We used multiple regression analysis to relate evapotranspiration( ET), com
puted from a water balance technique, to both thermal infrared and normaliz
ed difference vegetation index data obtained from the Advanced Very High Re
solution Radiometer (AVHRR) sensor on board on the National Oceanic and Atm
ospheric Administration (NOAA) satellite. This approach, based on only remo
tely sensed data, provided a reliable estimate of ET over the Pampas, the m
ain agricultural region of Argentina. The relationship between spectral dat
a and ET was more sensitive to the dates than to the sites used to generate
the models.