|
Analysing and Forecasting of Rainfall in Telangana State
|
|
| Author:
|
P.RAGURAMAN , C.SURESH KUMAR, P.KARUNAKAR REDDY, R.NAVYA SREE, K.VENKATA RATNAM
|
| Abstract:
|
India, which is in the tropical wet and dry area, receives a huge amount of precipitation each year, with the monsoon season being the main contributor. The environment resulting challenges also has led to significant improvements in the average rain distribution and its volatility, as well as the severity and frequency of severe rainfalls. Rain also exhibits great temporal and geographic volatility. Different Time-series prognostication models for predicting rain, including Holt's Linear Trend methodology (HLTM), Generalized Auto-Regressive Conditional Heteroskedasticity (GARCH), Holt's Winter Seasonal Methodology (HWSM), and seasonal Autoregressive Integrated Moving Average, were used in this study (SARIMA). In order to analyse the most basic time-series prognostication model, a comparison analysis was built. The HWSM model was determined to have the lowest error rate when compared to the other models. Mean square error (MSE), Root Mean square error (RMSE), and Mean Absolute Error are the analysis criteria used to compare different time-series rain forecast methods (MAE). The HWSM model exhibited, by far, the lowest error rate of the other models, with error rates of 4.767 and 4.343 RMSE, 1.51 and 1.432 MAE, and 25.65 and 24.75 MSE for datasets 1 and 2, respectively.
|
| Keyword:
|
Primarily on Agriculture, Concludes the Analysis, SARIMA.
|
| EOI:
|
-
|
| DOI:
|
https://doi.org/10.31838/ijpr/2019.11.01.308
|
| Download:
|
Request For Article
|
|
|