant K from <5 in dry soils to >25 in wet
soils. Emissivity varies as a nonlinear function of
K from approximately 0.95 in dry soils to 0.6 in
wet soils, for a decrease in soil brightness temperature of more than 100
0
K (Schmugge 1990).
Such large changes in TB are easily measurable
with microwave radiometers.
Several factors must be taken into account for
interpreting remotely sensed soil moisture values.
Surface roughness tends to increase emissivity and
decrease reflectivity because of the increased surface area. The following equation was developed
by Choudury et al. (1979) to incorporate roughness
into the estimate of emissivity:
(13.10)
where h is an empirical roughness parameter and
Ro is the smooth surface reflectivity «0.1 for dry
soils and up to 0.4 for wet). A rough surface (e.g.,
root mean square height variations of 5 cm) can
result in a TB that differs by 50
0
K compared with
the same soil with a smooth surface (Choudury et
al. 1979). A second complication arises from vegetation semitransparent to microwave radiation
(Jackson and O'Neill 1987). With sufficient leaf
area, the brightness temperature reflects the vegetation rather than the soil temperature. Corrections
based on the water content of the vegetation are
possible but complicated (Schmugge 1990). Additional factors for consideration are soil texture and
the angle of measurement (if different from
perpendicular) .
The use of remotely sensed data is certain to increase. Overall, remote sensing of the moisture in
the top 5 cm of the soil has great potential in deserts, crop systems, and some grasslands, and a
number of field campaigns have shown good correlations between field and remotely sensed data
across broad spatial scales (e.g., Lin et al. 1994).
However, there are three primary disadvantages for
microwave approaches in ecosystem research. The
first is that current technologies are difficult to use
in systems with high leaf area indices, such as forests, because the vegetation attenuates the microwave signal from the soil (Engman and Chauhan
1995). A dense thatch or dead vegetation layer
causes similar problems (Schmugge et al. 1988).
The second disadvantage is that microwave approaches are unlikely to provide information on soil
layers deeper than 10 cm in the near future. The
final disadvantage is that the approach is expensive,
requiring airplane flights, truck-or tower-mounted
equipment, or satellite imagery. Despite these problems, no existing technique has such potential for
landscape and regional integration of soil moisture
availability. Combining its use with ground-based
methods and modeling provides a powerful approach for the coming century.
Estimating the Vegetative
Component of Ecosystem
Water Fluxes
The above methods are used to determine soil water
potential, water content, and the availability of water to plants. Measuring the amount of water transferred to the atmosphere by plants is also important
for many ecosystem studies (e.g., Jackson et al.
1998). Methods for determining total canopy or
ecosystem water fluxes are described in Chapters
207
two broad categories, passive and active techniques. Passive techniques measure the natural microwave radiation emitted by the ground surface
(plants and the soil), while active techniques generate a pulse of microwave radiation and measure
the backscatter to the sensor. The wavelengths most
commonly used to minimize interference from the
atmosphere are those greater than 5 cm (or a frequency of <6 GHz, where v = ciA and v is frequency, A is wavelength, and c is the speed oflight).
Data are generally taken with truck-mounted equipment or from airplanes, but satellite measurements
are becoming increasingly common (e.g., Owe et
al. 1988).
Soil brightness temperature (T B ) as measured by
a radiometer is related to soil temperature (Tsoi1) by
the following relationship for a bare surface and
isothermal soil (the Rayleigh-Jeans approximation
to Planck's law):
(13.9)
where r is reflectivity, c is emmisivity, and
r = 1 - c. The emissivity is dependent on surface
roughness and the dielectric constant of the soil
(Engman and Chauhan 1995). The ability to assess
soil moisture remotely is therefore based on the
same dielectric effect used by time domain
reflectometers-an increase in the soil dielectric
constant K from <5 in dry soils to >25 in wet
soils. Emissivity varies as a nonlinear function of
K from approximately 0.95 in dry soils to 0.6 in
wet soils, for a decrease in soil brightness temperature of more than 100
0
K (Schmugge 1990).
Such large changes in TB are easily measurable
with microwave radiometers.
Several factors must be taken into account for
interpreting remotely sensed soil moisture values.
Surface roughness tends to increase emissivity and
decrease reflectivity because of the increased surface area. The following equation was developed
by Choudury et al. (1979) to incorporate roughness
into the estimate of emissivity:
(13.10)
where h is an empirical roughness parameter and
Ro is the smooth surface reflectivity «0.1 for dry
soils and up to 0.4 for wet). A rough surface (e.g.,
root mean square height variations of 5 cm) can
result in a TB that differs by 50
0
K compared with
the same soil with a smooth surface (Choudury et
al. 1979). A second complication arises from vegetation semitransparent to microwave radiation
(Jackson and O'Neill 1987). With sufficient leaf
area, the brightness temperature reflects the vegetation rather than the soil temperature. Corrections
based on the water content of the vegetation are
possible but complicated (Schmugge 1990). Additional factors for consideration are soil texture and
the angle of measurement (if different from
perpendicular) .
The use of remotely sensed data is certain to increase. Overall, remote sensing of the moisture in
the top 5 cm of the soil has great potential in deserts, crop systems, and some grasslands, and a
number of field campaigns have shown good correlations between field and remotely sensed data
across broad spatial scales (e.g., Lin et al. 1994).
However, there are three primary disadvantages for
microwave approaches in ecosystem research. The
first is that current technologies are difficult to use
in systems with high leaf area indices, such as forests, because the vegetation attenuates the microwave signal from the soil (Engman and Chauhan
1995). A dense thatch or dead vegetation layer
causes similar problems (Schmugge et al. 1988).
The second disadvantage is that microwave approaches are unlikely to provide information on soil
layers deeper than 10 cm in the near future. The
final disadvantage is that the approach is expensive,
requiring airplane flights, truck-or tower-mounted
equipment, or satellite imagery. Despite these problems, no existing technique has such potential for
landscape and regional integration of soil moisture
availability. Combining its use with ground-based
methods and modeling provides a powerful approach for the coming century.
Estimating the Vegetative
Component of Ecosystem
Water Fluxes
The above methods are used to determine soil water
potential, water content, and the availability of water to plants. Measuring the amount of water transferred to the atmosphere by plants is also important
for many ecosystem studies (e.g., Jackson et al.
1998). Methods for determining total canopy or
ecosystem water fluxes are described in Chapters