Recognition Of Physically Connected Independent Climate Variables For Seasonal Monsoon Rainfall And Future Forecast Through Ann Modeling Over Rewa District

Research Article
Tiwari, RK and Karmakar, SK
DOI: 
http://dx.doi.org/10.24327/ijrsr.2017.0812.1210
Subject: 
science
KeyWords: 
Neural Network, Monsoon Rainfall, Forecasting, Climate Variables, Modeling.
Abstract: 

Dependency of seasonal long range rainfall (in mm.) with parameters of climate variables such as Sun Spot Number, Cosmic Ray Intensity, Geomagnetic indices, Maximum Temperature, Minimum Temperature, Maximum Relative Humidity, Minimum Relative Humidity, Wind Speed, are examined over Rewa District. Wherein, merely Cosmic Ray Intensity, Geomagnetic indices, Minimum Temperature, and Maximum Temperature, have found physically connected with monsoon rainfall for long period while, Sun Spot Number, Relative Humidity, and wind speed are not providing any influence. Thus an ANN Modeling to forecast future monsoon rainfall over this region is established through these physically connected independent parameters. It is found that ANN modeling was performance up to 74% and 73% accuracy during training and testing period respectively. The skeleton of entire ANN modeling and its performances are presented through this research paper.