Among plant water stress criteria used for irrigation scheduling, crop water stress index (CWSI) has the main advantage to be of easy use.
On the other hand, it has the disadvantage to be highly dependent on weather conditions during time of measurement.
The authors present a way of calculation of CWSI which allows to obtain a relation to relative evapotranspiration which comes closer to theoritical values from Jackson (1982) and which is not as dependent on wind velocity conditions as is the classical way.
This is supported by results of measurements performed over Soybean during three drying cycles in summer 1989 at Mas d'Asport near Nîmes (South of France). Replacing VPD estimates of minimum temperature (Idso, 1982) by direct radiative temperature measurements over small irrigated plots not only increases the correlation to relative evapotranspiration (rET) but, in addition, partly corrects the influence of soil surface temperature for small LAI values.
Then, the accuracy of rET estimates by means of CWSI reaches that given by rET relation to predawn leaf water potential (Itier et al., 1990, 1992). consequently CWSI becomes a more reliable and competitive plant water stress index.