Submarine landslides pose a hazard to coastal communities as they can generate powerful tsunamis, and threaten critical offshore infrastructure such as seafloor cable networks that underpin global communications. Such events can be orders of magnitude larger than their onshore equivalents. Despite the hazard they pose, many aspects of submarine landslides remain poorly understood, such as why they fail on low angle (<2°), seemingly stable slopes. Many studies have proposed that failure on low slope angles, and the large areal extent of submarine landslides, may be controlled by the presence of laterally-extensive weak layers embedded within the slope stratigraphy, which precondition slopes to failure. Little remains known, however, about the characteristics and processes that control and form weak layers. We conducted a comprehensive review of published submarine landslide studies that examine failure planes and apparent weak layers associated with historical and ancient submarine landslides. Based on a new global landslide catalogue that comprises 64 case studies, this review aims to investigate the types of sediment that form weak layers and to understand the controls on their global variability. Existing classification schemes are based on mechanical process(es), and do not readily enable a diagnosis of weak layers from unfailed sediments. Here, a new and complementary classification of weak layers based on lithology is introduced. This classification enables weak layer recognition from sediment cores (including those sampling unfailed sediments), and allows us to attribute failure mechanisms to different environmental settings where distinct types of weak layers are more likely. The results show that failure planes usually form in the vicinity of an interface between distinct lithologies that together comprise a weak layer. The weak layers of 22 of the 64 case studies were related to characteristic sediment sequences within the slope stratigraphy, of which 19 were classified based on direct measurements from sediment cores and in-situ measurements: 16 weak layers were classified as siliciclastic, four as volcaniclastic, and two as fossiliferous sediment sequences. Only three submarine landsides were related to clay-dominated weak layers. In addition, failure along lithological contrasts was inferred for six case studies. Based on global depositional models likely locations of these different types of weak layer can be inferred. These include oceanic gateways where long-term circulation can create distinct permeability interfaces within siliciclastic sequences, areas of high productivity where biogenic sediments may dominate, and regions that experience widespread ash fall from volcanic eruptions. We highlight that many submarine landslide studies have historically not collected sediment cores that characterise weak layers within intact sedimentary sequences and instead have focused on characterising the slope failure deposit. As weak layers can collapse or become heavily modified during failure, there is a widespread omission of key information required for geotechnical analysis to determine where and why certain slopes are predisposed to failure. We conclude by highlighting the need to combine detailed geotechnical measurements with sedimentological and geophysical analyses including grain-scale observations (e.g. micro-Computed Tomography 3D imagery), and emphasise the importance of a uniform workflow that will allow for a better comparison between individual studies.

海底滑坡对沿海社区构成威胁，因为它们会产生强大的海啸，并威胁到支撑全球通信的海底电缆网络等关键的海上基础设施。此类事件可能比陆上事件大几个数量级。尽管它们构成了危险，但海底滑坡的许多方面仍然知之甚少，例如它们为什么会在低角度（<2°）、看似稳定的斜坡上失败。许多研究表明，低坡角和大面积海底滑坡的破坏可能受到斜坡地层中横向扩展弱层的控制，这是斜坡破坏的先决条件。然而，关于控制和形成弱层的特征和过程知之甚少。我们对已发表的海底滑坡研究进行了全面审查，这些研究检查了与历史和古代海底滑坡相关的失效平面和明显的薄弱层。基于包含 64 个案例研究的新全球滑坡目录，本综述旨在调查形成薄弱层的沉积物类型，并了解对其全球变化的控制。现有的分类方案基于机械过程，并且不容易从未破坏的沉积物中诊断薄弱层。在这里，介绍了一种基于岩性的新的、互补的弱层分类。这种分类可以从沉积物岩心（包括对未失效沉积物取样的那些）中识别弱层，并允许我们将故障机制归因于更可能出现不同类型的弱层的不同环境设置。结果表明，破坏面通常形成在不同岩性之间的界面附近，这些岩性共同构成了一个薄弱层。64 个案例研究中的 22 个弱层与斜坡地层内的特征沉积层序有关，其中 19 个是根据沉积岩心的直接测量和原位测量进行分类的：16 个弱层被归类为硅质碎屑岩，4 个被归类为火山碎屑岩，和两个作为化石沉积序列。只有三个海底陆侧与以粘土为主的薄弱层有关。此外，六个案例研究推断出沿岩性对比的失败。根据全球沉积模型，可以推断出这些不同类型弱层的可能位置。其中包括长期环流可以在硅质碎屑岩层序内形成明显渗透界面的海洋门户、生物沉积物可能占主导地位的高生产力区域以及火山喷发造成广泛火山灰坠落的区域。我们强调，许多海底滑坡研究历来没有收集表征完整沉积序列内弱层的沉积物核心，而是专注于表征斜坡崩塌沉积物。由于薄弱层可能会在破坏过程中坍塌或发生严重变化，因此岩土分析所需的关键信息被广泛遗漏，以确定某些斜坡易发生破坏的位置和原因。