High-quality seismological recorded dataset analysis for the estimation of peak ground acceleration in Himalayas

This research aims to see how well back-propagation neural network (BPNN) approaches perform in estimating peak ground acceleration in the Himalayan region using various forms of input data (i.e., magnitude, focal depth, hypocentre and S-wave avg. velocity). These criteria were derived from the Himalayan Earthquake Catalogue, which includes all minor and significant earthquakes and their aftershock sequences in prior years. The analysis of high-quality seismological recorded data is important for earthquake prediction. The BPNN is employed for predicting the strong ground motion parameters such as peak ground acceleration, peak ground velocity and peak ground displacement using different activation functions at different layers. This research paper focuses on the model's training to analyze the dataset, including a few parameters to estimate the acceleration of the ground. Translated data into seismicity indicators, which are mathematically derived parameters. These seismicity indicators were utilized for the training of the BPNN to improve decision-making and for peak ground acceleration estimation. Artificial neural network models were used to predict the ground motion intensity measures for future events. The predictive parameters' contributions to the prediction of the considered intensity measurements are finally explored via sensitivity analysis. This study uses the multi-layer BPNN model to discover components in the Himalayan region, using the actual incidence of earthquake seismicity indicators as input and goal vectors. Finally, the BPNN model produces a good estimate for the peak ground acceleration of earthquakes.