Abstract:
Artificial Neural Network based modeling technique has been used to study the influence of different combinations of meteorological parameters on evaporation. The data set used is taken from the records of NIH Observatory. The input combination selected based on the statistical properties of data was used in predicting the evaporation. The prediction accuracy of Artificial Neural Network has also been compared with the accuracy of multiple linear regressions for predicting evaporation. The comparison demonstrated superior performance of Artificial Neural Network over multiple linear regression approach. The findings of the study also revealed the requirement of all input parameters considered together, instead of individual parameters taken one at a time as reported in earlier studies, in predicting the evaporation
The RMSE of ANN model during calibration and validation was found to be 1.0151 and 1.0045 respectively, whereas for the MLR model, RMSE value during calibration and validation was
1.1874 and 1.1523 respectively, and also the ANN model efficiency during calibration and validation was 0.7800 and 0.7830 respectively, whereas the MLR model efficiency during calibration and validation was 0.6988 and 0.8501 respectively, indicates a substantial improvement in the model performance. In addition, comparison of the scatter plots of time series indicates that the values of Evaporation estimated by the ANN model are more precise than those found by the MLR.
The highest correlation coefficient (0.7800), (0.7830) along with lowest root mean square error (1.0151),(1.0045) during calibration and validation of best evaporation model [4-4-1] was obtained with the input combination of rainfall, maximum temperature, minimum temperature, and mean relative humidity. A graph between the actual and predicted values of evaporation suggests that most of the values lie within a scatter of ±15% with all input parameters. The findings of this study suggest the usefulness of ANN technique in predicting the evaporation losses.
