Spiking Neural Network based Water management in Drought Area of Maharashtra India: A Case Study of Osmanabad District

Devdatta K. Mokashi, Vidula Sohoni

International Journal of Intelligent Systems and Applications in Engineering · 17 February 2023

In one sentence. A spiking neural network downscales coarse precipitation to map meteorological drought (SPI) in Osmanabad, India; reported efficiency is claimed without rigorous benchmarking.

Abstract

This paper proposes a machine learning technique for drought planning and risk reduction in agricultural regions dominated by smallholders. The proposed machine learning technique is Spiking Neural Network (SNN). Analysis of spatiotemporal features and patterns of meteorological drought at resolution is the major goal of the proposed control technique, which will help local risk planning and mitigation decisions as well as activities. Using publically accessible coarse scale data, the proposed method is employed in this study to produce high-resolution gridded precipitation products, which are then downscaled for investigation of meteorological droughts. Meteorological drought is defined by the Standard Precipitation Index (SPI). The purpose of the SNN for predicts the rainfall data resolution. By using coarse-scale precipitation data as a starting point for the high-resolution, spatiotemporal analysis of droughts, the proposed method is applied and adequately general to be adopted in other areas to enable local drought risk planning and targeted mitigation-decisions as well as movements. The proposed model is run on the MATLAB/Simulink platform, and the performance is evaluated using the existing methods. The proposed method helps to predict the drought and it has more success rate than the existing methods. The efficiency of proposed method is 0.9 which is way better than the existing techniques.

Labels

Labels and summary are generated by Claude and reviewed weekly. Wrong? Edit this entry.