Numerous Tibetan lower-crustal and upper-mantle earthquakes, detected by Sn/Lg ratios, suggest crustal delamination or drip tectonics

Abstract

Whether intermediate-depth earthquakes beneath Tibet and the Himalaya are in the lower crust or upper mantle should provide insight into rheology, hence composition and temperature, of continental lower crust and upper mantle. Based on the waveguide theory that earthquakes respectively above or below the Moho will excite higher Lg or Sn energy, we develop an Sn/Lg amplitude-ratio analysis using a single permanent station to investigate earthquake depths with respect to the Moho. First, we use synthetics to show the ubiquitous sharp increase of the Sn/Lg ratio for hypocenters beneath the crust. A deep-crustal high-vs layer appropriate for eclogitized Indian lower crust causes a more gradual increase of Sn/Lg ratios for hypocenters within or below the eclogite layer. We measured Sn/Lg ratios of 595 Tibetan earthquakes from 1998 to 2022 with nominal (catalog) hypocentral depths >30 km and magnitudes >3.2 in five regions (west Tibet, southwest Tibet, south Tibet, southeast Tibet, and Qiangtang). As predicted by our normal-mode synthetics, earthquakes in west and south Tibet show Sn/Lg ratios that increase sharply as catalog depths approach the independently determined receiver-function Moho. Numerous earthquakes with high Sn/Lg ratios and nominal depths around the reported Moho are identified between the Karakoram fault and Altyn-Tagh fault (ATF) in west Tibet, which we attribute to an eclogitized Indian lower crust extending north to the ATF. Impingement of this underthrust Indian slab against the Tarim craton and consequent eclogite delamination or dripping may cause the observed below-Moho seismicity which occupies a region 100 km across and 200 km along orogenic strike and extends 30–50 km below published Moho depths. Similarly dense seismicity spans the Moho and occupies a similar volume below the Moho beneath the High Himalaya in south Tibet. The south Tibet Sn/Lg patterns may indicate an eclogitized Indian layer beginning to delaminate or drip south of the Yarlung-Zangpo suture. Our southeast Tibet and southwest Tibet Sn/Lg observations are less clear cut, perhaps due to less appropriate epicentral distances from the available observing stations. The Qiangtang region of Tibet likely has sparse mid-to-lower-crustal earthquakes, but no definitive below-Moho earthquakes. Our work expands the catalog of continental intraplate earthquakes below the Moho by >100 events with mb>3.2, as well as identifying tens of mb>3.2 lower-crustal events, so that Tibet fits neither an ideal crême-brulée nor an ideal jelly-sandwich crustal-strength paradigm. Two clusters of earthquakes that each span the Moho may be best explained by the delamination or dripping of a strong eclogitized lower crust into the upper mantle.