A logistic regression classifier for long-term probabilistic prediction of rock burst hazard

Rock burst is a complex dynamic process can lead to casualties, to failure and deformation of the supporting structures, and to damage of the equipment on site; hence, its prediction is of great importance in underground construction. We present a novel empirical method to predict rock burst based on the theory of logistic regression classifiers. An extensive database collected from the literature, which includes observations about rock burst occurrence (or not) in underground excavations in projects from all over the world, is used to train and validate the model. The proposed approach allows us to compute new class separation lines (or planes) to estimate the probability of rock burst, using different combinations of five possible input parameters—tunnel depth, H; maximum tangential stress, MTS; elastic energy index, W _et; uniaxial compressive strength of rock, UCS; uniaxial tensile strength of rock, UTS—among which it was found that the preferable model could be developed in H–W _et–UCS space. The proposed model is validated with goodness-of-fit tests and nine-fold cross-validation; results show that its predictive capability compares well with previously proposed empirical methods and confirm that, as expected, the probability of rock burst increases with excavation depth, and that both W _et and UCS have a similarly significant influence on rock burst occurrence. Finally, expressions are proposed for identification of conditions associated with several reference values of rock burst probability, which can be employed in preliminary risk analyses.     

Seismic source modeling and probabilistic seismic hazard analysis for Bangladesh

Natural Hazards - The probabilistic seismic hazard analysis (PSHA) has been performed for Bangladesh using background seismicity, crustal fault, and subduction zone source models. The latest ground...     

A proposed probabilistic seismic tsunami hazard analysis methodology

Natural Hazards - We expand on existing probabilistic tsunami hazard analysis (PTHA) approaches by presenting a methodology wherein the epistemic uncertainties associated with tsunami timing,...     

Deterministic and probabilistic seismic hazard analysis for Gwadar City, Pakistan

Gwadar City is located at the coastline of Pakistan. The city is currently in a phase of development, which is expected to become a future economic hub for Pakistan. This has led us to choose Gwadar for seismic hazard evaluation. Seismic hazard analysis for Gwadar is carried out using deterministic and probabilistic seismic hazard analysis techniques. The present study will help in sustainable development of a future large city and economic hub for Pakistan on ways of coping from a major threat of earthquake hazard. In deterministic seismic hazard analysis, line sources were identified close to Gwadar. Based on the analysis of maximum magnitude and closest distance (worse conditions), Makran subduction zone was identified out of all the line sources with earthquake potential of 8.2 at a distance of 30 km. This yielded a peak ground acceleration value of 0.38 g for Gwadar City. In second phase, probabilistic seismic hazard analysis technique with the area source modeling was adopted to acquire results at different return periods. For this purpose, seismic data were collected from the Pakistan Meteorological Department and International Seismological Center (2010) databases for development of a comprehensive data catalog. The a and b values were obtained using regression analysis for each source zone, and probabilistic analysis yielded the results of 0.34 g for a return period of 500 years. As per building codes of Pakistan, areas or cities with ground acceleration greater than 0.32 g are considered in seismic zone 4, and both deterministic and probabilistic hazard analysis place the city in seismic zone 4. These values correspond to rock site with shear wave velocity of 760 m/s.     

Near-field probabilistic seismic hazard analysis with characteristic earthquake effects

In conventional seismic hazard analysis, uniform distribution over area and magnitude range is assumed for the evaluation of source seismicity which is not able to capture peculiar characteristic of near-fault ground motion well. For near-field hazard analysis, two important factors need to be considered: (1) rupture directivity effects and (2) occurrence of scenario characteristic ruptures in the nearby sources. This study proposed a simple framework to consider these two effects by modifying the predictions from the conventional ground motion model based on pulse occurrence probability and adjustment of the magnitude frequency distribution to account for the rupture characteristic of the fault. The results of proposed approach are compared with those of deterministic and probabilistic seismic hazard analyses. The results indicate that characteristic earthquake and directivity consideration both have significant effects on seismic hazard analysis estimates. The implemented approach leads to results close to deterministic seismic hazard analysis in the short period ranges (T < 1.0 s) and follows probabilistic seismic hazard analysis results in the long period ranges (T > 1.0 s). Finally, seismic hazard maps based on the proposed method could be developed and compared with other methods.     

Regional attributes of hurricane surge response functions for hazard assessment

Accurate quantification of hurricane surge probabilities is critically important for coastal planning and design. Recently, the joint probability method has been shown to yield statistically reliable surge probabilities and has quickly become the method of choice for extreme-value surge analysis in the United States. A main disadvantage of the joint probability method is the requirement to have accurate computational surge simulations for a large array of hurricane conditions. Recently, this shortcoming has been overcome by using a variety of interpolation schemes to reduce the number of surge simulations required to an optimal sample for joint probability analysis. One interpolation scheme uses response functions, or physically based dimensionless scaling laws, that consider the relative impact of hurricane landfall position, central pressure, and storm size on surge magnitude at the location of interest. Here, the influence of regional changes in bathymetry on the physically based response function form is investigated. It will be shown that the influence of continental shelf width on surge generation along a continuous coast is coupled with the influence of storm size and that this coupled physical effect can be treated within the response functions via dimensionless scaling. The surge response function model presented here has an algebraic form for rapid calculation. This model performs well for the entire 600-km Texas coast, yielding accurate surge estimates (root-mean-square errors less than 0.22?m and R ² correlations better than 0.97) with virtually no bias (mean error magnitudes less than 0.03?m).     

Region-specific deterministic and probabilistic seismic hazard analysis of Kanpur city

A seismic hazard map of Kanpur city has been developed considering the region-specific seismotectonic parameters within a 500-km radius by deterministic and probabilistic approaches. The maximum probable earthquake magnitude (M _max) for each seismic source has been estimated by considering the regional rupture characteristics method and has been compared with the maximum magnitude observed \(\left ({M_{\max }^{\text {obs}}}\right )\), \(M_{\max }^{\text {obs}} +0.5\) and Kijko method. The best suitable ground motion prediction equations (GMPE) were selected from 27 applicable GMPEs based on the ‘efficacy test’. Furthermore, different weight factors were assigned to different M _max values and the selected GMPE to calculate the final hazard value. Peak ground acceleration and spectral acceleration at 0.2 and 1 s were estimated and mapped for worst-case scenario and 2 and 10% probability of exceedance for 50 years. Peak ground acceleration (PGA) showed a variation from 0.04 to 0.36 g for DSHA, from 0.02 to 0.32 g and 0.092 to 0.1525 g for 2 and 10% probability in 50 years, respectively. A normalised site-specific design spectrum has been developed considering three vulnerable sources based on deaggregation at the city center and the results are compared with the recent 2011 Sikkim and 2015 Nepal earthquakes, and the Indian seismic code IS 1893.     

Aftershock probabilistic seismic hazard analysis for Bushehr province in Iran using ETAS model

Natural Hazards - Aftershock probabilistic seismic hazard analysis (APSHA) has a key role in risk management after a major earthquake. The main goal of the current study is to assess aftershock...     

Non-Poisson probabilistic seismic hazard assessment

The development of the new seismic hazard map of metropolitan Tehran is based on probabilistic seismic hazard computation using the non-Poisson recurrence time model. For this model, two maps have been prepared to indicate the earthquake hazard of the region in the form of iso-acceleration contour lines. They display the non-Poisson probabilistic estimates of peak ground accelerations over bedrock for 10 and 63 % probability of exceedance in 50 years. To carry out the non-Poisson seismic hazard analysis, appropriate distributions of interoccurrence times of earthquakes were used for the seismotectonic provinces which the study region is located and then the renewal process was applied. In order to calculate the seismic hazard for different return periods in the probabilistic procedure, the study area encompassed by the 49.5–54.5°E longitudes and 34–37°N latitudes was divided into 0.1° intervals generating 1,350 grid points. PGA values for this region are estimated to be 0.30–0.32 and 0.16–0.17 g for 10 and 63 % probability of exceedance, respectively, in 50 years for bedrock condition.     

Site-specific probabilistic seismic hazard analysis for northern part of the Qeshm Island,Iran

This paper presents results of a site-specific probabilistic seismic hazard analysis for northern part of the Qeshm Island, one the most seismic prone areas of Iran. Seismotectonic and seismicity properties of seismic sources in the study area were characterized and used for evaluation of various strong ground motion parameters implementing the classical Cornell’s PSHA approach. The results show that peak rock accelerations for 475-year return period are 0.4 and 0.27 g, respectively, for 84th and 50th percentiles while being about 0.37 and 0.61 g for 2475-year return period. These values are slightly smaller than those read from national seismic zonation maps which can be attributed to the considered conservatism for development of such design maps. In order to incorporate local site conditions, a series of dynamic site response analyses based on the equivalent linear approach were also employed. The results indicate that the presence of soft subsurface deposits at the site significantly alters the fundamental characteristics of the response spectra. The obtained median (50th percentile) peak ground accelerations for 975-year return period range between 0.49 and 0.54 g at different locations in the study site showing minor amplifications relative to their corresponding bedrock acceleration of 0.48 g. Finally, the obtained site-specific spectrum was compared with the standard spectrum mandated by the design codes. In this regard, the agreement was found to be reasonable at period ranges shorter than about 0.5 s, while the differences were more obvious at longer periods. This reveals the need for implementation of site-specific design spectrum to avoid underestimation or overestimation of seismic forces for designing critically important structures especially when softer deposits are encountered.     

Fuzzy probabilistic seismic hazard analysis with applications to Kunming city,China

China is prone to highly frequent earthquakes due to specific geographical location, which could cause significant losses to society and economy. The task of seismic hazard analysis is to estimate the potential level of ground motion parameters that would be produced by future earthquakes. In this paper, a novel method based on fuzzy logic techniques and probabilistic approach is proposed for seismic hazard analysis (FPSHA). In FPSHA, we employ fuzzy sets for quantification of earthquake magnitude and source-to-site distance, and fuzzy inference rules for ground motion attenuation relationships. The membership functions for earthquake magnitude and source-to-site distance are provided based on expert judgments, and the construction of fuzzy rules for peak ground acceleration relationships is also based on expert judgment. This methodology enables to include aleatory and epistemic uncertainty in the process of seismic hazard analysis. The advantage of the proposed method is in its efficiency, reliability, practicability, and precision. A case study is investigated for seismic hazard analysis of Kunming city in Yunnan Province, People’s Republic of China. The results of the proposed fuzzy logic-based model are compared to other models, which confirms the accuracy in predicting the probability of exceeding a certain level of the peak ground acceleration. Further, the results can provide a sound basis for decision making of disaster reduction and prevention in Yunnan province.     

A method for producing digital probabilistic seismic landslide hazard maps

The 1994 Northridge, California, earthquake is the first earthquake for which we have all of the data sets needed to conduct a rigorous regional analysis of seismic slope instability. These data sets include: (1) a comprehensive inventory of triggered landslides, (2) about 200 strong-motion records of the mainshock, (3) 1:24 000-scale geologic mapping of the region, (4) extensive data on engineering properties of geologic units, and (5) high-resolution digital elevation models of the topography. All of these data sets have been digitized and rasterized at 10 m grid spacing using ARC/INFO GIS software on a UNIX computer. Combining these data sets in a dynamic model based on Newmark's permanent-deformation (sliding-block) analysis yields estimates of coseismic landslide displacement in each grid cell from the Northridge earthquake. The modeled displacements are then compared with the digital inventory of landslides triggered by the Northridge earthquake to construct a probability curve relating predicted displacement to probability of failure. This probability function can be applied to predict and map the spatial variability in failure probability in any ground-shaking conditions of interest. We anticipate that this mapping procedure will be used to construct seismic landslide hazard maps that will assist in emergency preparedness planning and in making rational decisions regarding development and construction in areas susceptible to seismic slope failure.     

Seismic hazard analysis of Mersin Province,Turkey using probabilistic and statistical methods

Nearly 108-km lengths of Mersin shores are composed of natural beaches. The region is located between major tourist centers. In the future, this region is thought to be built with a great number of tourist facilities. Turkey’s largest seaport, Ata? refinery (Mersin International Port) is located in Mersin. Recently, Mersin is becoming of great importance to Turkey as the latter plans to construct its second nuclear power plant in the region. Therefore, as nuclear power plants are built to withstand environmental hazards, it is very important to analyze the seismic risk of the areas where the nuclear power plant will be constructed. The region is located between the East Anatolian Fault Zone and Center Anatolian Fault Zone. Based on the Turkey Earthquake Regions Map, Mersin is divided into second-, third-, and fourth-degree earthquake regions. In this study, we sampled earthquakes of magnitude of 4.0 or greater between 01 Jan 1900 and 31 Dec 2010 in the area; seismic hazard of Mersin province was estimated with probabilistic and statistical methods. The study area was selected as the coordinates between 36.03° and 37.42° North and 32.57° and 35.16° East. On the study area, different scaled magnitude values in the last 110 years converted to a common scale (Mw) and earthquake catalog was re-compiled and also seismic sources that may affect the area was determined. In this study, the seismic hazards of the region were obtained using the methods of probability and statistics. This study used three different attenuation relationships. Using the attenuation relationships suggested by Boore et al. (Seismol Res Lett 68(1):128–153, 1997) and Kalkan and Gülkan (Earthquake Spectra 20:1111–1138, 2004), the largest ground acceleration which corresponds to a recurrence period of 475 years was found as 0.08–0.09 g and Akkar and Ça?nan (Bull Seismol Soc Am 100 6:2978–2995, 2010), 0.04 g for bedrock at the central district. When computing for seismic hazard curves, Mut district appears to have a greater seismic hazard compared with other districts. Moreover, according to the attenuation relationships, seismic hazard curves corresponding to a recurrence period of 475 years were obtained for the Mersin Central, Mut, Erdemli, Çaml?yayla, and Tarsus districts.     

Seismic hazard analysis of Sinop province,Turkey using probabilistic and statistical methods

Using 4.0 and greater magnitude earthquakes which occurred between 1 January 1900 and 31 Dec 2008 in the Sinop province of Turkey this study presents a seismic hazard analysis based on the probabilistic and statistical methods. According to the earthquake zonation map, Sinop is divided into first, second, third and fourth-degree earthquake regions. Our study area covered the coordinates between 40.66°– 42.82°N and 32.20°– 36.55°E. The different magnitudes of the earthquakes during the last 108 years recorded on varied scales were converted to a common scale (Mw). The earthquake catalog was then recompiled to evaluate the potential seismic sources in the aforesaid province. Using the attenuation relationships given by Boore et al. (1997) and Kalkan and Gülkan (2004), the largest ground accelerations corresponding to a recurrence period of 475 years are found to be 0.14 g for bedrock at the central district. Comparing the seismic hazard curves, we show the spatial variations of seismic hazard potential in this province, enumerating the recurrence period in the order of 475 years.     

滑坡位移的动态概率预测模型

滑坡位移预测是实现滑坡灾害预报的有效手段,文章利用回声状态网络建立动态预测模型来预测滑坡位移。与传统预测方法相比,动态模型能更好地反映出滑坡演化的动态系统本质;动态模型的建立不需要计算互信息来实现变量选择,也使得预测过程得到简化。考虑到确定性预测的局限性,文章进一步将概率预测的思想引入滑坡位移预测的研究中,提出一种动态预测模型的概率预测方法。所提动态预测模型的预测精度,在白水河和石榴树包两个滑坡位移预测的具体案例中得到了检验;而通过概率预测得到具有明确置信度水平的预测区间,从而也对滑坡状态发展变化趋势给出了更全面的描述。     

Spatial variation of probabilistic seismic hazard for Mumbai and surrounding region

Mumbai city, the economical capital of India, is located on the west coast of stable intra-plate continental region of Peninsular India which has an experience of significant historical earthquakes in the past. The city stood as the fourth most populous city in the world. Recent seismo-tectonic studies of this city highlighted the presence of active West coast fault and Chiplun fault beneath the Deccan basalt. In the present study, spatial variability of probabilistic seismic hazard for Mumbai region (latitudes of 18.85–19.35°N and longitudes of 72.80–73.15°E at a grid spacing of 0.05°) which includes Mumbai city, Suburban, part of Thane district and Navi Mumbai, in terms of ground motion parameters; peak horizontal acceleration and spectral acceleration at 1.0-s period for 2 and 10 % probability of exceedance in 50 years are generated. The epistemic uncertainty in hazard estimation is accounted by employing seven different ground motion prediction equations developed for worldwide shallow crustal intra-plate environments. Further, the seismic hazard results are deaggregated for Mumbai (latitude 18.94°N, longitude 72.84°E) to understand the relative contributions of earthquake sources in terms of magnitude and distance. The generated hazard maps are compared with the zoning specified by Indian seismic code (IS1893: Part 1 in Indian standard criteria for earthquake-resistant design of structures, Part 1—General provisions and buildings. Bureau of Indian Standards, New Delhi, India, 2002) for rocky site. Present results show an underestimation of potential seismic hazard in the entire study region by non-probabilistic zoning prescribed by IS1893: Part 1 with significantly higher seismic hazard values in the southern part of Navi Mumbai.     

A new probabilistic seismic hazard study of Spain

In this paper, we present a probabilistic seismic hazard analysis (PSHA) for mainland Spain that takes into account recent new results in seismicity, seismic zoning, and strong ground attenuation not considered in the latest PSHA of the Spanish Building Code. Those new input data have been obtained as a three-step project carried out in order to improve the existing hazard map for mainland Spain. We have produced a new earthquake catalogue for the area, in which the earthquakes are given in moment magnitude through specific deduced relationships for our territory based on intensity data (Mezcua et al. in Seismol Res Lett 75:75–81, 2004). In addition, we included a new seismogenetic zoning based on the recent partial zoning studies performed by different authors. Finally, as we have developed a new strong ground motion model for the area García Blanco (2009), it was considered in the hazard calculation together with other attenuations gathered from different authors using data compatible with our region. With this new data, a logic tree process is defined to quantify the epistemic uncertainty related to those parts of the process. A sensitivity test has been included in order to analyze the different models of ground motion and seismotectonic zonation used in this work. Finally, after applying a weighting scheme, a mean hazard map for PGA, based on rock type condition for 10% exceedance probability in 50 years, is presented, including 15th and 85th percentile hazard maps. The main differences with the present official building code hazard map are analyzed.     

Comparative study of three probabilistic methods for seismic hazard analysis: case studies of Sochi and Kamchatka

Natural Hazards - This study examines the effect of the procedures used in three different probabilistic seismic hazard analysis (PSHA) methods for estimating the rates of exceedance of ground...     

Flow-based seismic risk assessment of a water transmission network employing probabilistic seismic hazard analysis

Natural Hazards - In this study, a seismic risk assessment model was proposed to evaluate the seismic reliability of a water transmission network. The proposed risk assessment model involves...     

Incorporation of directivity effect in probabilistic seismic hazard analysis and disaggregation of Tabriz city

Near-field ground-motion records affected by directivity may show unusual features in the signal resulting in low-frequency cycle pulses in the velocity time history. Current probabilistic seismic hazard analysis (PSHA) is not able to predict such effects well; recent studies thus have proposed modified frameworks to incorporate pulse effect in modified PSHA. This paper attempts to carry out the seismic hazard mapping of Tabriz city according to modified and ordinary PSHA for different return periods. Tabriz, located in northwest of Iran, is situated in the vicinity of the North Tabriz Fault, which is one of the major seismogenic faults. Disaggregation results indicate that including pulse-like effects in PSHA, increases the relative contribution of close distances and small epsilons (?). Another major probable result is the high contribution of pulse periods close to spectral period. The contributions to each earthquake scenario at long-period spectral acceleration shift to larger magnitudes with including the pulse effects.