**Assessment and Prediction of Evapotranspiration Based on Scintillometry and Meteorological Datasets**

Antonin Poisson, Angel Fernandez, Dario G. Gomez, Régis Barillé and Benoit Chorro

Additional information is available at the end of the chapter

http://dx.doi.org/10.5772/intechopen.68538

#### Abstract

Two methods are used for estimating the evapotranspiration (ET) rate: scintillometry and meteorological measurements using the FAO-PM56 model with the reference evapotranspiration for the crop (ETO) and the specific coefficient (Kc) for corn at its stage development. Measurements were done on a field with homogeneous corn crop at the stage of 3 months before the final harvest (65 % of maximum plant growth). The two methods are compared with environmental parameters to determine the most influential on the final result of ET.

A coefficient of 0.78 is found between the two methods resulting of an underestimation of the evapotranspiration values with FAO.

The sensitivity for the two measurements are compared in order to determine how sensitive the output calculation of evapotranspiration could be with respect to the calculation elements which are subject to uncertainty of variability in the input environmental parameters. The scintillometer uncertainty is lower than the FAO-56 uncertainty.

Finally, a model based on an artificial neural network (ANN) forecasting ET is developed in order to anticipate the necessary action for water management. It leads to the conclusion that scintillometry is more able to predict evapotranspiration on short and medium time than the FAO-PM56 method.

Keywords: evapotranspiration, FAO-PM56, scintillometry, artificial neural network
