Consistently computed ground motion intensity measures for liquefaction triggering assessment

The state of practice for assessing soil liquefaction relies on manifestation models developed from databases of liquefaction field case histories. A key component of these case histories is the seismic demand that led to soil liquefaction triggering or lack thereof. Traditionally, this seismic demand is represented by the peak ground acceleration (PGA), supplemented with the earthquake magnitude (M). The PGA values have generally been estimated using various methods. For instance, some PGAs were considered equal to the recorded PGA at a nearby seismic station, or estimated using ground motion models, 1D site response analyses, or judgment. This data release consists of PGA values at case history sites in the Next Generation Liquefaction (NGL) database, computed using a consistent approach based on Kriging interpolation and spatial correlation models developed using Bayesian inference. Values for three additional ground motion intensity measures (IMs) are also shared: peak ground velocity (PGV), cumulative absolute velocity (CAV), and Arias Intensity (AI). All the data associated with the calculation of these ground motion IMs are made available, including (1) recorded ground motion IMs for 31 major earthquake events, (2) coefficients and plots for a thousand spatial correlation models for each ground motion IM, and a maximum a posteriori model; and (3) gmKriger, a user-friendly Jupyter Notebook is included to allow users to compute ground motion intensity measures for events and locations of their interest. This dataset aims at making these data available to a broader audience of researchers.

当前评估土壤液化（Soil Liquefaction）的行业实践现状，依赖于基于液化现场案例历史数据库开发的液化表征模型。此类案例历史的核心要素，为引发土壤液化触发与否的地震动需求（Seismic Demand）。传统上，该地震动需求以峰值地面加速度（Peak Ground Acceleration, PGA）表征，并辅以地震震级（Earthquake Magnitude, M）。PGA值通常通过多种方法估算：部分PGA值被视作邻近地震台站记录的PGA，或通过地震动模型（Ground Motion Models）、一维场地响应分析（1D Site Response Analyses）或经验判断进行估算。本次发布的数据包含下一代液化数据库（Next Generation Liquefaction, NGL）中案例历史场地的PGA值，这些值通过基于克里金插值（Kriging Interpolation）与贝叶斯推断（Bayesian Inference）构建的空间相关模型的统一方法计算得到。本次发布还包含另外三类地震动强度指标（Ground Motion Intensity Measures, IMs）的数值：峰值地面速度（Peak Ground Velocity, PGV）、累积绝对速度（Cumulative Absolute Velocity, CAV）以及阿里亚斯强度（Arias Intensity, AI）。所有与这些地震动强度指标计算相关的数据均已公开，包括：(1) 31次重大地震事件的记录地震动强度指标；(2) 每个地震动强度指标对应的1000个空间相关模型的系数与绘图结果，以及最大后验模型（Maximum A Posteriori Model）；(3) 易用的Jupyter Notebook工具gmKriger，可供用户为自身关注的地震事件与场地计算地震动强度指标。本数据集旨在将这些数据开放给更广泛的研究人员群体。