INDIAN JOURNAL OF PURE & APPLIED BIOSCIENCES
ISSN (E) : 2582 – 2845
Indian Journal of Pure & Applied Biosciences (IJPAB)
Year : 2019, Volume : 7, Issue : 5
First page : (105) Last page : (112)
Article doi: : http://dx.doi.org/10.18782/2320-7051.7690
Application of Artificial Neural Networks for Rainfall Modelling
Paradkar V. D.*, Mittal H. K., Singh P. K., Mahesh Kothari and Jain H. K.
College of Technology and Engineering, Maharana Pratap University of Agriculture and Technology,
Udaipur, Rajasthan, India
*Corresponding Author E-mail: email@example.com
Received: 11.07.2019 | Revised: 16.08.2019 | Accepted: 24.08.2019
Artificial Neural Networks (ANNs) have been tried by researchers for rainfall forecasting in different parts of the world. In this paper, three categories of model are tried for rainfall forecasting of Pratapgarh district by varying number of inputs. In the first model (Model-A), rainfall of same week of previous three years and rainfall of three preceding weeks of same year used as input. In the second model (Model-B), rainfall of same week of previous four years and rainfall of four preceding weeks of same year used as input. While, in the third model (Model-C), rainfall of same week of previous five years and rainfall of five preceding weeks of same year used as input. The number of hidden layer neurons varied from 1 to 20. The performance of models was tested by using statistical indices such as R, RMSE and MAE. The obtained results of the models showed that, increasing number of inputs significantly improved the performance of models. It was concluded that, the model performance was not significantly affected by increase in number of neurons in hidden layer.
Keywords: Rainfall, Modelling, Artificial Neural Network.
Full Text : PDF; Journal doi : http://dx.doi.org/10.18782
Cite this article: Paradkar, V.D., Mittal, H. K., Singh, P.K., Kothari, M. and Jain, H.K. (2019). Application of Artificial Neural Networks for Rainfall Modelling, Ind. J. Pure App. Biosci.7(5), 105-112. doi: http://dx.doi.org/10.18782/2320-7051.7690