temperature for Smoke Creek Dailey and Railroad Valley respectively. Effective surface
temperature seems to be slightly lower than radiative surface temperature especially for large
Figures 3, and 4 chow a cross-plot between effective and radiative surface
temperatures. This is most probably due to the fact the soil under the shrubs and the shrubs
temperatures are not exactly equal. However, the RMSE between radiative and effective
surface temperature does not exceed 1.6 (C) for both sites.
These 2 figures suggest that a thermal radiometer pointed vertically above
such heterogeneous surfaces may actually measure an effective surface temperature that can
be directly used to estimate surface fluxes. Similar result recently have been reported by
Chehbouni et al., (1993) in their investigation during the Hapex Sahel experiment. However,
this is in direct contrast with what has been reported by Hall et al., (1992) about the
unfeasibility of deriving accurate surface fluxes from remotely sensed Thermal Infrared data.
In our view, accurate estimation of surface fluxes over heterogenous surfaces relies on 3 key
points:
1) A proper assessment of all the surface component's that interact with the
atmosphere.
2) An accurate parameterization of the resistances to heat and water transfer
(Seguin, 1993).
3) A better understanding of the information contained in radiometric
measurements made over heterogeneous surfaces.
over two sites with sparse and non-homogeneously distributed shrubs in Nevada. The surface
was represented by two adjacent compartments. The first compartment is made up of two
components (shrubs and underlying soil), and the second is made up of open "illuminated"
soil. The results show that this model can provide an accurate estimation ot sensible heat
flux over this type of surface. However the generalizations of this model to other
heterogeneous surfaces requires 2 conditions:
1- The .scale of surface heterogeneity allowed atmospheric forcing parameters such as win
speed, air temperature and humidity, incoming short and longwave radiotion, to be
considered as constant over the entire surface (des-orgamsed variability).
2- Each compartment interacts with the atmosphere in quasi-independent manner from the
other. In other words honzontal transfer are too small compared the vertical transfer
therefore inter-compartment advection can be neglected.
This study also show that effective surface temperature can be legitimately approximated by
radiometric surface temperature. This has considerable implications for the use of remotely
sensed surface temperature to estimate surface fluxes. However, additional studies are
needed to confirm the pertinence of these results.
ACKNOWLEDGMENTS
This research was conducted at the Jet Propulsion Laboratory, California Institute of Technology, under contract
with the National Aeronautics and Space Administration (NASA). This research is situated within the framework
of NASA-EOS hydrology project (OSSA- 1 / 88 ). The first author is currently at JPL. as a NRC Resident
Research Associate. Funding was also provided by ORSTOM and the French PNTS
5- CONCLUDING REMARKS
In this study, a three-component model was used to estimate sensible heat flux
783
