Estimation of pan evaporation using ANN – a case study

Abstract

A study on ‘Estimation of pan evaporation using Artificial Neural Network’ was carried out in K. C. A. E. T, campus, Tavanur. The meteorological parameters considered were monthly data of wind speed, dry bulb temperature, wet bulb temperature, maximum temperature, minimum temperature and evaporation for six years (February 2003 to January 2009) for the model development. The entire data was divided in to three based on the season, as pre- monsoon, monsoon and post monsoon. The five different strategies were considered for the study viz. M-1, M-2, M-3, M-4, and M-5. Strategy M-1 included all the input parameters. M-2 included all the input parameters except wind speed and M-3 included wind speed, dry bulb temperature and wet bulb temperature. Dry bulb temperature and wet bulb temperature were only considered in strategy M-4 where as strategy M-5 included only maximum temperature and minimum temperature. From the available data 70% data were taken for training and 15% each for testing and validation. From the results of the study, it can be seen that the strategy M- 3, combination of wind speed, dry bulb and wet bulb temperature has greater influence on evaporation rate the least MSE value and high R 2 value.