Earthquake cycle far from plate boundaries: Learning from Mongolia earthquakes

Earthquake cycle is a well-accepted concept when dealing with active faults bounding tectonic plates or large lithospheric blocks. Usually, along those faults the slip-rate is large enough, in the range of few cm/yr to few mm/yr, to produce earthquakes often enough at the geological timescale, thus allowing to discuss earthquake cycle.

Away from active plate boundaries, fault systems are less structured, slip-rate can be only few tenths of mm/yr, and earthquake return-time gets longer. Thus, discussing earthquake cycle becomes more difficult. In fact, even the possibility that successive earthquakes occur along the same fault becomes arguable.

Mongolia, at the northern limit of the India-Eurasia collision zone, far from plate boundaries, presents a unique opportunity to examine the relevance of the concept of earthquake cycle in intra-plate context.

The 1967 M7.1 Mogod earthquake occurred in central Mongolia. No unambiguous evidence of past earthquakes could be identified for certain in the morphology, suggesting that this event occurred as an isolated event on some remanent older geological structure. However, paleoseismological investigation shows that at least a previous event occurred along the same fault about 25 kyr BP.

In the NorthWest of Mongolia, in 1905, two M8 earthquakes occurred 14 days apart along respectively the Tsetserleg and the Bulnai faults. The rupture traces associated with each of those two events are only few kilometers apart. Slip-rate along the Bulnai fault was estimated to be about 3 mm/yr. Here we have determined that the slip-rate along the Tsetserleg fault is one order of magnitude lower, about 0.3 mm/yr. Accordingly, paleoseismological trenches along the Tsetserleg fault have revealed that the average earthquake return-time along that fault is about 6 ky, two to three times longer than along Bulnai. Our recent investigation along the Bulnai fault, using lacustrine paleoseismology, shows that such doublet as in 1905 is not unique in the history of this fault system and that, in fact, the fault system shows a pattern resembling a super cycle, similar to what has been document along more active fault systems. When integrating the Bulnai-Tsetserleg fault system together with other documented faults in western Mongolia, it appears that such earthquake super cycle might in fact affect the entire regional fault system, and not only Bulnai-Tsetserleg. The reason why those two faults, which are almost touching each others, did not rupture during the same earthquake remains unclear to date. Our recent monitoring of the microseismicity in the area where those two large faults intersect shows that the current regime of microseismicity is very different between Bulnai and Tsetserleg. Using this microseismicity, we might be able to better constrain the geometry of the Tsetserleg fault at depth, as well as the general fault structure in the intersection area. It might be the key to understand the 2 weeks time-delay between those two events and, overall, how stress build-up in this complex fault system to produce earthquake super cycles.