In the U.S. Pacific Northwest (PNW), the historic earthquake record is often insufficient to provide inputs to seismic-hazard analyses or to inform ground-motion predictions for certain seismic sources (e.g., the Cascadia Subduction Zone, CSZ). As a result, paleoseismic studies are commonly used to infer information about the seismic hazard. However, among the many forms of coseismic evidence, soil liquefaction provides the best, if not only, evidence from which the intensities of previous ground motions may be constrained. Accordingly, the overarching goal of this research is to use paleoliquefaction to elucidate previous ground motions in the PNW – both for CSZ events and others – and to further constrain the locations, magnitudes, and recurrence rates of such ruptures. Towards that goal, this paper: (i) reviews current paleoliquefaction inverse-analysis methods and their limited, prior applications in the PNW; (ii) compiles all PNW paleoliquefaction evidence from the literature into a GIS database, resulting in data from 185 study sites (e.g., feature locations, types, sizes, and ages); and (iii) develops maps – specific to the CSZ – that forecast paleoliquefaction for 30 different simulations of a CSZ event. These maps can be used to guide field explorations for new evidence, such that they are conducted efficiently and strategically, considering the apparent utility of evidence towards constraint of CSZ ground-motion models. Of additional utility, this process provides regional ground-motion predictions for physics-based simulations of an M9 event, to include expected site effects. Collectively, the maps of expected shaking intensity and liquefaction may be useful in downstream hazard modeling, regional loss estimation, policy development, and science communication. Ultimately, as more paleoliquefaction evidence is identified and studied, better constraint of regional ground-motion hazards will result.

在美国太平洋西北部 (PNW)，历史地震记录通常不足以为地震危害分析提供输入或为某些地震源（例如，卡斯卡迪亚俯冲带，CSZ）的地面运动预测提供信息。因此，古地震研究通常用于推断有关地震危险的信息。然而，在多种形式的同震证据中，土壤液化提供了最好的证据，如果不是唯一的证据，先前地面运动的强度可能会受到限制。因此，这项研究的首要目标是利用古液化来阐明 PNW 先前的地面运动——包括 CSZ 事件和其他事件——并进一步限制此类破裂的位置、幅度和复发率。为实现这一目标，本文：(i) 审查当前的古液化逆分析方法及其在 PNW 中有限的先前应用；(ii) 将文献中的所有 PNW 古液化证据编译到 GIS 数据库中，从而生成来自 185 个研究地点的数据（例如，特征位置、类型、大小和年龄）；(iii) 开发地图 - 特定于 CSZ - 预测 CSZ 事件的 30 种不同模拟的古液化。考虑到证据对 CSZ 地面运动模型约束的明显效用，这些地图可用于指导新证据的实地探索，以便有效和有策略地进行。此外，该过程还为 M9 事件的基于物理的模拟提供了区域地面运动预测，以包括预期的场地效应。集体，预期震动强度和液化的地图可能对下游灾害建模、区域损失估计、政策制定和科学传播有用。最终，随着更多古液化证据的识别和研究，将产生更好的区域地面运动灾害约束。