The study attempted to estimate the actual crop evapotranspiration from minimum weather data of Tavanur, Kerala and resulted in an Artificial Neural Network (ANN) model, which makes use of average temperature data to estimate the actual evapotranspiration. The effectiveness of this model was evaluated using various statistical indices. The results of this model were compared with various existing techniques. The analysis led to the conclusion that the ANN models were performing better than all existing techniques for computing the actual evapotranspiration. However, the study was based on a single season lysimeter data and more research work may be required to reinforce this conclusion.
A neural network model was developed and analyzed to estimate the daily values of rice crop evapotranspiration from minimum meteorological data. A radial basis function network was employed in the study. Three combinations were selected based on a detailed review of research work in this area of study. The results from each of these models were compared with actual lysimeter observations and it was found that the ANN model with temperature as sole input neuron estimated the lysimeter measured crop evapotranspiration effectively.
Download the report here:
/articles/investigation-capability-artificial-neural-network-technique-estimating-evapotranspiration