Paleoseismic assessment of the Hat Creek fault using cosmogenic He-3 surface exposure dating in basalt, northeastern California: A proof of concept study

Thursday, 17 December 2015: 10:20
302 (Moscone South)
Dylan Hunter Rood1, William H Amidon2, Ryan E McKeon3, John N Baldwin4, Brian Gray4, William D Page5 and Kenneth A Farley6, (1)Imperial College London, London, United Kingdom, (2)Middlebury College, Middlebury, VT, United States, (3)Dartmouth College, Hanover, NH, United States, (4)Lettis Consultants International, Inc., Walnut Creek, CA, United States, (5)Pacific Gas and Electric Company, San Francisco, CA, United States, (6)JPL/NASA/Caltech, Pasadena, CA, United States
We demonstrate the successful application of cosmogenic He-3 surface exposure dating to construct high-resolution paleoseismic records in the low-potassium Hat Creek basalt. We dated multiple events on the Hat Creek fault, a 47 km-long segmented normal fault at the transition between the Walker Lane and Cascades tectonic provinces. He-3 age dating methods are used to resolve fault slip rate, earthquake timing, and slip-per-event of ground rupturing earthquakes. We are able to separate individual earthquake ruptures using dating of exposed bedrock fault scarps and large fallen blocks. Age and displacement of the lava flow are resolved with high precision and accuracy (+/- 5-10%). The coeval He-3 ages of bedrock fault scarps and toppled blocks, and the occurrence of coeval rock falls on multiple distant fault sections, are both consistent with the hypothesis that earthquake ruptures topple large numbers of blocks near the fault scarp. Thus, the ages of rock fall events can provide the timing of earthquake ruptures. By comparing the timing of ruptures on multiple fault sections, rupture length and regional spatial patterns of strain release can be interpreted.

He-3 results on the Hat Creek basalt flow surface (23.8 +/- 1.4 ka; 1SD, N=7) agree with previous Ar-40/Ar-39 ages; however, the cosmogenic exposure age has higher precision. Exposure dating of bedrock fault scarp profiles suggest ages for ground-rupturing earthquakes at 13.1 +/- 1.0 ka and 10.0 +/- 2.4 ka, and a slip-per-event of >2.6-2.9 m. At the same site, dating of toppled megablocks (>2 m in height) and adjacent cliff surfaces suggest exposure ages of scarp-derived rock falls at 12.0 +/- 0.5 (N=2) and 9.1 +/- 1.5 (N=4). Exposure ages of additional rock falls at multiple sites suggest multi-section events at 3.3 +/- 1.1 (N=9) and ~0.2 ka (N=3).

Our results show that cosmogenic He-3 can be used to reconstruct late Pleistocene to late Holocene earthquakes and fault behavior with age resolution comparable to traditional paleoseismic methods, and represents a novel yet robust dating tool for earthquake hazard analysis in basalt terranes.