Skip to main content
Unpublished Paper
REFINE SLIP RATES AND SEGMENTATION OF THE NORTHERN SANGRE DE CRISTO FAULT, COLORADO’S LARGEST ACTIVE FAULT
USGS NEHRP Grant 03HQGR0076, Final Technical Report (2006)
  • James P. McCalpin
Abstract
We excavated a 25 m-long trench on the northern part of the Crestone section of the Northern Sangre de Cristo fault zone, to see whether the Crestone section might be composed of more than one (independent) rupture segments. The Carr Gulch trench exposed evidence for 3 paleoearthquake displacements in the past 27.4 ka. The existence of 3 colluvial deposits (units 2, 3, 4) indicates that the ca. 4.5 m of surface offset was formed during 3 surface-faulting events with an average displacement of about 1.5 m each. The events occurred at about 8 ka, 20 ka, and between 22.5 and 27.5 ka. In contrast, at the Major Creek site on the southern part of the Crestone section, there has been 8.9 m of displacement in a similar time period (since early Pinedale time). The displacement-per-event, reconstructed from trenches at the Carr Gulch and major Creek sites, is similar (1.5 m at Carr Gulch; 1.4-2.2 m at Major Creek). Therefore, the difference in net displacement implies that fewer events have occurred at Carr Gulch than at Major Creek. The latest event at both sites occurred at nearly the same time, and may represent a 90 km-long rupture of the entire Crestone section. The penultimate event at Major Creek, however, occurred between about 11 ka and 15 ka, much younger than the penultimate event at Carr Gulch (ca. 20 ka). At Major Creek we have no direct dating control on the earlier events, but the youngest of these 3 events could conceivably correlate with the penultimate event at Carr Gulch. The oldest event at Carr Gulch could correlate with one of 2 (but not both) early events at Major Creek. Despite the ambiguities of event-matching with such fragmentary dating control, it is clear that there have been 5 inferred faulting events at Major Creek since early Pinedale time, but only 3 events at Carr Gulch. Only the latest event appears to be a possible simultaneous rupture at both sites. The penultimate events at Carr Gulch and Major Creek cannot be the same event, if the age constraints at each site are correct. Some of the earlier events may have ruptured at both sites, but all of them could not have. This pattern of fault behavior suggests that there is a nonpersistent segment boundary between the Major Creek and Carr Gulch sites. We propose that this segment boundary lies at the junction of the Villa Grove fault zone and the NCSF. Based on the record at Carr Gulch, less than half of the range-front ruptures south of the VGFZ have continued north past the segment boundary. The other half of the ruptures have been deflected out onto the VGFZ, and are responsible for the late Quaternary fault scarps there. Therefore, we propose to subdivide the Crestone section (as defined by Widmann et al, 1998) into 2 rupture segments, with a segment boundary where the VGFZ joins the NCSF. This subdivision is identical to the suggestion of Jack Benjamin and Associates and Geomatrix Consultants (1996), that the Crestone section should be subdivided into a 38-km-long segment north of the Major Creek/Kerber Creek thrust fault zone, and a 52-km-long segment south of it. We propose the name Hayden Segment for the northern segment and Major Segment for the southern segment. The Major segment is 52 km long, but most of the time it probably ruptures with the VGFZ, which is an additional 19 km long. According to Wells and Coppersmith (1994), a rupture length of 71 km implies an earthquake of M7.3, whereas the average displacement of 2.15 m implies an earthquake of M7.0, and the maximum displacement of 2.9 m implies an earthquake of M6.9. The Hayden Segment is 38 km long and has average per-event displacements of 1.5 m (latest 3 events) at Carr Gulch. The only other displacement estimate is the MRE surface offset at Rock Creek (1.7 m; McCalpin, 1982). We use these values as Davg and Dmax, respectively. According to Wells and Coppersmith (1994), a rupture length of 38 km implies an earthquake of M6.9, the average displacement of 1.5 m implies an earthquake of M6.9, and the maximum displacement of 1.7 m implies an earthquake of M6.8. These values yield an estimate of M6.9 for Mchar. The average recurrence interval over the latest 2 closed seismic cycles (5 ka and 12 ka) is 8.5 ka. During 2-segment ruptures, the combined length of the Major and Hayden segments ruptures for a total rupture length of 90 km. According to Wells and Coppersmith (1994), a rupture length of 90 km implies an earthquake of M7.4. For the 2-event rupture we use the same Mchar-D values as for the Major segment. Therefore, we compute a preferred value of the magnitude of the characteristic 2-segment earthquake (Mchar) as 7.3. The space-time diagram suggests that only about 2 out of 5 (40%) of events that rupture at Major Creek also rupture at Carr Gulch. Thus, if only about 40% of ruptures are 2-event ruptures, the recurrence period of such ruptures should be approximated by dividing the recurrence interval of the individual segments by 40%. This yields a recurrence for 2-segment ruptures of ca. 12-30 ka. At present the National Seismic Hazard Map assumes a slip rate of 0.18 mm/yr for the entire Sangre de Cristo fault. We propose that, in the next iteration of the map, the NCSF be modeled as being composed of 2 rupture segments, each with a different Mmax and recurrence time, and including the occurrence of occasional 2-segment ruptures with an Mchar of 7.3 and a return period of 12-30 ka.
Keywords
  • active fault,
  • Colorado,
  • Sangre de Cristo
Publication Date
2006
Citation Information
James P. McCalpin. "REFINE SLIP RATES AND SEGMENTATION OF THE NORTHERN SANGRE DE CRISTO FAULT, COLORADO’S LARGEST ACTIVE FAULT" USGS NEHRP Grant 03HQGR0076, Final Technical Report (2006)
Available at: http://works.bepress.com/james_mccalpin/15/