Previous   |    Next

doi:10.2204/iodp.proc.340.202.2015

Materials and methods

The method used to generate a composite depth scale and splice is adopted from and largely follows the strategy developed for paleoceanographic IODP expeditions (e.g., Expedition 303 Scientists, 2006; Expedition 323 Scientists, 2011; Expedition 339 Scientists, 2013; Jaeger et al., 2014).

A note on depth scales

In its simplest terms, the CSF-A depth scale is based on the length of the drill string plus the length of the material recovered during the process of coring. The zero depth point in the CSF-A scale is defined by the uppermost sediment in the first core (commonly referred to as the “mudline” core); the depth to any point along that core is then determined by adding the distance that point occurs from the mudline. The length of each recovered core is then used to advance the drill string, setting a new datum for subsequent cores from the same hole. The CSF-A scale is inaccurate because of ship heave (which is not corrected for in APC coring), tidal and nontidal variations in sea level, and other sources of measurement error.

The goal of constructing a composite depth scale is to place coeval, laterally continuous stratigraphic features into a common frame of reference by manipulating the CSF-A depth of individual cores to maximize physical property correlation between holes. This is a rather more elegant method of depth scale construction because instead of relying on the drill string measurement, the composite core depth below seafloor (CCSF-A) is built by continuously correlating physical features downhole from the mudline. The core with the most representative mudline defines the top of the stratigraphic section and becomes the anchor in the CCSF-A depth scale. It is typically the only core in which the depths are the same for both the CSF-A and CCSF-A scales. Each core downhole is tied to the composite section by adding or subtracting a depth offset (a constant) that best aligns the observed lithologic variations for cores from adjacent holes. Using this method, no cores are stretched or squeezed to facilitate correlation but are instead “hung” next to each other to build the composite section. It is therefore not usually possible to align all features perfectly between holes; in such cases, offsets are chosen to maximize correlations over the whole core.

In the process of constructing the composite section, the CCSF-A scale is always almost expanded relative to the CSF-A scale. This expansion is typically reported as 5%–20% (e.g., Expedition 303 Scientists, 2006; Expedition 323 Scientists, 2011; Expedition 339 Scientists, 2013; Jaeger et al., 2014). Although it is often difficult to identify the specific cause for the growth factor, it is widely thought to result from the coring process and includes (but is not limited to) stretching and squeezing, decompression and degassing, and curation. In response to this growth, the CCSF-B (composite core depth below seafloor, method B) depth scale is intended to correct the CCSF-A scale for any empirically observed expansion (Jaeger et al., 2014). CCSF-B depths are produced by correcting for the average growth of the CCSF-A scale relative to the CSF-A scale over a sufficiently long interval that random variations in drill pipe advance due to ship heave, tides, and other factors are averaged to be negligible (Jaeger et al., 2014). This scaling produces a complete composite section sequence that is the same length as the total cored interval. As the CCSF-B scale is a closer representation of the actual drilling depths than the CCSF-A scale, it should be the scale used for estimation of sediment accumulation rate. However, because the CCSF-B scale provides an estimate of sediment thickness, intervals that are targeted for sampling within each core section are better represented by the uncompressed CCSF-A scale (Jaeger et al., 2014).

Correlation

Given the diverse lithology of the sediments recovered at Site U1396, Holes U1396A–U1396C were primarily correlated using shipboard magnetic susceptibility (MS), which was acquired at 2.5 cm intervals prior to the core being split on the Whole-Round Multisensor Logger (WRMSL). To independently corroborate MS correlations and assist decisions if MS correlation became ambiguous, density data measured through gamma ray attenuation (GRA) on the WRMSL at 2.5 cm intervals and natural gamma radiation (NGR) radiation measured at 10 cm intervals on the NGR system were also used for correlation. The first and last data points from each measurement were masked in the data set because they contain volumetric edge effects in MS, GRA, and NGR that did not result from lithologic variances. Physical property data were imported into the Corewall Correlator software (version 2.0; http://www.corewall.org), which provides depth shifting capabilities and associated correlation values that were used to maximize the correlation between cores from different holes.

During correlation, it was noted that the MS data reported for Sections 340-U1396C-9H-6 and 9H-7 suffered positioning errors during measurement on the WRMSL. The WRMSL-assigned length of Section 9H-7 in the LIMS database is 143 cm, but the section is only 70 cm in length. This is manifested in the WRMSL-derived MS data as a 72.5 cm interval of repeated measurements between 83.35 and 84.05 m CSF-A (Fig. F1A). Comparison of the WRMSL MS data (Fig. F1A) to the MS point source data collected on the Section Half Multisensor Logger (SHMSL) confirms this discrepancy (Fig. F1B). Removal of 72.5 cm of repeated MS data in the WRMSL record produces new WRMSL CSF-A depths for Sections 9H-6 and 9H-7 (Fig. F1C; Table T1) and a data set that now reflects the true core length and better replicates the SHMSL data. The amended CSF-A data for Core 9H is used for correlation and in construction of the CCSF-A depth scale.

Splice development

Once both holes are mapped onto the common CCSF-A depth, the most representative continuous section can be chosen to reflect the complete record of the site. The spliced record is sampled to avoid missing and disturbed intervals by using sections from more complete holes. At Site U1396, Core 340-U1396B-2H was used to replace disturbed Core U1396A-2H and the remainder of the record switches between Hole U1396A and Hole U1396C to bridge WR and core gaps. The choice of splice tie points is somewhat subjective, but attempts were made to avoid the uppermost and lowermost parts of core, which may be more susceptible to disturbance, and to utilize, where possible, the longest possible sections within individual cores to reduce the number of necessary tie points. Data within the splice are assigned an additional CCSF-D depth. These depths are exactly the same as the CCSF-A depths; assignment of a CCSF-D scale to an interval simply implies it is incorporated within this spliced record. Because 10 cm WR samples were taken at 3–6 m intervals in Hole U1396C, the resulting splice will likely lean more heavily on cores from Hole U1396A.

Top of page   |    Previous   |    Next