Deep earthquakes occur down to 700 km depth where pressure is up to two orders of magnitude greater than in the crust. Rupture characteristics and propagation mechanisms under those extreme conditions are still poorly constrained. We invert seismic waveforms for the spatial dimensions, duration and stress drop of deep‐focus earthquakes (Mw 6.7–7.7) in the Kuril subduction zone. We find stress drops of ∼1–10 MPa and rigidity‐corrected spatial dimensions and durations similar to crustal earthquakes. Radiated efficiency >1 is observed, suggesting that undershooting is prevalent in deep earthquakes, consistent with laboratory derived weakening mechanisms. Comparisons with subduction models suggest across‐slab propagation within regions with temperatures T < ∼ 1,000°C, similar to shallow events. Hence, despite different triggering mechanisms, the same physics seems to control the rupture propagation of both shallow and deep earthquakes.

No abstract

During the Apollo missions in the 1970s, four seismic stations were placed on the Moon's near side. These seismic stations recorded hundreds of deep moonquakes, which occur at depths of 700-1,200 km.Deep moonquakes tend to occur in clusters called "nests." In each nest, moonquakes occur at monthly intervals; they are seemingly influenced by the Earth-induced tidal stresses (

&lt;p&gt;Repeating earthquakes are thought to represent the repeated rupture of loaded patches surrounded by regions that are slipping aseismically; they provide a natural laboratory to study interactions between seismic and aseismic processes. These events occur less often than one would expect if these earthquakes accommodate all of the long-term slip. Recent crack models using rate-and-state friction&lt;strong&gt; &lt;/strong&gt;(Cattania and Segall, 2019; Chen and Lapusta, 2009 ) suggest a possible explanation: for small events, a larger amount of the slip budget on the patch being taken up by aseismic slip. For larger events where most of the slip budget is seismic, the patch experiences partial ruptures, also leading to the deviation from expected scaling. We aim to test the predictions of this model of repeating ruptures by searching for the proposed partial ruptures. We choose to search using the Northern California earthquakes catalogue, which contains many well-located repeating earthquake sequences. Preliminary results suggest that partial ruptures in the Parkfield region are not common. If preliminary results pass additional tests, it may suggest that partial ruptures do not make up a significant proportion of the slip budget of larger repeating earthquakes in this region.&amp;#160;&lt;/p&gt;

We probe the present‐day stresses in the lunar interior by examining the slip directions of moonquakes in the A01 nest. In this nest, some deep moonquakes appear to slip “backwards,” in the opposite direction to other events. We assess whether these changes in slip direction result from a spatial variation in the tectonic stress or from a temporal variation in the tidal stress. To test these two options, we first show that a dominant tectonic stress implies deep moonquakes can only slip in one direction: forwards and backwards, while a dominant tidal stress could allow moonquakes to slip in more directions: any combination of forwards, backwards, left, and right. Then we look for the number of slip directions; we separate the deep moonquake waveforms into slip directions using a principal component analysis technique. We find two slip directions present in the A01 deep moonquake nest. The moonquakes slip in a variety of directions as time evolves. This observation implies that the tidal stresses drive deep moonquakes. Additionally, these results place a new constraint on the magnitude of the tectonic stresses at depth; they must be smaller than the modeled tidal stress of ∼0.1 MPa.