Ground motion intensity measures for New Zealand

Vemula, Sreenath; Yellapragada, Meenakshi; Podili, Bhargavi; Raghukanth, S. T. G.; Ponnalagu, Alagappan

doi:10.1016/j.soildyn.2021.106928

Cited by 11 publications

(14 citation statements)

References 37 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…As the network is fixed, the input and target data are normalized using the modified Tanh estimator. The detailed procedure and the mathematical equations used for normalizing the data in this study are taken from Vemula et al (2021). The normalized input and target data used in the network for both the developed models are given in Tables 2 and 3.…”

Section: Framework and Design Of Annmentioning

confidence: 99%

“…These performance parameters are shown in Tables 6 and 7, respectively. The formulae for these performance parameters are obtained from Vemula et al (2021). From these tables, it is observed that 1.…”

Section: Performance Of the Modelmentioning

confidence: 99%

“…It is known that GA is less prone to getting stuck at local minima; the overall network has a higher chance of converging to global minima. Furthermore, this hybrid model has been successfully implemented by Vemula et al (2021) for New Zealand. However, their model computed site-to-site residuals using the Rodriguez-Marek et al (2011) approach, an approximate method.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Ground motion model for Peninsular India using an artificial neural network

et al. 2023

Self Cite

View full text Add to dashboard Cite

Ground motion models (GMMs) are an essential tool for seismic hazard analysis. They are used for developing predictive relationships to estimate the expected levels of seismic ground shaking through the ground motion parameters (GMPs). There is limited recorded data on the stable continental regions (SCR) such as Peninsular India (PI), and Central and North Eastern America (CENA), and GMMs are developed either by hybrid or stochastic methods. In this study, an effort has been made to develop a GMM for the PI region by compiling the recorded data from the CENA and PI regions, as well as from the Near Source Strong (NESS) data. An artificial neural network coupled with the genetic algorithm is considered to develop GMMs. So, the GMP-1 model is developed for various GMPs for both RotD50 and vertical components, while the GMP-2 model is developed for RotD50 components of peak ground acceleration (PGA), peak ground velocity (PGV), and the 5% damped pseudo-spectral acceleration (PSA) for periods between 0.01 and 3 s. The developed models are valid for magnitudes between 2–7.7 Mw and 0 to 1500 km Joyner–Boore distance. In addition, a parametric study is performed with the developed models for various combinations of predictor variables, and it is noticed that both the models capture the trend and attenuation pattern observed in the recorded data for all the GMPs. Moreover, the GMP-2 model predictions agree well with the recorded data compared with the candidate GMMs developed in the NGA-EAST project. A comparison of the GMP-2 models with the recorded data in the PI region indicates the robustness and effectiveness of the developed model in providing reasonable estimates of GMPs, implying its applicability to tectonic environments similar to the CENA region.

show abstract

Section: Framework and Design Of Annmentioning

confidence: 99%

Section: Performance Of the Modelmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Ground motion model for Peninsular India using an artificial neural network

et al. 2023

Self Cite

View full text Add to dashboard Cite

show abstract

“…The developed NGA‐West2 model is compared with a few records of Chi‐Chi and Christchurch ground motion along with other developed GMMs: Abrahamson et al., 3 Boore et al., 4 Campbell and Bozorgnia, 5 and Chiou and Youngs 6 in Figures 12 and 13, respectively, for different distance and soil types. Additionally, Dhanya and Raghukanth 18 model prediction is plotted for Chi‐Chi and Vemula et al 20 . prediction for Christchurch data.…”

Section: Network Performancementioning

confidence: 99%

“…Their model had lower residuals than those parametric models developed for the Central and Eastern North America (CENA) region. Vemula et al 20 developed an ANN model for the New Zealand region to predict GMPs and spectra. Their model could capture attenuation characteristics.…”

Section: Introductionmentioning

confidence: 99%

Stochastic ground motion models to NGA‐West2 and NGA‐Sub databases using Bayesian neural network

Vemula

Raghukanth

2022

Earthq Engng Struct Dyn

View full text Add to dashboard Cite

Ground motion models (GMMs) are traditionally developed from a frequentist approach. The Bayesian framework has received recent attention in developing nonergodic models, measuring uncertainty, or updating the model with additional data. However, no neural networks are developed to date in this framework to predict ground motion parameters or spectra. Hence, the present work develops a probabilistic Bayesian neural network (PBNN) to next‐generation attenuation – West2 and Subduction databases using variational inference with mean‐field assumption. Network inputs are magnitude, rupture distance, hypocentral depth, shear wave velocity, style of faulting, and region flags; outputs are peak ground values and response spectra. Both models have two hidden layers with seven neurons in each hidden layer. The models are verified for potential overfit, and their performance is validated through the parametric study by varying inputs. The output of a deterministic model is a point estimate. Considering probabilistic layers in hidden and output layers enables the model to capture within‐model epistemic uncertainty and aleatory variability. Obtained aleatory standard deviations are consistent with other models. Mean epistemic uncertainty and aleatory variability are in the range 0.07–0.10 and 0.62–0.78 (ln units) for NGA‐West2 and 0.09–0.16 and 0.67–0.95 for NGA‐Sub models, respectively. The correlation coefficients between recorded and overall mean predictions ranged from 0.94 to 0.97 for NGA‐the West2 model and from 0.91 to 0.95 for the NGA‐Sub models. Network performance for out‐of‐training inputs showed increased epistemic deviations with no effect on aleatory deviations.

show abstract

A hybrid non‐parametric ground motion model for shallow crustal earthquakes in Europe

Vemula

Podili

Raghukanth

2023

Earthq Engng Struct Dyn

View full text Add to dashboard Cite

In the current study, ground motion models (GMMs) are derived using the European Strong Motion (ESM) database for pseudo-spectral acceleration (PSA), peak ground acceleration (PGA), peak ground velocity (PGV), peak ground displacement (PGD), cumulative absolute velocity (CAV), arias intensity (I a ), and significant duration. In addition to addressing random effects associated with ground motion regression, such as inter-event, inter-site, inter-locality, and inter-region variabilities, the current study also aims at reducing the standard deviations (STDs) of the GMMs through development of a hybrid non-parametric GMM. The hybrid model is derived through an ensemble-weighted method of five non-parametric machine learning models: shallow neural network, deep neural network (DNN), gated recurrent unit (GRU), support vector, and random forest (RF) regression techniques; with weights based on model performances.The resulting hybrid model, which also accounts for epistemic uncertainty, is compared against other regional models and is found superior for all output variables. The inter-event, inter-site, inter-locality, and inter-region deviations, and total ergodic sigma of PSA for the ensemble model lies between 0.3164-0.4478, 0.4156-0.5339, 0.1449-0.3687, 0.0819-0.2421, and 0.668-0.8545, respectively. The coefficient of determination (R 2 ) between predicted and recorded values lies between 0.8435-0.9114 for all the output variables.

show abstract

Ground motion intensity measures for New Zealand

Cited by 11 publications

References 37 publications

Ground motion model for Peninsular India using an artificial neural network

Ground motion model for Peninsular India using an artificial neural network

Stochastic ground motion models to NGA‐West2 and NGA‐Sub databases using Bayesian neural network

A hybrid non‐parametric ground motion model for shallow crustal earthquakes in Europe

Contact Info

Product

Resources

About