The first two figures show the use of sedimentation rates in age refinement for submarine earthquake generated turbidites. Figure 1 shows two examples from the NSAF and Cascadia using interseismic sedimentation to trim PDF's, and compares these results to the known ages of two earthquakes. Figure 2 shows the age model for NSAF core 49PC, using 14C PDF's and sedimentation rates and the fit of PDF peaks to moving window and regressed age curves.
Figure 3 shows the correlation details for the NSAF system used in this paper, revised with RGB data from Goldfinger et al. (2007).
Figure 4 shows the age deviation from zero for NSAF and Cascadia earthquakes. Zero deviation represents coincident events on both systems. This shows that for four methods of determining event timing, all show a strong in phase relationship with Cascadia earthquakes preceding NSAF earthquakes.
Figure 5 shows coseismic displacements at the surface from 14 m of slip on the Cascadia megathrust. Figure 6 shows displacements at the surface from 7 m of deep afterslip on the Cascadia megathrust. Figure 7 shows cumulative displacements at the surface from 60 years of viscoelastic relaxation following14 m of slip on the Cascadia megathrust.
Figure 1. OxCal methods test using the well constrained 1906 earthquake and associated paleoseismic data onshore and offshore. The left panel shows the hemipelagic (H) data determined from visual observation, physical property data mineralogy and x-radiography. H data is then input to OxCal and with raw 14C ages are converted to time via sedimentation rate curves developed for each site. Right panel shows four ways to calculate the age of the 1906 earthquake, with the preferred method being the use of underlying and overlying hemipelagic intervals, historical data (no written record of an earthquake between the date of the first San Francisco Mission built in 1769 and the 1838 earthquake). Lower panel shows result of similar application of constraining hemipelagic data to the AD 1700 Cascadia earthquake, the age of which is known independently (Satake et al. 2003). This method commonly resolves the ambiguities inherent in radiocarbon dating where PDF’s have multiple peaks or broad distributions due to the slope or complexities of the calibration curve. In this example, the overlying 300 years or hemipelagic sediment restricts the PDF to the earlier of two peaks. Such constraints are typically not as strong for events deeper in the core section because the present day upper boundary layer is absolute in these examples.
Figure 2. Age model for core M9907-49PC at Noyo Channel. OxCal Probability Density Functions for 14C ages shown in black. Gray PDF’s are shown for hemipelagic ages. PDF’s are plotted with their bases at the cumulative hemipelagic depth point (blue diamonds) for that event. Spline curve fit in red is a moving window average sedimentation rate. Linear regression shown as dashed line. Discordant ages or event occurrence within the low probability parts of the PDF would require abrupt sedimentation rate changes not supported by the relatively smooth rate curve, or expected in marine sedimentation. This can serve as a check of the reasonableness of ages in a series, and supports event occurrence near the probability peaks in this series, as shown by Goslar et al. (2005). Maximum cumulative thickness error (assuming each measurement was off by 0.5 cm in the same direction) for each event shown as dashed red line.
Figure 3. Correlation diagram for the uppermost ~ 3000 years in NSAF system cores, modified after Goldfinger et al. (2007). Channel locations shown in Figure 2. Light blue traces are Gamma density, dark blue are magnetic susceptibility. RGB color traces from line scan imagery are also shown. Areas affected by coring artifacts are shown for all stratigraphic data. 1 cm resolution magnetic susceptibility data are shown for Noyo cores, and 3 mm high-resolution data are shown for all other cores. Event numbers in blue at core centers. OxCal age peaks and ranges shown on cores 49PC and 54KC at left. Gualala ages have been corrected for sample thickness only. Correlation for weak events T7a and T9a uncertain and not shown beyond Noyo Channel. T7a is datable, and included in our time series; T9a appears to be limited to two adjacent proximal core sites, and is not included in the recurrence statistics. 49PC high-resolution magnetic trace is punctuated by sample voids, indicated by yellow squares.
Figure 4. Histogram of pre and post NSAF earthquake times for Southern Cascadia earthquakes.
Figure 5. Coseismic displacements at the surface from 14 m of slip on the Cascadia megathrust as described in the text. Colored contours and colorbar show vertical displacement, scaled arrows show horizontal displacement.
Figure 6. Displacements at the surface from 7 m of deep afterslip on the Cascadia megathrust from the post seismic afterslip model described in the text. Colored contours and colorbar show vertical displacement, scaled arrows show horizontal displacement.
Figure 7. Cumulative displacements at the surface from 60 years of viscoelastic relaxation following 14 m of slip on the Cascadia megathrust. Colored contours and colorbar show vertical displacement, scaled arrows show horizontal displacement.
Table 1. Land ages, error ranges and OxCal input files for all sites and OxCal combines. References for both tables 1 and 2 are given on the reference tab in each sheet.
Download: Table1_Land_data.xls [Microsoft Excel file; 1.6 MB]
Table 2. Marine ages and error ranges for all sites. Tabs include erosion corrections, and hemipelagic age calculations, 14C data, average calculations, calculated ages, reservoir age model, recurrence averages and sedimentation rate calculations.
Download: Table2_Marine_data.xls [Microsoft Excel file; 664 KB]
Table 3. Marine Oxcal combines.
Table 4. Rupture length limits estimated from turbidite latitudinal extents.
[ Back ]
