Application of artificial neural networks to predict the probability of Extreme rainfall and comparison with the probability by Fisher Tippet Type-II distributions- Case study at Anand station of Gujarat, India

摘要

An attempt has been made here to predict the return period for highest one day maximum rainfall of 58 stations of 14 districts of Gujarat state of India covering eight agro climatic zones. The rainfall data of highest one day maximum rainfall of 58 stations from 1901-1992 were subjected to Fisher and Tipett Type-II distribution and Artificial Neural Network (ANN) method was applied. Some results of predicted return periods by the Fisher and Tipett Type-II ($T_F$) were used for training of the neural network. Results by Artificial Neural Network were compared with the $T_F$. During the analysis Standard Errors were computed. Return period $T$ obtained by Artificial Neural Network with supervised networks has non- significant difference with? $T_F$?at 14 places. Here, computations of S.E.s were found very less that ranges between 0.2 to 2.2 mm.Predicted return period by ANN, at some places namely, Modasa and Prantij stations of Sabarkantha district has very high return period (210 yrs and 200 yrs). This is because of very high value of daily maximum rainfall (1026.2 mm, 781.6mm).?Artificial Neural Network is applicable to predict the return period except at some places of Gujarat state of India.