2018-04-02_REVISION - M1977342 - Laserfiche WebLink

Home Browse Search

SEO NFOUR Seismic Nuard Inputs and seismogenic of 1.0 (Table 1). Exceptions include faults that may be secondary and dependent on other faults (e.g., the unnamed faults near Leadville and the unnamed faults in Williams Fork Valley), faults or fault features that may have a non-seismogenic origin (e.g., unnamed faults near Burns), and faults that may be too short (_< 10 km) to independently generate significant earthquakes (e.g., unnamed faults in Granby basin). The probability of activity for faults that do not show evidence for repeated Quaternary activity was individually judged based on the available data and the criteria explained above. The probability of activity values range from 0.2 to 1.0(Table 1). All of the faults are assumed to be dominantly normal-slip faults, although some show components of oblique or possible reverse slip. All faults are modeled as single, independent, planar sources extending the full extent of the seismogenic crust, with the exception of the unnamed faults in Granby basin and unnamed faults near Leadville (Table 1). These faults are modeled as zones with multiple parallel planes within defined zone boundaries. Additionally, due to their proximities to the sites, we used a curvilinear surface model for the WFMF and Mosquito fault in the analyses. This better represents their curved geometry and results in more accurate calculations of distances. To model the curvilinear nature of normal faults,we digitized the primary, most active fault trace and projected these curves down-dip using a weighted mean strike. This method eliminates the overlap created at concavities and fills. The gaps created at convexities when creating the down-dip projection. It results in a smoothed curvilinear surface that generally mimics the trend of the fault trace. Our model assumes constant dip with depth and listric geometries are not considered. Based on an analysis of Front Range seismicity(Bolt et al.,2003),we model the thickness of the seismogenic crust as 20 t 5 km with the best estimate given a weight of 0.6 and 0.2 weight given to the one sigma values. Fault dips are assumed averages throughout the entire seismogenic crust. Dips range from 35'to 90'(Table 1)and are based on available data and consideration of the relationship of the faults to other structures(e.g. intrabasin faults are generally assumed to be steeper). However, subsurface data is generally lacking and so default dips of 50' ±15' were assumed for many range-bounding faults, including the WFMF (Table 1). This distribution is consistent with the dip inferred by Kellogg et al.(2011) of the WFMF in their cross section,and is the default distribution recommended by the Basin and Range Province Earthquake Working Group 11 to the USGS for use in the 2014 National Seismic Hazard Maps (Lund,2012;see Issue G4). It is noteworthy, however, that fault dips often appear steeper in the northern Rio Grande rift than typical faults in the rest of the Basin and Range Province, perhaps due to the complex geometries of reactivating older reverse faults that were active in the Laramide orogeny, and the Mosquito fault is a good example of this(Table 1). Preferred characteristic (or maximum) magnitudes (weighted 0.6) were estimated using two empirical relationships weighted equally: (1) Wells and Coppersmith (1994) for surface rupture length (SRL) and all types of faults; and (2) the "censored" relation of Stirling et al. (2002) for SRL. Although the Wells and Coppersmith (1994)relation is somewhat outdated (Sirling et al., 2012), it is well-established and based on a well-documented dataset. However, many paleoseismic studies of normal faults have noted a discrepancy between displacement-based magnitudes and SRL-based magnitudes from the Wells and Coppersmith (1994) relation (e.g., Olig et al., 1994; Mason, 1996; Carpenter, et al., 2012; DuRoss et al., 2015), with the latter being consistently lower. Unfortunately,displacement per event data is not available for most of the faults in this study. Therefore, we also used the Stirling et al.'s (2002) censored relation, '274MR.18N 11