Agricultural water requirement in arid and semi-arid environments represents an important fraction of the total water
consumption, suggesting the need of appropriate water management practices to sparingly use the resource. Furthermore
the quality and quantity of some crops products, such as grape, is improved under a controlled amount of water stress.
The latter is related, on a side to actual evapotranspiration (ET) through water demand, on the other side to plant water
content through leaf water potential. Residual energy balance approaches based on remote sensing allow to estimate the
spatial distribution of daily actual ET at plant scale, representing an useful tool to detect its spatial variability across
different cultivars and even within each parcel. Moreover, the connection between actual ET and leaf water potential is
still not well assessed, especially under water stress conditions, even if farmers use leaf water potential to plan irrigation.
However residual energy balance methods are based on the hypothesis that storage terms are negligible, at least during
the remote sensor overpass. Indeed, energy balance approaches estimate daily actual ET from the instantaneous value at
the overpass time using a daily integration method. The paper first verifies this latter assumption using field data
acquired by a flux tower on a whole phenological period. Then, the actual ET values measured by eddy covariance tower
were analyzed together with water potential measured using a Scholander chamber; the analysis highlights that, under
water stress conditions, daily actual ET is inversely linearly related with water potential. These results suggest the
possibility to use remote sensing-based ET as support for irrigation management at plot scale.