Monsoon Rain Prediction for Year 2020 for Marathawada India
Anand M. Sharan
Department of Mechanical Engineering, Faculty of Engineering, Memorial University of Newfoundland, ST. John’s, Newfoundland, Canada.
*Author to whom correspondence should be addressed.
Abstract
This work was undertaken to predict expected rainfall in year 2020 because Marathawada suffers frequently from drought and as a consequence, many farmers committ suicide. If rainfall is not enough then it is the farmer who suffers the most as compared to others who lend money, sell seeds etc. In doing so, sufficiently large period is used (past 32 years) in three methods other than the Artificial Neural Network (ANN) method where rainfall history back to AD 1872 is used in the analysis. In addition, some details about the four methods are also mentioned here.
To arrive at the predictions, four methods were used and their average value was used for predictions of rain for various months. The methods used were – the Fast Fourier Transform (FFT) method, the Artificial Neural Network (ANN) method, the Time Series method and the Root Mean Square (RMS) method. The result shows that this year there will be slightly more rainfall than the average of last 32 years.
Keywords: Monsoon rain prediction, annual rainfall, rainfall frequency spectrum, flood control, hydro-power generation