International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Vol XXXV, Part B7. Istanbul 2004 
The potential of using drainage systems, i.e. the management 
of water table depth, to function as sub-irrigation system 
without increasing the accumulation of salts in the root zone 
and is a valuable agricultural resource in arid regions (Benz 
et al., 1984). 
Plant water stress can limit productivity in both natural and 
agronomic plant communities.: Short term as well as long 
term water stress has the same effects on plant physiology 
and canopy architecture. Changes in water status of a canopy 
can have indirect effects on remotely sensed optical 
reflectance and thermal emittance. 
Surface temperature (T,) is a major component in the energy 
balance equation. Several models have been developed to 
evaluate crop water use, water stress, crop yield and soil 
moisture (Reginato et al, 1976; Idso et al, 1981; Price, 
1982; Reginato et al., 1985; Jackson ef al., 1987). 
Several authors have investigated the combination of the 
thermal band 6 of Landsat TM with the reflective bands in 
the Red and NIR, band 3 and band 4, respectively. The 
relationship between NDVI and T, is linear with negative 
slope (Gurney ef al., 1983; Hope, 1986; Moran ef al, 1990), 
This relationship could be diagnostic of plant water stress, 
particularly, at large NDVI values. For a given 
meteorological condition, the surface of bare soil would tend 
to have a stable maximum surface temperature once soil 
moisture is depleted. Accordingly, spatial variability in plant 
available moisture (in the root zone) would not be reflected 
in the T, of pixels dominated by bare soil (at small NDVI 
values). Moreover, an increase in vegetation moisture stress 
will cause the T, of pixels with large NDVI values to 
increase. For partially irrigated fields and homogeneous crop 
cover, the NDVI would be relatively unaffected by soil 
moisture difference, whereas T, values would be low over 
the irrigated portion and high over the dry portion (Moran et 
al, 1990). Moreover, this was explained in terms of the 
increase in the latent heat flux (LE) associated with greater 
amount of transpirationally active vegetation (Hope and 
McDowell, 1992). Table 1 list and define these vegetation 
indices. 
1144 
In recent studies, the soil adjusted vegetation index (SAVI) 
has been used in the combination of spectral and thermal 
bands (Choudhury, 1994 and Moran ef al., 1994). The SAVI 
has the advantage of being more sensitive to the increase in 
vegetation cover and less sensitive to spectral changes in the 
soil background than the NDVI (Huete, 1988). Moran ef al. 
(1994) 
index/temperature trapezoid (VIT) in an attempt to combine 
proposed the concept of the vegetation 
spectral vegetation index with composite surface 
temperature. Since spectral vegetation indices are non- 
linearly related to vegetation cover (Vc) then Vc is 
substituted in the Y-axis in the trapezoidal relationship 
between Vc and the temperature difference 
(T. -T.). 
Recent research has examined technologies involving remote 
sensing to quantify water stress. Moran et al. (1989) 
investigated the effect of water stress on canopy architecture 
in alfalfa (Medicago sativa L.) and the sequential effect on 
canopy reflectance. They found water-stressed canopies to 
have a lower spectral reflectance in the NIR and red 
wavebands when compared with unstressed canopies. A ratio 
of the two wavebands was most successful in estimating the 
onset of stress. Moran et al. (1994) investigated the concept 
of a water deficit index, which is defined as the ratio of actual 
to potential ET. This index exhibits the ability to predict ET 
rate and relative field water deficit for both full-cover and 
partially-vegetated sites. The measurement can be calculated 
from remotely sensed data (red and NIR) gathered with 
ground, aircraft, or satellite-based sensors. On-site 
measurements used in the calculation include net radiation, 
air vapor pressure deficit, air temperature, and wind speed. 
The results of Shakir and Girmay-Gwahid (1998) showed 
that in the wave length range of 850 - 1150 nm the stressed 
plots showed lower reflectance than unstressed plots. 
However the reflectance of stressed plots was higher above 
the 1150 nm. 
This study aims to studiy the effect of varying water table 
depth alfafa spectral reflectance. And to examine if water 
stress is likely to occure due to the variations of watertable 
levels.