Global Search for D" Discontinuity Structure

在地球表面下大约1600英里处，有一个与地球性质形成鲜明对比的地方，被称为“D”不连续区，已经在100多项研究中被证明存在。然而，地球科学家对这种不连续性的物理含义仍然知之甚少。这主要有两个原因：用于检测这种不连续性的大多数地震传感器在空间上仅限于陆地上的传感器（我们的地球只有30%被陆地覆盖），我们目前的技术只能让我们从地震的最深处检测这种不连续性；地震记录中的额外噪音先前阻碍了我们使用浅层地震的能力。然而，浅层地震比深层地震数量更多，发生在地球上更多的地方，因此，如果浅层地震可以被利用，它们将大大提高地球的覆盖范围，我们可以在其中寻找不连续。因此，该项目寻求开发新技术，使地球科学家能够利用浅层地震感知与这种不连续有关的微妙地震信号。此外，该项目将开发一种新的建模能力，以便更好地理解构成这种不连续性的材料，并具有更高的准确性。能够确定D ‘ ’不连续代表的位置和内容至关重要，因为它可能在地球内部正在进行的过程中起着重要作用。例如，根据D ‘ ’不连续的位置和性质，可能会强烈影响整个地幔对流过程，并控制我们从夏威夷、冰岛或黄石等热点火山观测到的火山活动。这项研究将成为犹他大学博士后学者的研究重点。该项目将进一步资助一名学生2年的本科研究经历，并将额外支持一名学生的研究经历，作为犹他大学中学教师硕士项目的一部分。本研究中进行的广泛数据分析的所有数据收集和结果将在犹他大学的hive数据存储库上公开共享，并且为执行本提案中描述的阵列引导技术而开发的所有软件将通过GitHub免费提供。地震阵列处理技术旨在增强低振幅地震波到达，非常适合搜索横向可变、低对比度的地震不连续面。然而，只有少数研究利用了这些阵列处理方法提供的大量地震数据。此外，当使用标准阵列处理方法时，人们经常观察到潜在的到达，这些到达可能与已知或未知的不连续结构有关，但也可能是由于接收机子集上的相关噪声条件。在这里，研究人员提出了一种新的阵列处理方法的发展和应用。特别是，他们将地震台站分组为虚拟子阵列，并使用自举重采样方法计算每个子阵列的速度地震图（vespagrams），最终为每个子阵列获得多个vespagrams。对于每个子阵列，研究人员自动识别(1)地震波到达，(2)传播时间和慢度的95%置信限，以及(3)估计到达的可能性不是由噪声引起的。他们进一步证明，他们可以使用这种方法对浅层地震成像，这有可能大大增加下地幔的数据覆盖范围。研究人员进一步开发了一个贝叶斯反演方案来模拟与他们的观测相关的地震速度剖面，从而进一步估计与观测相关的误差、不确定性和权衡。通过结合许多子阵列的结果，他们进一步发展了不连续结构的置信度约束。通过这种方法，研究人员希望在地幔最底部1000公里处识别、编录和绘制横向可变但一致的sh波地震不连续面，并根据误差估计和正演模拟提供这些不连续面存在的深度范围和地震速度结构的约束。该奖项反映了美国国家科学基金会的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。

Approximately 1600 miles beneath the surface of our planet a sharp contrast in Earth properties, called the D" discontinuity, has been shown to exist in more than 100 studies. However, geoscientists still don’t have great knowledge of what this discontinuity physically represents. This is for two primary reasons: most seismic sensors used to detect this discontinuity are spatially restricted to those on land (only 30% of our planet is covered by land) and our current technology only allows us to detect this discontinuity from the deepest of earthquakes; extra noise in the seismic recordings have previously hindered our ability to use shallower earthquakes. Shallow earthquakes, however, are more numerous and occur in many more places on the Earth than deep earthquakes, so if shallow earthquakes could be utilized, they would greatly improve the coverage of the Earth in which we can search for the discontinuity. As such, this project seeks to develop new technology that allows geoscientists to sense the subtle seismic signals associated with this discontinuity using shallow earthquakes. In addition, the project will develop a new modeling capability that will allow a better understanding of what the materials making up this discontinuity are with greater accuracy. Being able to determine where and what the D” discontinuity represents is critical because it likely plays a large role in the ongoing processes inside the Earth. For example, depending on where and what the D" discontinuity is, could strongly affect the whole mantle convection process and control where we observe volcanic activity from hot spot volcanoes such as observed in Hawaii, Iceland, or Yellowstone. This research will be the focus of research for postdoctoral scholar at the University of Utah. This project will furthermore fund 2 years of undergraduate research experience for one student and will additionally support a research experience for one student as a part of the MS for Secondary School Teachers program at the University of Utah. All data collections and results from the extensive data analyses performed in this study will be shared openly on the University of Utah’s hive data repository and all software developed to carry out the array bootstrap technique described in this proposal will be made freely available through GitHub.Seismic array processing techniques are designed to enhance low amplitude seismic wave arrivals and are ideally suited for searching for laterally variable, low contrast seismic discontinuities. Yet only a limited number of studies have taken advantage of the vast amounts of seismic data available using these array processing methods. In addition, when using standard array processing approaches, one often observes potential arrivals that could be associated with known or unknown discontinuity structure but may also be due to correlated noise conditions on a subset of receivers. Here the investigators propose the development and application of a new array processing methodology. In particular, they group seismic stations into virtual subarrays, and compute velocity seismograms (vespagrams) for each subarray using a bootstrap resampling approach ultimately attaining multiple vespagrams for each subarray. For each subarray, the investigators automatically identify (1) seismic wave arrivals, (2) 95% confidence limits in travel-time and slowness, and (3) estimates on the likelihood that the arrival is not due to noise. They further demonstrate that they can image Dʺ discontinuity seismic arrivals using this methodology for shallow earthquakes which has the potential to increase data coverage of the lower mantle substantially. The investigators further develop a Bayesian inversion scheme to model the seismic velocity profile associated with their observations which provides further estimates of the errors, uncertainties and trade-offs associated with the observations. By combining the results from many subarrays, they further develop confidence constraints on discontinuity structures. With this approach the investigators hope to identify, catalog and map laterally variable yet consistent SH-wave seismic discontinuities in the bottommost 1,000 km of the mantle and to provide constraints on the depth range and seismic velocity structure at which these discontinuities exist based on error estimates and forward modeling.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.