Velocity structure and seismicity of the Garm region, Central Asia

Abstract

S U M M A R YWe have examined arrival-time data from two seismic networks operating within the Pamir-Tien Shan region of Central Asia in order to evaluate crustal structure and earthquake distribution in one of the most seismically active zones in the region. The data are from two networks surrounding the Soviet earthquake-prediction laboratory at Garm, Tadjikistan: a 15-station local network operated by the Soviet Complex Seismological Expedition (CSE) since the mid-l950s, and a dense telemetered network, nested within the larger CSE network, operated jointly by the US Geological Survey (USGS) and the CSE since 1975. A simultaneous inversion procedure was used to calculate l-D (layered) velocity models and station corrections independently for each of the two networks. Using a variety of data subsets and starting models, we obtained stable estimates of the velocity structure, independent of the initial velocity models. For both the CSE and the USGS network data the velocities of two upper-crustal layers proved to be indistinguishable from a model with a single layer over a half-space. For the CSE network, the best P-wave velocity model comprises an 8 km thick upper crustal layer of 5.2 km sK', overlying a 5.9 km s K ' half-space. Using data from the USGS network alone, a similar velocity model was obtained, with the exception of a significantly higher velocity (V, = 6.3 km sK ') for the mid-crustal half-space. These results contradict previously published velocity models featuring anomalously low or high seismic velocities in the upper to mid-crust. Hypocentral parameters were redetermined for approximately 26 000 events occurring during the period 1975-82 using the newly obtained velocity models and station corrections. Most of the seismicity in the study area is associated with active deformation of the sedimentary rocks of the Peter the First Range. The distribution of hypocentres within the range suggests the presence of a southeastdipping zone of seismicity that can be traced updip to the Petrovsky thrust fault. Beneath this upper crustal seismicity, at a depth of 12-17 km, is an aseismic zone that may be related to the presence of low-strength evaporites at the base of the sedimentary section. This aseismic zone is in turn underlain by anomalously deep crustal seismicity, at a depth of 17-30 km. The deeper seismicity appears to be associated with a southeast-dipping feature approximately 10 km in thickness. This dipping seismic zone is interpreted as a zone of intracontinental subduction, whose surface expression is the Vakhsh thrust fault, at the frontal edge of the Peter the First Range.

Seismotectonics of the Tajik Basin and Surrounding Mountain Ranges

“…In contrast, the Tajik basin's lithosphere between these mountain ranges was proposed to behave as a relatively rigid block (van Hinsbergen et al, ). Intense crustal seismicity (Burtman & Molnar, ; Fan et al, ; Hamburger et al, ; Lukk & Yunga, ; Nersesov & Popandopulo, ), including seven shallow magnitude 7+ earthquakes within the last century, demonstrates the high seismic propensity within the basin and along its margins (Kulikova, , and references therein; Metzger et al, ). This poses the question to what extent the crustal seismicity is driven by subcrustal processes.…”

The fractal description of seismicity

“…The Garm catalogue (see e.g. Hamburger et al 1993) contains very small seismic events with the Rautian energy class K ≥ 4 [the empirical relation to magnitude being given by m = ( K − 4)/1.8].…”

A hybrid multiple event location technique to obtain ground truth event locations

“…Even though we have only three stations, the RCA geometry is so favourable that the combined secondary azimuthal gap is only 110 • . We used a local velocity model by Hamburger et al (1993) to predict traveltimes for the 44 Pg and Sg readings. Again, we had no close-in stations; so, we ran RCA with depth fixed at 12 km.…”