Wednesday, 2 March 2022

Computational Rainfall Forecasting Models | Chapter 03 | Recent Recent Advances in Mathematical Research and Computer Science Vol. 8

 Rainfall forecasting is critical for planning and understanding rainfall variability, which aids agricultural management in making decisions. Rainfall forecasting is a key scientific concern in meteorology. Many academics sought to develop linear models to predict rainfall, but the revelation of nonlinearity in the nature of weather data has changed the attention to nonlinear rainfall prediction. Through a simple procedure utilising historical data, neural networks can automatically construct a forecasting model. This type of training allows the neural system to capture complex and non-linear correlations that are difficult to examine using traditional approaches. Neural networks are a novel type of computational technology that can be used to investigate the dynamics of a wide range of hydrological applications. An Artificial Neural Network (ANN) is a massively parallel, distributed information processing system with a highly flexible architecture that captures nonlinearity well. In hydrologic modelling, Artificial Neural Networks with one hidden layer are often utilised. The standard data pre-processing approach of Moving Average (MA) was combined with Artificial Neural Network as MA – ANN model in this research study to improve rainfall forecast in Tamilnadu. The MA – ANN hybrid model is a better tool than the Moving Average and ANN models when used separately, according to the experimental results.



Author(S) Details

M. Nirmala
Sathyabama Institute of Science and Technology, Chennai, India.

View Book:- https://stm.bookpi.org/RAMRCS-V8/article/view/5743

No comments:

Post a Comment