Pan Evaporation Estimation Using Artificial Neural Network (ANN) and Fuzzy Logic Models for Raichur Region, Karnataka: A Case Study

Abstract

Aims: Accurate estimates of evaporation by employing efficient and proven soft computing techniques that involve least number of influencing variables are important to tackle present water crisis.
 Place and Duration of Study: In the present study, Artificial Neural Network (ANN) and fuzzy logic models were developed to predict the pan evaporation (Ep) in Raichur, Karnataka, using six input parameters viz., maximum and minimum temperatures, maximum and minimum relative humidity, sunshine hours and wind speedfor the period of 30 years (1990-2019).
 Methodology: Comparison between models was done to select best suitable model to predict pan evaporation. The ANN models were trained withthree training algorithms. Gaussian membership function was used in fuzzy logic (FL) model.
 Results: The results revealed that, the ANN-GDX model performed better over ANN-LM, ANN-BR and fuzzy logic models during validation period. The correlation coefficient (r), coefficient of efficiency (CE), mean absolute error (MAE) and root mean square error (RMSE) were observed to be 0.7637, 0.5831, 1.3880 and 1.8541 respectively during validation period between actual and predicted pan evaporation (Ep) with 1.3880 mm root mean square error. Therefore, ANN-GDX model was chosen for predicting pan evaporation in the study area.
 Conclusion: ANN-GDX model was chosen for predicting pan evaporation in the study area.