Time series analysis of monthly rainfall at Pantnagar using ARIMA and ANN
Loading...
Date
2017-06
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
G.B. Pant University of Agriculture and Technology, Pantnagar - 263145 (Uttarakhand)
Abstract
Planning and design of the water resources projects need accurate information about different hydrological events which are not governed by known physicochemical laws. Hydrologic modelling of climatological data, as a stochastic process, is of interest for a variety of hydrological research area such as flood routing reservoir operation and agricultural planning at watershed scale. Time series analysis plays an important role at this regard. Keeping this in view, an attempt has made to develop a time series model for Pantnagar by using ARIMA and ANN. The monthly rainfall data from (1961-2014) were used for the analysis. Various tests such as Mann-Kendell’s trend test for trend analysis. Stationarity in time series was tested by Augmented Dickey Fuller (ADF), Kwiatkowski-Philips-Schmidt-Shin (KPSS) were used for development of ARIMA model. Gamma test is done for input selection for ANN model. The performance of developed model is judged by using statistical indices such as Root Mean Squared Error (RMSE), Coefficient of Determination (R2), and coefficient of Efficiency (CE). Akaike Information Criterion (AIC), Bayesian Information Criterion (BIC) or Schwarz criterion, Pooled Average Relative Error (PARE). Data from 1961 to 2012 were used for development of model. By using developed model for the years 2013-2014 were forecasted. The results showed good agreement between observed and predicted. Overall ANN performance is superior to ARIMA.
Description
Keywords
null