Previous   |    Next

doi:10.2204/iodp.proc.320321.209.2013

Methods

Sediment records from different holes can be aligned at high spatial resolution using detailed physical properties measurements generated for each hole. For Sites U1336, U1337, and U1338, we used three primary sets of measurements: digital photographs, gamma ray attenuation (GRA) density, and magnetic susceptibility (MS). It is noteworthy that physical properties records for Neogene sediment at these sites are far superior to those generated in contemporaneous sediment at previous drill sites in the EEP.

Preceding Expedition 320/321, new depth scales were introduced for scientific drill holes (see “IODP Depth Scales Terminology” at www.iodp.org/program-policies/). The most fundamental depth unit is core depth below seafloor (CSF), which comes from drilling logs. This depth is the best estimate for the true depth of sediment cores and sediment samples below the seafloor. For Expedition 320/321, CSF (in meters) is the same as meters below seafloor (mbsf) used on many previous scientific drilling cruises. As emphasized above, however, successive cores in a single hole do not lead to a complete stratigraphic record.

Composite depth sections require a different depth scale. This scale is designated core composite depth below seafloor (CCSF) and typically exceeds CSF by ~10%. The reason for this difference lies in the methodology of splicing and core expansion during retrieval (Lisiecki and Herbert, 2007). Upon splicing, gaps between successive sediment cores are accounted for, but not the cause. The gaps arise because sediment generally expands during core recovery such that a full 9.5 m core represents about 9.0 at depth with some sediment lost.

Prior to constructing a spliced section the data collected by the various instruments in the core laboratory must be cleaned of “spikes” and bad intervals. Most of the spikes occur at the end of core sections where part of the measurement includes some air. Bad intervals generally are related to coring disturbance or incomplete recovery. As an initial step in analysis, core section JPEG images were converted to x-y data using a modification of the process described in Wilkens et al. (2009). Data were cleaned manually by overlaying data, such as GRA or MS, on top of the core images. Data that were judged to be anomalous that corresponded to either a section end or visual core disturbance were deleted (Fig. F2). In cases where GRA density is a little low but MS data or images can be correlated we did not eliminate the GRA data, as they possibly indicate a part of the core that was stretched during recovery.

To build a composite depth scale we first fix the core that best seems to capture the mudline from all of the holes cored at a site. Cores from other holes are then moved along a depth axis (CCSF) so that correlative horizons align. A tie point near the base of the first fixed core is chosen, generally at some prominent feature that is also seen near the top of a deeper (floating) core. The core gap between the top two cores in the first hole is then spanned by the floating core from the second hole. The process is continued until a depth is reached where correlative features cannot be recognized. The result of the splicing operation is a table of offsets for each individual core at a site and a second table of the locations of the tie points that define the composite section (e.g., Tables T1, T2). The results of applying the splice are illustrated with scanned core images in Figure F3 and with core data in Figure F4.

In the case where no correlation from the bottom of one core to the tops of any of the others at the site can be made, the protocol is to “append” the next deeper core of the same hole to the bottom of the composite section. This means that the offset of the upper core is carried without change to the lower core. Appends are most common in intervals of only partial core recovery. It is important to note that this procedure does not account for core expansion and gaps between cores that are seen in intervals where recovery is 100%—that is, the core liners are full. Thus, too many appends may skew calculation of overall core expansion.

In revising the composite depth scales developed during Expedition 321 coring operations, we attempted to maintain the original tie points whenever possible so that samples that were originally selected to be within the composite section will remain so, even if the depth has shifted slightly. In general, within the shallower (<300 m CCSF) advanced piston corer (APC) cored intervals, changes in offsets from the original splice and/or tie points are on the order of a few tens of centimeters. Deeper at Sites U1337 and U1338 the revisions are greater because of the increased difficulty in aligning cores with partial recovery or cores that have undergone drilling disturbance during extended core barrel (XCB) coring operations. Although some intervals of deeper sections correlate well, this is not always the case. Users of these offset and splice tables should evaluate the data provided in the “Appendix” before using them as a basis for critical observations.

Once the composite section has been defined, a final stretch/compress process is required for intervals of core that are not included within the splice (off-splice). The problem and the solution process are illustrated in Figure F5. Above the interval of Core 321-U1337A-4H selected for the composite section, a boundary at ~21.5 m CCSF corresponds to a feature at ~22.0 m CCSF in the composite. The composite interval at these depths comes from Core 321-U1337B-3H (Fig. F3). For future sampling of off-splice sediment, it is imperative that features such as this boundary be mapped to the composite so that equivalence in material can be maintained. This is accomplished by identifying tie points between off-splice intervals and the composite section and then stretching or squeezing the CCSF depth scale through interpolation to produce an adjusted CCSF scale. Tables of off-splice tie points are provided for each hole in the “Appendix.” Adjusted CCSF depths are provided in the cleaned data files as well.

Top of page   |    Previous   |    Next