8 Evaporation
171
8.4 Current trends: improved observations and improved
parameterizations.
8.4.1 Local maximum evaporation and land surface temperature [6]
When dealing with heterogeneous land surfaces two issues have to be dealt with
separately:
A. Surface temperature provides information on actual evaporation in a relative
sense only: given the energy available (Rn-G) and Planetary Boundary Layer
(PBL) conditions, including the aerodynamic properties of the surface, a
higher surface temperature indicates lower evaporation.
B. Accurate estimates of evaporation require accurate determination of land surface variables, particularly of the maximum evaporation and minimum surface
temperature attainable under the given conditions (available energy and PBL
conditions) .
In other words the maximum evaporation and the corresponding range from minimum to maximum surface temperature changes spatially for a given radiative and
PBL forcing. To use surface temperature as a measure of evaporation, i.e. of water
availability, both maximum evaporation and the range from minimum to maximum
surface temperature must be determined first.
Implicitly, both the linear relationship and the look-up table methods attempt to
address this issue, although evaporation mapping would require mapping of
(Rn-G), Band n. On the other hand the early results (e.g. Stone et al., 1975;
Jackson et aI., 1977) indicated that observations of surface temperature were
highly correlated with actual evaporation under precisely defmed experimental
conditions. Analysis of multi-spectral data sets, i.e. concurrent measurements of
surface temperature, albedo and spectral indices brought evidence of well identified clusters of data points related to the range of actual evaporation. Goward et ai.
(1985) provided evidence of the correlation of NDVI with T rad. Menenti et ai.
(1986) studied patterns of temperature and albedo in the oasis-playa-desert environment of Tunisia. Menenti et ai. (1989) exploited the correlation of surface albedo with surface temperature to map actual evaporation.
Observations of surface temperature give a measure of evaporation, relative to
maximum evaporation, at constant green vegetation. This concept, i.e. the vegetation index! temperature trapezoid (VITT; Fig. 8.3) was described in detail by
Moran et ai. (l994a). Carlson et ai. (1995) used a SVAT model to determine linear
isopleths (at constant soil moisture availability and relative evaporation) relating
NDVI to surface radiant temperature. Once the range of surface temperature, the
maximum evaporation and the isopleths have been determined, this method makes
straightforward use of observed surface temperatures possible.
171
8.4 Current trends: improved observations and improved
parameterizations.
8.4.1 Local maximum evaporation and land surface temperature [6]
When dealing with heterogeneous land surfaces two issues have to be dealt with
separately:
A. Surface temperature provides information on actual evaporation in a relative
sense only: given the energy available (Rn-G) and Planetary Boundary Layer
(PBL) conditions, including the aerodynamic properties of the surface, a
higher surface temperature indicates lower evaporation.
B. Accurate estimates of evaporation require accurate determination of land surface variables, particularly of the maximum evaporation and minimum surface
temperature attainable under the given conditions (available energy and PBL
conditions) .
In other words the maximum evaporation and the corresponding range from minimum to maximum surface temperature changes spatially for a given radiative and
PBL forcing. To use surface temperature as a measure of evaporation, i.e. of water
availability, both maximum evaporation and the range from minimum to maximum
surface temperature must be determined first.
Implicitly, both the linear relationship and the look-up table methods attempt to
address this issue, although evaporation mapping would require mapping of
(Rn-G), Band n. On the other hand the early results (e.g. Stone et al., 1975;
Jackson et aI., 1977) indicated that observations of surface temperature were
highly correlated with actual evaporation under precisely defmed experimental
conditions. Analysis of multi-spectral data sets, i.e. concurrent measurements of
surface temperature, albedo and spectral indices brought evidence of well identified clusters of data points related to the range of actual evaporation. Goward et ai.
(1985) provided evidence of the correlation of NDVI with T rad. Menenti et ai.
(1986) studied patterns of temperature and albedo in the oasis-playa-desert environment of Tunisia. Menenti et ai. (1989) exploited the correlation of surface albedo with surface temperature to map actual evaporation.
Observations of surface temperature give a measure of evaporation, relative to
maximum evaporation, at constant green vegetation. This concept, i.e. the vegetation index! temperature trapezoid (VITT; Fig. 8.3) was described in detail by
Moran et ai. (l994a). Carlson et ai. (1995) used a SVAT model to determine linear
isopleths (at constant soil moisture availability and relative evaporation) relating
NDVI to surface radiant temperature. Once the range of surface temperature, the
maximum evaporation and the isopleths have been determined, this method makes
straightforward use of observed surface temperatures possible.