90
Multiscale Hydrologic Remote Sensing: Perspectives and Applications
where LE, R n , and G have been previously defined; ρ is the water vapor density (in
kilograms per cubic meter); C p is the specific heat of air (in joules per kilogram); r a is
the aerodynamic resistance (in second per meter); and T c and T a are the canopy and air
temperatures respectively (in degrees Celsius). Estimation of LE requires a measure
of basic energy balance components, canopy temperature, and air temperature and an
estimate of the resistance term for sensible heat transfer. In this approach, there has to
be an estimate of the r a term that is often derived from fairly simple approximations of
canopy turbulence parameters (z o and d) and wind speed. Canopy temperature–based
models for ET estimation have been evaluated by comparing direct measurements of
LE from lysimeters and those estimates from Equation 5.2 for a number of locations
and crops (Hatfield et al. 1984). Standard error of the regression lines was 75 W m –2 ,
indicating agreement between the two ET methods.
Regional estimates of evaporation are possible with remotely sensed data, and
most of the approaches are considered to be single-source ET models. Zhang et al.
(1995) applied Equation 5.2 to regional ET estimates in France and found that the dif5.2 to regional ET estimates in France and found that the dif.2 to regional ET estimates in France and found that the difference between a remote sensing model and area ET averages obtained from groundbased stations was within 28 W m –2 . The application of a single-source model begins
to encounter problems when there is a partial canopy-covered surface, which often
exists in agricultural systems (Kustas et al. 1989; Hall et al. 1992; Vining and Blad
1992). Problems with single-source models can be attributed to differences between
the “aerodynamic” and “radiometric surface temperature” (Norman and Becker
1995). Other “two-source” approaches that consider energy exchanges from both soil
and vegetation components account for the differences between radiometric temperature and aerodynamic temperature and thus represent an advance over single- and
two-source models (e.g., Norman et al. 1995; Kustas and Norman 1999). The use of
canopy or surface temperatures as direct inputs into large-scale ET models provides
a spatial representation of water use that is not possible with single energy balance
systems. In the application of canopy temperatures, the problem of incomplete ground
cover is critical because of the potential differences in temperature between the soil
and the crop. The development of multiscale approaches combining thermal, visible,
and near-infrared imagery from multiple satellites to partition the fluxes between the
soil and canopy offers the potential for future improvements in the use of surface
temperature at a range of scales from 1 m to 10 km (Anderson et al. 2007). This type
of method shows the further refinement in the ability to use remote sensing as an
assessment tool for ground-based observations as well as a method for regional-scale
measurements. Application of remote sensing of thermal radiation offers a potential
for new advances in our understanding of the complexities of the surface–atmosphere
exchanges. These data are available in the study we conducted; however, the focus
was directed toward quantifying the advection and turbulence dynamics of the surface before comparing methods applicable to regional-scale assessments.
5.2.2 eddy covaRiance
Turbulent fluxes of sensible and latent heat can be measured directly with fast- response
sonic anemometers and infrared gas analyzers (IRGAs). Sonic anemometers measure
the wind velocities in three-dimensional (3D) space (x, y, and z), and in meteorological
Multiscale Hydrologic Remote Sensing: Perspectives and Applications
where LE, R n , and G have been previously defined; ρ is the water vapor density (in
kilograms per cubic meter); C p is the specific heat of air (in joules per kilogram); r a is
the aerodynamic resistance (in second per meter); and T c and T a are the canopy and air
temperatures respectively (in degrees Celsius). Estimation of LE requires a measure
of basic energy balance components, canopy temperature, and air temperature and an
estimate of the resistance term for sensible heat transfer. In this approach, there has to
be an estimate of the r a term that is often derived from fairly simple approximations of
canopy turbulence parameters (z o and d) and wind speed. Canopy temperature–based
models for ET estimation have been evaluated by comparing direct measurements of
LE from lysimeters and those estimates from Equation 5.2 for a number of locations
and crops (Hatfield et al. 1984). Standard error of the regression lines was 75 W m –2 ,
indicating agreement between the two ET methods.
Regional estimates of evaporation are possible with remotely sensed data, and
most of the approaches are considered to be single-source ET models. Zhang et al.
(1995) applied Equation 5.2 to regional ET estimates in France and found that the dif5.2 to regional ET estimates in France and found that the dif.2 to regional ET estimates in France and found that the difference between a remote sensing model and area ET averages obtained from groundbased stations was within 28 W m –2 . The application of a single-source model begins
to encounter problems when there is a partial canopy-covered surface, which often
exists in agricultural systems (Kustas et al. 1989; Hall et al. 1992; Vining and Blad
1992). Problems with single-source models can be attributed to differences between
the “aerodynamic” and “radiometric surface temperature” (Norman and Becker
1995). Other “two-source” approaches that consider energy exchanges from both soil
and vegetation components account for the differences between radiometric temperature and aerodynamic temperature and thus represent an advance over single- and
two-source models (e.g., Norman et al. 1995; Kustas and Norman 1999). The use of
canopy or surface temperatures as direct inputs into large-scale ET models provides
a spatial representation of water use that is not possible with single energy balance
systems. In the application of canopy temperatures, the problem of incomplete ground
cover is critical because of the potential differences in temperature between the soil
and the crop. The development of multiscale approaches combining thermal, visible,
and near-infrared imagery from multiple satellites to partition the fluxes between the
soil and canopy offers the potential for future improvements in the use of surface
temperature at a range of scales from 1 m to 10 km (Anderson et al. 2007). This type
of method shows the further refinement in the ability to use remote sensing as an
assessment tool for ground-based observations as well as a method for regional-scale
measurements. Application of remote sensing of thermal radiation offers a potential
for new advances in our understanding of the complexities of the surface–atmosphere
exchanges. These data are available in the study we conducted; however, the focus
was directed toward quantifying the advection and turbulence dynamics of the surface before comparing methods applicable to regional-scale assessments.
5.2.2 eddy covaRiance
Turbulent fluxes of sensible and latent heat can be measured directly with fast- response
sonic anemometers and infrared gas analyzers (IRGAs). Sonic anemometers measure
the wind velocities in three-dimensional (3D) space (x, y, and z), and in meteorological