|
|
||||||||
1 Laboratoire de Tectonique
Institut
de Physique du Globe de Paris
BP89, 4 place Jussieu
75005 Paris,
France
(J.L.-Z., Y.K., P.T.)
2 Institute of Tibetan Plateau
Research
Chinese Academy of Sciences
P.O. Box 2871, 18 Shuang Qing
Road
Beijing 100085, China
(J.L.-Z.)
3 Insitute of Geology
China
Earthquake Administration
Beijing 100029, China
(X.X., G.C.,
W.C.)
4 Laboratoire de
Géologie
Ecole Normale Supérieure
24 Rue Lhomond
75231
Paris Cedex 05, France
(C.L.)
5 Tianzhu Bureau of Earthquake
Administration
92 Tuanjie Road Tianzhu County
Gansu Province 733200,
China
(B.Z.)
The Haiyuan fault is a major active left-lateral fault along the northeast
edge of the Tibet-Qinghai Plateau. Studying this fault is important in
understanding
current deformation of the plateau and the mechanics of continental deformation
in
general. Previous studies have mostly focused on the slip rate of the fault.
Paleoseismic
investigations on the fault are sparse, and have been targeted mostly at the
stretch of the fault that ruptured in the 1920 M
8.6 earthquake in
Ningxia Province.
To investigate the millennial seismic history of the western Haiyuan fault, we
opened
two trenches in a small pull-apart basin near Songshan, in Gansu Province. The
excavation exposes sedimentary layers of alternating colors: dark brown silty to
clayey deposit and light yellowish brown layers of coarser-grained sandy
deposit.
The main fault zone is readily recognizable by the disruption and tilting of the
layers.
Six paleoseismic events are identified and named SS1 through SS6, from youngest
to oldest. Charcoal is abundant, yet generally tiny in the shallowest parts of
the trench
exposures. Thirteen samples were dated to constrain the ages of paleoseismic
events.
All six events have occurred during the past 35003900 years. The
horizontal offsets
associated with these events are poorly known. However, events SS3 to SS6 appear
to be large ones, judging from comparison of vertical separations and widths of
fault
zones. The youngest event SS1 instead seems to be a minor one, probably the 1990
Mw 5.8 earthquake. Thus, four large events in
35003900 years would imply a recurrence
interval of about 1000 years. Three events SS2 to SS4 prior to 1990 occurred
sometime during 14401640 A.D., shortly after 8901000
A.D. and 0410 A.D., respectively.
We tentatively associate them with the 1514 A.D., 1092
A.D., and 143 or
374 A.D. historical earthquakes. Taking 10 ± 2 m of slip for
large events (SS3 and
SS4), comparable to the 1920 M >8 Haiyuan earthquake, their
occurrence times
would be consistent with the long-term 12 ± 4 mm/yr estimate of
Lasserre et al.
(1999). However, a more realistic evaluation of slip rate and its
possible change with
time requires a more rigorous determination of coseismic slip amounts of past
earthquakes.
Online material: Trench photos with interpretation.
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |