S T G Raghukanth

This work aims at developing a hybrid ground motion prediction equation (GMPE) for spectral acceleration in Western Himalayas and North-Eastern India. The GMPE is derived using an efficient nonparametric modelling based on neural network algorithm. In this study, owing to sparsity in the recorded ground motions (498 recordings) for the region, the available information is combined with 13,294 records from the well-tested NGA-West 2 database. For the methodology adopted in the study, regional flags are assigned to the records. Thus, given a magnitude, distance, shear wave velocity, fault type and region, the model is able to predict the possible spectral acceleration. The developed GMPE is observed to be unbiased with respect to region. Further, the inter- and intra-event standard deviations are also in acceptable ranges. It is observed that developed GMPE for Western Himalayas and North-Eastern India is able to capture all the known ground motion characteristics. Additionally, the GMPE is compared with the existing GMPE for rock-type soil condition available for the Western Himalayas and North-Eastern India. Furthermore, applicability of the developed GMPE model in estimating hazard is analysed by obtaining the uniform hazard response spectra for Delhi and Guwahati.

The present work aims at developing the first probabilistic fling hazard map of India and adjoined regions. First, we developed a new ANN-based ground motion prediction equation (GMPE) for fling corresponding to horizontal and vertical directions. The developed GMPE is based on permanent ground residual displacement form 556 scenario events considered consistent with the regional characteristics. The corresponding simulations are performed by suitably combining the Okada’s solutions. Developed GMPE is comparable with the existing relations and the few available data which contained fling characteristics. Further, the developed GMPE, along with the other two available prediction equations for the fling, is incorporated to represent the ground motion characteristics in the estimation of hazard using a suitable logic tree. In addition to the fling prediction equations, the evaluation of regional fling hazards requires identifying the location of all the probable seismic sources and their seismicity characteristics. In this study, we used the linear-fault model, as the fling is a near field phenomenon. We report the resultant probabilistic fling hazard map for 10%,2% 1%, and 0.5% probability in 50 Years for the region. The maps showed that the active regions in Himalayas, North-Eastern India, and Andaman experience higher values for fling than stable Peninsular India. Thus, this study develops the fling hazard map for the first time, and the results are essential in the design and rehabilitation of important structures in the region.

A deep learning model is developed for the Next Generation Attenuation–Subduction database for predicting spectral accelerations and peak amplitude measures. The developed model satisfies the statistical criteria necessary for prediction. Standard deviations lie in 0.2864–0.3809, 0–0.2696, and 0.4514–0.7892, range for inter-event, -region, and intra-events, respectively. Transfer learning is applied to the New Zealand region. Probabilistic seismic hazard analysis is performed for the Andaman-Nicobar region and obtained a peak ground acceleration of 0.6–0.7 g and 0.4–0.5 g at the Andaman and the Nicobar Islands, respectively, for a 2475-year return period.