Preliminary investigation of factors affecting the seismic potential of the Bartlett Springs fault zone, northern California

The Bartlett Springs fault (BFS) extends 170 km from its south end, a large releasing bend in the Hunting Creek fault, to its north end, another large releasing bend in the Lake Mountain fault. The seismic potential of BFS is poorly known because of incomplete mapping, poorly constrained geologic slip rate and creep rates. Before our study only part of BFS was mapped as Holocene-active for a variety of reasons including the heavy rainfall and steep slopes causing extreme erosional conditions (many landslides), heavy vegetation and lack of detailed aerial photography. We acquired new 1:12,000-scale aerial photography of the entire BFS, from which we interpreted geomorphic features that indicate Holocene faulting extends along the entire BFS. The new mapping, formatted for use in geographic information systems, clarifies possible geometric constraints on fault segmentation. To better constrain the spatial variation of creep rate along BFS, we increased creep monitoring sites from two to four, presenting the latest results in our poster. The 4-yr average creep rate for our Lake Pillsbury site is 2.8 ± 0.4 mm/yr (1-SD), comparable to a creep rate estimated from a step in the velocity field (3.4 ± 0.8 mm/yr) using all USGS GPS array points across the central BSF. Lacking a geologic slip rate for BSF, we estimate an average velocity across the fault using a rigid block model of the GPS site velocities. This yields ~6.5 mm/yr, which is comparable to the 6 mm/yr long-term rate observed on the Northern Calaveras fault (NCF). Much NCF slip and probably additional slip from the Greenville fault transfers indirectly to the BSF via the Concord-Green Valley fault (CGVF). The NCF and CGVF have long-term creep rates ranging from 1.8-4.4 mm/yr, comparable to our estimates for BSF. For seismic hazard estimation, the segmentation of the BSF may depend on many factors, including the spatial variation in aseismic moment release, the size and 3D structure of the largest geometric discontinuities, the state of stress relative to the failure stress of each section and the timing and extent of previous ruptures. Microseismicity distribution is also relevant to seismic potential, because it tends to correlate with fault creep. The central third of the BSF has high microseismicity and with respect to with the plate boundary its strike ranges from transtensional to parallel, thus favoring stable sliding and unclamping. Our mapping indicates the largest BSF discontinuities are a 2.5-km left stepover at the Middle Fork of the Eel River and a 2.5-km right stepover at Wilson Valley. Some previous hazard models have assumed the BFS is likely to have a complete 170-km rupture of Mw~7.2, however, shorter rupture scenarios may be much more likely. For example, one more plausible rupture, for the ~70-km central section having high microseismicity, yields Mw~6.8, because its rupture would tend to terminate in the adjacent sections oriented favorably to locking. Many other plausible scenarios for segments of varying length can be posed, each needing to be weighted inversely to the likely difficulty of each rupture to occur.