RAPID/Collaborative Research: Investigation of False Positive Liquefaction Triggering Predictions from the Canterbury Earthquake Sequence

2010年至2011年，新西兰克赖斯特彻奇市遭受了一系列强烈地震的破坏。这些事件在近三分之一的城市地区造成了严重的土壤液化破坏，有些地方液化了六次或更多。最终，土壤液化造成的损失约占地震直接经济损失的50%（即400亿美元总额中的约200亿美元）。在地震之后，一个前所未有的、开放的地下数据库，称为坎特伯雷岩土技术数据库（CGD），已经被编制出来，以确定城市的哪些地区容易（或不容易）受到土壤液化的影响。CGD目前包含了来自大克赖斯特彻奇地区的大约18000个圆锥体穿透测试（CPT）的探测结果，可以对每次地震中观测到的土地性能（即液化引起的破坏或缺乏）和基于CPT的液化触发的回顾性预测进行详细的空间比较。在使用这个广泛的CPT数据库来估计整个城市的液化触发/易感性之后，得出了以下观察结果：(1)CPT已被证明是评估液化触发/表现潜力的一个很好的筛选工具。除了相对较少的例外，如果观测到一个地点在其中一次地震中液化，基于cpt的方法是回顾性预测液化的触发，并且(2)基于cpt的方法倾向于预测整个基督城一些相当大的地区的液化触发/易感性，这些地区没有观测到液化引起的损害。因此，基于cpt的方法产生了相当多的假阳性回顾性预测。虽然对液化严重程度的明显过度预测给克赖斯特彻奇市民在未来的建筑努力方面带来了独特的挑战，但这个问题可能不是克赖斯特彻奇土壤所独有的，因此可能会在美国和世界各地的类似土壤中推动过于保守（即过于昂贵）的液化设计。因此，这是一个具有国际影响的工程问题。为了确定是否可以在当前基于cpt的程序中实施合理的调整以最大限度地减少误报，需要迅速进行进一步的研究。这个快速反应研究（RAPID）合作项目也将有助于加强美国和新西兰之间的国际研究合作，并将为美国研究生提供有益的国际旅行经验，这将有助于平衡他们的技术教育，并使他们接触到地震工程中仍然存在的全球联系问题。简而言之，这项工作的广泛影响远远超出了新西兰的边界，并将影响美国和国外的地震危险措施。在新西兰基督城通过直推井间测试收集的一些假阳性地点的初步剪切波速（Vs）和压缩波速（Vp）数据显示，地下水位以下的土壤顶部几米通常具有高于预期的v值和/或低于预期的Vp值（表明非饱和状态）。这些初步发现，如果在大范围内得到证实，可以解释假阳性液化预测。因此，本研究将利用组合CPT-Vs/Vp方法来调和液化触发预测与负面土地破坏观测。为了研究这个问题，我们计划在大约30个额外的地点进行以下测试：(1)直接推入井间，(2)地震CPT，(3)通过声波钻井进行连续采样。我们相信这项工作将导致高质量病例历史的发展，可用于改进当前基于cpt的液化触发程序，以尽量减少过于昂贵的假阳性预测。虽然这项研究目前只能在新西兰进行，但由于他们不幸的环境，最终目标是通过提高抗震设计方面的民用基础设施的弹性、可持续性和可负担性，使整个社会受益。

In 2010-2011, the city of Christchurch, New Zealand was devastated by a series of powerful earthquakes. These events induced severe soil liquefaction damage in nearly one third of the city area, with some sites liquefying six or more times. Ultimately, soil liquefaction damage was blamed for approximately 50% of the direct economic losses from the earthquakes (i.e., approximately $20B of the $40B total). In the aftermath of the earthquakes, an unprecedented, open-access subsurface database, called the Canterbury Geotechnical Database (CGD), has been compiled in order to determine which areas of the city are susceptible (or not susceptible) to soil liquefaction. The CGD currently contains about 18,000 cone penetration test (CPT) soundings from the greater Christchurch area, which allow for detailed spatial comparisons between observed land performances in each earthquake (i.e., liquefaction-induced damage or a lack thereof) and CPT-based retrospective-predictions of liquefaction triggering. After using this extensive CPT database to estimate liquefaction triggering/susceptibility across the city, the following observations have been made: (1) The CPT has proven to be a great screening tool for assessing the potential for liquefaction triggering/manifestations. With relatively few exceptions, if a site was observed to have liquefied in one of the earthquakes the CPT-based methods are retrospectively-predicting triggering of liquefaction, and (2) The CPT-based methods tend to predict liquefaction triggering/susceptibility in some fairly large areas throughout Christchurch where liquefaction-induced damage was not observed. Therefore, the CPT-based methods are producing quite a few false positive retrospective-predictions. While the apparent over-prediction of liquefaction severity poses unique challenges to the citizens of Christchurch regarding future building efforts, this problem is likely not unique to Christchurch soils, and therefore may be driving overly-conservative (i.e., overly-expensive) liquefaction design in similar soils in the United States and world-wide. As such, it is an engineering problem with international consequences. Further studies are rapidly needed in order to determine if a rational adjustment can be implemented in current CPT-based procedures to minimize false positives. This Rapid-Response Research (RAPID) collaborative project will also serve to strengthen international research collaborations between the U.S. and New Zealand, and will provide U.S. graduate students with rewarding international travel experiences that will serve to balance their technical education and expose them to the globally-connected problems that still exist in earthquake engineering. In short, the broader impacts of this work stretch far beyond the borders of New Zealand, and will impact seismic hazard practices in the U.S. and abroad.Preliminary shear wave velocity (Vs) and compression wave velocity (Vp) data collected at a number of false positive sites in Christchurch, New Zealand via direct-push crosshole testing has revealed that the top several meters of soil generally has higher than expected Vs values and/or lower than expected Vp values below the ground water table (indicating unsaturated conditions). These preliminary findings, if corroborated on a wide-scale, could explain the false positive liquefaction predictions. Therefore, this research will utilize a combined CPT-Vs/Vp approach for reconciling positive liquefaction triggering predictions with negative land damage observations. We plan to conduct the following tests at approximately 30 additional sites in order to investigate this issue: (1) direct-push crosshole, (2) seismic CPT, and (3) continuous sampling via Sonic drilling. We believe this work will lead to the development of high-quality case histories that can be used to refine current CPT-based liquefaction triggering procedures in order to minimize overly-expensive false positive predictions. While this research can presently only be conducted in New Zealand, due to their unfortunate set of circumstances, the end-goal is to benefit society at-large through increased resiliency, sustainability, and affordability of our civil infrastructure in regards to seismic design considerations.