176
M. Menenti
estimates of evaporation with measurements to fmd differences significantly smaller
than 100 Wm· 2 •
Two land surface variables appear in many parameterizations of evaporation and
turbulent heat fluxes:
the Leaf Area Index (LAI);
the aerodynamic resistances for turbulent transport of momentum and of sensible
heat fluxes
Accurate determination of these variables made the difference between good and
poor agreement with observations in many studies of latent and sensible heat
fluxes.
Leaf Area Index. Wollenweber (1995) used a one-dimensional model to study the
impact of spatial variability of land surface variables on heat fluxes. The LAI was
used to parameterize roughness length, zero plan displacement and stomatal resistance. Since these parameterizations are not linear, spatial aggregation of either
LAI or of the calculated heat fluxes will give different results. Kaneko and Hino
(1996) proposed a method to estimate evaporation based on concurrent estimates
ofLAI (inferred from NDVI), of vapour pressure deficit at leaf temperature (which
requires an estimate of the latter) and of stomatal resistance. On the other hand
Nemani et al. (1993) developed a procedure to estimate LAI using a corrected
NDVI by including spectral reflectance in the mid-infrared. They concluded that
relation of NDVI with LAI was poor in their forest area and that the proposed
method significantly improved the LAI estimates. The impact of the correction
was significant on estimates of evaporation.
Aerodynamic resistances. The most widely used parameterizations of momentum
and heat fluxes are derived from flux-profile relationships and relate mean (along
the vertical direction) flux to the difference in either wind speed or temperature
between the surface and a reference height. The accuracy of such parameterizations depends on the accuracy of the aerodynamic resistance (ra, the ratio of the
flux density to the difference in the state variable).
Kalma (1989) evaluated several parameterizations to conclude that the estimation of ra is difficult. Under conditions of large spatial variability of bare and
vegetation patches, accurate values of the resistance for momentum require accurate values of the aerodynamic roughness length (Humes et aI., 1994). The use of
thermal infrared radiometers to estimate surface temperature and sensible heat flux
density brings an additional difficulty. The proper reference for the flux-profile
relationship is an aerodynamic surface temperature, Taer, while Trad relates to directional measurements of radiance. The difference between T aer, and T rad is significant and depends on environmental conditions (Choudhury et aI., 1986).
To determine the resistance for heat fluxes the difference in temperature between
foliage and bare soil must, therefore, be taken into account. Norman et al. (1995)
described an explicit dual source heat transfer model. A different approach to estimate sensible heat flux using observations ofTrad was proposed by Lhomme et al.
M. Menenti
estimates of evaporation with measurements to fmd differences significantly smaller
than 100 Wm· 2 •
Two land surface variables appear in many parameterizations of evaporation and
turbulent heat fluxes:
the Leaf Area Index (LAI);
the aerodynamic resistances for turbulent transport of momentum and of sensible
heat fluxes
Accurate determination of these variables made the difference between good and
poor agreement with observations in many studies of latent and sensible heat
fluxes.
Leaf Area Index. Wollenweber (1995) used a one-dimensional model to study the
impact of spatial variability of land surface variables on heat fluxes. The LAI was
used to parameterize roughness length, zero plan displacement and stomatal resistance. Since these parameterizations are not linear, spatial aggregation of either
LAI or of the calculated heat fluxes will give different results. Kaneko and Hino
(1996) proposed a method to estimate evaporation based on concurrent estimates
ofLAI (inferred from NDVI), of vapour pressure deficit at leaf temperature (which
requires an estimate of the latter) and of stomatal resistance. On the other hand
Nemani et al. (1993) developed a procedure to estimate LAI using a corrected
NDVI by including spectral reflectance in the mid-infrared. They concluded that
relation of NDVI with LAI was poor in their forest area and that the proposed
method significantly improved the LAI estimates. The impact of the correction
was significant on estimates of evaporation.
Aerodynamic resistances. The most widely used parameterizations of momentum
and heat fluxes are derived from flux-profile relationships and relate mean (along
the vertical direction) flux to the difference in either wind speed or temperature
between the surface and a reference height. The accuracy of such parameterizations depends on the accuracy of the aerodynamic resistance (ra, the ratio of the
flux density to the difference in the state variable).
Kalma (1989) evaluated several parameterizations to conclude that the estimation of ra is difficult. Under conditions of large spatial variability of bare and
vegetation patches, accurate values of the resistance for momentum require accurate values of the aerodynamic roughness length (Humes et aI., 1994). The use of
thermal infrared radiometers to estimate surface temperature and sensible heat flux
density brings an additional difficulty. The proper reference for the flux-profile
relationship is an aerodynamic surface temperature, Taer, while Trad relates to directional measurements of radiance. The difference between T aer, and T rad is significant and depends on environmental conditions (Choudhury et aI., 1986).
To determine the resistance for heat fluxes the difference in temperature between
foliage and bare soil must, therefore, be taken into account. Norman et al. (1995)
described an explicit dual source heat transfer model. A different approach to estimate sensible heat flux using observations ofTrad was proposed by Lhomme et al.