Forecasting monthly rainfall using autoregressive integrated moving average model (ARIMA) and artificial neural network (ANN) model: A case study of Junagadh, Gujarat, India
Journal Title: Journal of Applied and Natural Science - Year 2019, Vol 11, Issue 1
Abstract
The onset, withdrawal and quantity of rainfall greatly influence the agricultural yield, economy, water resources, power generation and ecosystem. Time series modelling has been extensively used in stochastic hydrology for predicting various hydrological processes. The principles of stochastic processes have been increasingly and successfully applied in the past three decades to model many of the hydrological processes which are stochastic in nature. Time lagged models extract maximum possible information from the available record for forecasting. Artificial neural network has been found to be effective in modelling hydrological processes which are stochastic in nature. The ARIMA model was used to simulate and forecast rainfall using its linear approach and the performance of the model was compared with ANN. The computational approach of ANN is inspired from nervous system of living beings and the neurons possess the parallel distribution processing nature. ANN has proven to be a reliable tool for modelling compared to conventional methods like ARIMA and therefore ANN has been used in this study to estimate rainfall. In this study, rainfall estimation of Junagadh has been attempted using monthly rainfall training data of 32 years (1980-2011) and testing data of 5 years (2012-2016). A number of ANN model structures were tested, and the appropriate ANN model was selected based on its performance measures like root mean square error and correlation coefficient. The correlation coefficient Seasonal ARIMA (1,0,0)(3,1,1)12 and ANN back-propagation model (5-12-1) on the testing data was found to be 0.75 and 0.79 respectively. Seasonal ARIMA (1,0,0)(3,1,1)12 and ANN back-propagation model (5-12-1) were used for forecasting rainfall of 5 years (2017-2021).
Authors and Affiliations
D. Dwivedi, J. H. Kelaiya, G. R. Sharma
Keywords
Related Articles
- EP ID EP485125
- DOI 10.31018/jans.v11i1.1951
- Views 82
- Downloads 0
Biocontrol of toxigenic strain of Aspergillus flavus isolated from the root tubers of safed musli (Chlorophytum borivilianum Sant. F) using its rhizospheric mycoflora
Miraculous herb safed musli (Chlorophytum borivilianum Sant. F), family liliaceae, is well recognized for its immense potential as an aphrodisiac. The root tubers of this herbal drug were found to be invested with Asperg...
Grain yield of aerobic rice as influenced by seed rate and row spacing in aerobic situation under changed climate
Aerobic rice system is the method of cultivation, where the rice crop is established by direct seeding in un-puddle field. The grain yield of aerobic rice in aerobic situation realized by the farmers is still lower. Amon...
Impact of organic techniques of seed crop management on seed yield and quality in rice
A study was conducted at the Department of Seed Science and Technology, Tamilnadu Agricultural University, Coimbatore in June - July season of two different years to find out suitable organics for enhanced productivity o...
Influence of agronomic practices on severity of late blight of potato (Phytophthora infestans)
An experiment was carried out to find out the influence of such agronomic practices on severity of the disease and to seek a proper cultural practice for management late blight of potato (Phytophthora infestans). The exp...
Compatibility of biocontrol agents and N fixing organisms with post emergence pre-mix herbicide-bispyribac sodium + metamifop 14 % SE
The experiments were conducted in vitro in the Agricultural Microbiology laboratory at College of Agricul-ture, Vellayani, Thiruvananthapuram, Kerala, India to evaluate the compatibility of biocontrol agents Pseudomonas...