The Uncertainty Factors Research Of Near-field Ground Motion Numerical Simulation

The numerical simulation of ground motion is the fundamental study of earthquake occurrence,developing characteristic and regularity,which contains source,propagation path and site condition etc.The numerical simulation conclusion could be used to conform the basic parameters of strong ground motion.It is a very important basic work in seismic prevention and disaster mitigation.Now,considering indeterminate factors in the earthquake predicting process has become popular,such as many simulation method pay much attention to choose source parameters by stochastic finite fault method or empirical green method.But in the fact,most numerical simulation of strong ground motion in the modern time is earthquake inversion.Most source parameter using in simulation are deterministic.The key point to predict ground motion characteristic is choosing the appropriate source parameters.There are many large uncertainties in the future earthquake,so these uncertainty factors should be serious consideration in the predicting process of ground motion in the near field.Many methods,such as stochastic finite fault method,empirical green function,hybrid method etc,has become relatively mature,but these method are dererministic with little uncertainty factor consideration,especially some source parameters are accurate values.The simulated conclusion can’t indicate ground motion characteristic of futre earthquake clearly.Prediction equation also could indicate the attenuation relationship of many ground motion parameters,but it is also influenced by observed data,model or stochastic factors.Because of lacking the information of earthquakes and geological,people’s epistemic to earthquake occurring regularity,such as earthquake predicting,propagation path and site response,is limited.Each section of earthquake predicting has big uncertainty.For instance,different experts have different opinion about how to build a source model when they try to simulate earthquake.These different opinion and suggestion usually contain large man-made uncertainty.The integrity and reliability of earthquake information,also containing theory and model building,could result in many uncertainty factors.In this paper,the mainly research content is studying the uncertainty factors of earthquake simulation process.We research the uncertainty factors which have large influence on earthquake simulation systematic.We hope to build a source model which contains some crucial uncertainty parameters and applied these models into factual engineers.There are generally two kinds of uncertainty factors which are stochastic and epistemic factors.For the stochastic uncertainty factors,the best approach is using statistics method to get the empirical scaling law between source parameters.At the same time,because of the earth’s impenetrability and the observation approach’s limitation,there are also some epistemic uncertainty factors in source model of earthquake.We will use logical tree method to manage these epistemic uncertainty factors and take the optimal scheme as the prediction results of earthquake ground motion.The concept of strong ground motion zone was introduced.For the epistemic uncertainty factor,we mainly analyzed the asperity.The asperity source model was also established1.Using Empirical Green Function method to simulated the Kagoshima Mjma6.5 earthquake of Kyushu,Japan,in 1997.This simulated process proved that this method was useful when simulated strong earthquake.All the data come from K-NET of Japan.The reason to choosing this earthquake is that this earthquake has plentiful earthquake data and their quality is very good.We also analyzed much more engineering parameters which were widely used in earthquake engineering field.The simulated results indicated the ground motion characteristic.2.Basing on a lot of earthquake record and articles,we used stochastic method to get the empirical relationship of source parameters in earthquake dense regions,especially the Longmen mountains fault zone.3.We used the Kumamoto earthquake which the magnitude is 7.3 to analyze the key source parameters which may had great influence on ground motion by sensitivity approach.We will allot much weight to this important source parameters in the future earthquake predicting process to improved the accuracy degree and reliability of simulated results.4.Basing the earthquake data of Wenchuan,we studied the influence degree of epistemic uncertainty of asperity by logical tree method.Then we build the asperity source model.We regarded the asperity equal to strong motion generation area,which this area was proposed by Hiroe Miyake,We mainly study the area of asperity,numbers,stress drop,rise time.We finally got the best model of asperity source model.The biggest innovation point is putting the uncertainty factors processing method which is used in probability seismic hazard analysis into future earthquake numerical simulation in the near field.The stochastic uncertainty factors will be managed by statistical method while epistemic uncertainty factors will be managed by logical tree method.The numerical simulated results could indicate well the ground motion characteristic of future large earthquake.The simulated results have a better and higher reference values in the future earthquake resistance and disaster prevention.Considering the uncertainty factors of predict the future devastating earthquakes,so we can evaluate the comprehensive characteristic of ground motion quantitatively when the future earthquakes occur.It could provide scientific reference to earthquake prevention and disaster reduction.It also could provide scientific reference to engineering design decisions.