Imaging deep mantle structure beneath Alaska using full waveform tomography

地球表面的板块在不断移动，并在板块边界被消耗到地球上。在北美，大洋太平洋板块在阿拉斯加南海岸下俯冲，在地球表面产生强烈的地震和活火山。那么这个俯冲的板块在地球内部去往何处呢？追踪这块板块进入地球的轨迹，可以让我们了解数百万年来地球表面板块运动的近代史，以及地球自形成以来数十亿年来的演变过程。要了解现在，我们必须研究过去。就像超声波被用来观察人体内的器官一样，地震学家利用地震产生的地震波来研究地球的内部运作。图像的清晰度或模糊度是由所使用的波的音调（或频率）控制的。更高的频率提高了焦点，但需要在整个地球上进行大量的计算。假设可以用来简化计算，将繁重的计算减少到我们想要成像的地球部分。这种简化意味着我们可以提高我们的注意力，更好地跟踪阿拉斯加俯冲的历史。该研究将支持研究生的培训，并为早期职业研究者提供支持。pi将与社区共享代码和模型，并将参与湾区的当地推广活动。阿拉斯加在过去200 Ma的演化以俯冲、碰撞和增生为特征。这一漫长的俯冲历史的残余应该存在于最底部的地幔中，但过去的区域和全球层析模型解决了不一致的结构，可能是由于方法的局限性和有限的采样。我们对北太平洋板块构造史的了解目前还不完整。剩下的问题包括：板块在阿拉斯加地下渗透有多深，板块的几何形状和厚度是多少，以及它如何与过渡带相互作用？该提案的主要目标是利用在有限区域内应用的全波形反演方法（称为“盒”层析成像）来提高阿拉斯加下方整个地幔区域地震图像的分辨率。全波形反演需要考虑多路径和波前愈合的影响，否则会掩盖强烈的局部非均质性，如板块和周围的地幔楔。此外，空间受限的“盒子”方法结合了一个快速的1D和较慢的3D波场求解器，从而减少了计算时间，这使得团队能够使用更高频率的区域和远震体波。首先利用三分量面波、泛音和体波的组合，构建剪切速度模型。随着迭代的进行，增加计算的最大频率，并增加对纵速敏感的体波形，将获得纵速图像。这种分析将大大提高现有的地震波速度和径向各向异性图像，特别是在过渡带深度和中、下地幔。这些高分辨率的阿拉斯加地幔层析成像将有助于(1)限制该地区俯冲和地幔动力学的历史，(2)计算从南桑威奇群岛到阿拉斯加监测站的极地路径上观测到的更精确的地幔相位校正，这些路径呈现出特别大的走时异常，至少部分可能是由于阿拉斯加的板块结构。通过更好地识别俯冲板块，这项工作的结果将使地球动力学家对俯冲带下地幔的流动建模和板块运动重建感兴趣。地球动力学家和矿物物理学家也会对研究内核各向异性的模式和起源感兴趣。此外，一个强大的、高分辨率的阿拉斯加层析成像模型和对其影响进行校正的方法将提高USArray在阿拉斯加站在其他深层地球研究中的效用。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。

The Earth’s surface plates are constantly moving and are being consumed into the Earth at plate boundaries. In North America, the oceanic Pacific plate is subducting beneath the south coast of Alaska, generating powerful earthquakes and active volcanoes on Earth’s surface. But where inside the Earth does this subducted plate then go? Tracking this plate into the Earth tells us about the more recent history of plate movement on the surface over millions of years, as well as how the Earth has evolved since it formed over billions of years. To understand the present, we must investigate the past. Much like ultrasound waves are used to look at organs inside a body, seismologists use seismic waves generated by earthquakes to investigate the inner workings of the Earth. The sharpness or blurriness of the images is controlled, among others, by the pitch (or frequency) of the waves that are used. Higher frequencies improve the focus but require heavy computations in the whole Earth. Assumptions can be used to simplify the calculations, reducing the heavy computations to the part of the Earth that we want to image. This simplification means that we can improve our focus and better track the history of subduction under Alaska. The study will support the training of a graduate student and provide support for an early-career investigator. The PIs will share codes and models with the community, and will be involved in local outreach in the Bay Area.The evolution of Alaska over the past 200 Ma features multiple episodes of subduction, collision and accretion. The remnants of this long subduction history should be present down to the lowermost mantle, but past regional and global tomographic models resolve inconsistent structures, likely owing to methodological limitations and limited sampling. Our understanding of the plate tectonics history of the Northern Pacific is currently incomplete. Remaining questions include: how deep do slabs penetrate beneath Alaska, what is the slab geometry and thickness, and how does it interact with the transition zone? The primary objective of this proposal is to improve the resolution of whole mantle regional seismic images beneath Alaska using a Full Waveform Inversion method applied within a restricted region, referred to as “box” tomography. Full Waveform Inversion is required to account for the effects of multipathing and wavefront healing that otherwise mask strong and local heterogeneity, such as slabs and surrounding mantle wedges. Moreover, the spatially restricted “box” approach couples a fast 1D and slower 3D wavefield solver thus reducing computation time, which enables the team to use higher frequency regional and teleseismic body waves. Using a combination of 3-component surface wave, overtone and body waveforms, a shear velocity model will first be constructed. Increasing the maximum frequency of the computations as iterations progress, and with additional body waveforms sensitive to compressional velocity, compressional velocity images will be obtained. This analysis will significantly sharpen existing images of seismic wavespeeds and radial anisotropy, particularly at transition zone depths and the mid and lower mantle. These higher resolution tomographic images of the mantle beneath Alaska will help to (1) constrain the history of subduction and mantle dynamics in this region, and (2) compute more accurate mantle corrections for core phases observed on polar paths from the south Sandwich Islands to stations in Alaska, which present a particularly large spread of travel time anomalies, at least part of which is likely due to Alaska slab structure. The results of the work will be of interest to geodynamicists for modeling flow in the mantle beneath subduction zones, and in plate motion reconstructions, by allowing better identification of subducted slabs. It will also be of interest to geodynamicists and mineral physicists investigating the pattern and origin of inner core anisotropy. Moreover, a robust, high resolution tomographic model of Alaska and a method for providing corrections for its effects will improve the utility of the USArray stations in Alaska for other deep Earth studies.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.