SHIPBOARD SCIENTIFIC PROCEDURES

The drill sites are numbered consecutively and may be associated with one or more holes drilled at a specific location. A letter suffix is used to identify an individual hole drilled at a given site. For example, the first hole drilled on the Southeast Greenland (SEG) shelf is designated as Hole SEG01A, the second hole drilled at the same site is Hole SEG01B. Every hole drilled is registered in the ODIN database (see "ODIN Database") whether or not core was recovered. This procedure differs slightly from that used by the Deep Sea Drilling Project for Sites 1–624 but follows more recent ODP notation designed to prevent ambiguity between site and hole numbering. The important difference in how a site is defined by ODP and during Leg 163X relates to the different systems used for the dynamic positioning of the ship. On the JOIDES Resolution, the ship's position during drilling is dynamically controlled by means of an acoustic beacon located on the seabed. An ODP site is defined by the target ground accessible using the same beacon. All holes drilled without redeploying an acoustic beacon are indexed alphabetically. Deploying a new seabed beacon constitutes targeting a new site.

On the Norskald, the dynamic positioning system is controlled in several ways: (1) using the signal from an acoustic beacon mounted to the seabed template, (2) tension on the tautwires of the template, and/or (3) a tautwire positioned outboard of the template (see "Drilling from the Norskald" section in the "Drilling Operations" chapter). During normal operations, the wireline core barrel and drill bit can be retracted from the hole and redeployed in the same hole without repositioning the seabed template. It is also possible to reposition the template for a new hole by first raising the wireline and outer API and drill bit to the seabed, elevating the seabed template off the seafloor, and finally raising the heave compensator–rooster box with the attached drill string to the top of the derrick. This procedure provides a maximum of 20 m of clearance from the seabed. In calm seas and areas of smooth ground the drill ship can be moved as far as 400 m from its original location with the drill string and seabed template elevated. For more remote repositioning of the ship, the entire drill assembly must be brought on deck. If the latter is required, we define the new station as a new site.

Dynamic positioning of the Aranda is controlled either using a tautwire or Differential Global Positioning System (DGPS). The DGPS was used exclusively with the BGS Rockdrill (see "Drilling from the Aranda" section in the "Drilling Operations" chapter). When using DGPS, a drill site is defined by a unique set of navigational coordinates (latitude and longitude). Under normal operations the drilling platform is not launched until the location can be confirmed using these coordinates. Once on station, the drilling platform is lowered to the seabed by the umbilical cable and winch. The location of the drill site is taken as the ship's position at the time that the drill platform touches bottom, with a correction applied to the horizontal separation of the GPS receiver and the A-frame pulley. If the drill platform was deployed two or more times without changing the ship's navigation coordinates, successive holes were referenced to the same site (e.g., Holes SEG01A and SEG01B). A new site was designated when the ship relocated to a new location using a new set of navigation coordinates. This scheme allowed us to return to a previously occupied site at the same navigation coordinates and continue with the alphabetical lettering of holes (e.g., the third hole drilled at Site SEG01 would be designated Hole SEG01C).

The cored interval within a hole is measured in meters below seafloor (Fig. F1). The maximum recovery from a single core barrel is 3 m of rock or sediment, but recovery was commonly less. Onboard the Norskald, the total length of the drill string is measured from the rig floor and corrected to sea level by subtracting the height of the rig floor above sea level. The water depth is computed by subtracting the depth below the rig floor when the sea bottom was first encountered. Similarly, when drilling from the Aranda, depth to the seabed is calculated from the length of cable paid out until touchdown of the Rockdrill on the seabed and correcting for the height of the A-frame above sea level. The drilled interval is referenced to the seabed, while the depth of the cored interval is referenced to the depth below the seabed at which coring begins. The initiation of coring is usually marked by an increase in torque on the drill bit and reduction in the rate of penetration as recorded by the drillers (see Fig. F2). We assume that advance of the drill bit to basement is through unconsolidated sediments, only parts of which may be recovered in the core barrel.

Cores taken from a hole are numbered sequentially from the upper part of the hole downward (Fig. F1). Core numbers and their associated cored intervals in meters below seafloor are usually unique for a given hole; however, this may not be true if an interval is cored twice because of caving of the walls during retraction of the drill string or other technical reasons. A recovered core is divided into ~1.0-m-long sections and placed in three-bay core boxes (Fig. F3) with the upper section of each core located in the left bay with the upper part of the section to the left. The second section is placed in the second bay with its upper part toward the top of the box. The third section is similarly oriented in the third bay. When full recovery is obtained, the sections are numbered from one to three (or higher during the 1998 cruise), with the last section possibly being shorter than 1.0 m. In cases where the recovery is not complete, the core is moved to the top of the core tray and divided as before into as many sections as necessary to accommodate the core. For example, 1.2 m of core would ideally be divided into one 1.0-m length core (referred to as Section 1) and a 0.2-m length core (referred to as Section 2) (Fig. F4). By convention, the top of the recovered core is equated with the top of the cored interval. In some cases sections <1.0 m may be cut to preserve features of interest (e.g., lithologic contacts). To save on storage space, cores collected from several holes were stored in individual bays of the same box. In some cases where only very short cores were obtained, the cores may have been stored in the same bay separated by a plywood divider.

Each piece of the core is numbered sequentially within each core section, beginning at the top of the first section. Consecutive fragments with obvious features allowing realignment are considered to be a single piece. These pieces are given the same number and lettered consecutively (e.g., 1A, 1B, and 1C). Sequentially numbered pieces may be separated by intervals of unrecovered core. Pieces in the core are located by reference to the section containing them and the positions of the top and bottom of the piece measured in centimeters from the upper part of the section. The location of a sample or piece in a core is defined by the following information: Leg, Site and Hole, Core, and Section followed by the top and bottom interval (in centimeters) and piece number. For example, 163X-SEG02A-5-2 (Piece 3, 58–60 cm) uniquely locates a specific sample in the core. The same referencing scheme, with or without the inclusion of the piece number, is used to locate any specific feature of interest in the core.

In the standard ODP sample designation protocol, a letter suffix is added to the core number to identify the type of drilling technique used (e.g., "R" for rotary coring). Because we only used diamond coring during Leg 163X, we have dropped this notation in the sample identification code.

When core is received on deck, details of the hole location, time of recovery, seafloor depth, and coring intervals are noted on the Core Receiving Log form (Fig. F5). This information is entered into the ODIN database (see "ODIN Database") to generate core box labels that are affixed to the end of the core boxes. For each piece of core, the section number and top and bottom interval of the piece (in centimeters) is recorded on the Piece Registration form (Fig. F6). This information is entered immediately into the ODIN database for printing of piece labels. The core is then split using a diamond saw into a working half and an archive half. The bottoms of oriented pieces (i.e., pieces that clearly could not have rotated top to bottom within the core barrel) are marked with a red wax pencil and cut lines are determined in order to preserve important features of the core in both halves. Cores recovered with the wireline diamond coring systems on the Norskald have a diameter of 56 mm and are split into equal halves. Cores recovered with the BGS Rockdrill have a diameter of 49 mm. These cores are split asymmetrically with the working half having a thickness of 30 mm to permit sampling for paleomagnetic analysis. When the recovered core is too small to split but enough material was available for paleomagnetic analysis, the complete piece is logged and stored with the working half. After splitting, the core is rinsed, dried, and placed in the working half and archive half core boxes. Piece labels are then glued to the external surface of the split core with the working half designated with a "W" and the archive half designated with an "A." After labeling, the archive core half is packed for storage. The working core half is photographed with a digital camera before being made available for shipboard petrographic description. Sediments are handled in a similar fashion, although they are not split and the entire core is stored in the working half box. Mud and diamicton are sealed in plastic before storage.

A visual core description (VCD) form (Fig. F7) is used to describe both igneous and sedimentary lithologies for each section of core. These forms are electronically scanned and loaded into the ODIN database (see "ODIN Database"). The first column on the left of the VCD form gives the piece number corresponding to the sketch of the core section in the column to the right. The graphic representation of the core indicates the shape of individual pieces, notable unit boundaries, and any internal features of merit. The column to the right of the sketch shows the orientation of each piece, if determinable. The next two columns to the right show the location of any samples taken for thin sectioning, geochemical analyzes, paleomagnetic studies, and age determinations. All samples taken from the core are recorded on the Sample Registration form (Fig. F8) and entered into the ODIN database for tracking. The third column from the left on the VCD form is used to record the location of close-up photographs and, for the 1998 cruise, the location of shipboard XRF analyses. The far right column shows the unit designations and boundaries. We assign unit numbers sequentially from top to bottom using "I" to denote igneous units, "S" to denote sedimentary units, and "C" to denote cored basalt clasts.

When igneous basement was cored, unit boundaries are identified by the occurrence of major lithologic changes (i.e., color, texture, structure, grain size, mineralogic occurrences, and abundances). Most of these units are cooling-rate controlled and represent individual extrusive flows, intrusive sills, or dikes. Significant primary internal variations in grain size, phenocryst type, vesiculation, or other notable features within individual units are designated as subunits using a letter suffix (e.g., 1A and 1B). Secondary features (deformation and alteration) are not considered part of the unit designation.

The presence of glaciomarine sediment of variable thickness covering much of the basement on the East Greenland shelf made drilling with the BGS seabed platform particularly challenging. Often it was necessary to maintain a high flushing rate to keep the drill bit clear during advance through the sedimentary overburden. This meant that any soft mud matrix was washed from the hole leaving a concentrate of clasts. When the matrix was not recovered it was impossible to unequivocally classify the sediment. Of importance for our work was to establish whether the fragments recovered represented basement lithologies or exotic material. To maintain objectivity, we classify all clasts as sediments ("S" units), but we further note whether the collection of rock fragments is monolithic or polylithic. If all the fragments from the core barrel are composed of lithologically identical basalt, we referred to the sediments as "basaltic gravel" and described the materials in the same way as an igneous unit. This classification proved useful for relating basalt gravel to underlying cored material, basalt gravel from other holes at the same site, or material recovered from neighboring sites. If the fragmented basalt was large enough to be cored but we are not fully certain it represented in situ basement, we applied the unit prefix "C." This material is also described using igneous unit nomenclature. The "C" designation acknowledges the tenuous nature of this fragmented material but carries with it the implication that fractured basement was probably drilled. However, we only concluded that these basaltic clasts represent the local basement lithology when indistinguishable material was recovered from adjacent holes at the same site.

Thin sections were prepared on board using standard procedures. The dried rock billets (25 mm x 15 mm x 8 mm) are glued to glass plates using an Araldith epoxy and cured at 60°C. Sections are polished to a thickness of ~45 µm with a Logitech rotary lap using a mixture of ethylene glycol (to avoid sample swelling) and 15-µm alumina powder. Glass covers were omitted to permit future polishing for shorebased microprobe studies.

The visual description of igneous rock units was recorded on the Lithologic Unit Description form (Fig. F9) and entered into the ODIN database (see "ODIN Database"). A separate entry was made for every igneous lithologic unit of each core section. The required information falls into one of three categories: (1) summary information, (2) primary mineralogy, and (3) secondary features. The terminology used by the Shipboard Scientific Party and summarized in Table T1 is consistent with that used on ODP Legs 152 and 163.

Igneous rocks are classified according to the International Union of Geological Sciences scheme (Le Maitre, 1989), with modifiers applied to specify the phenocrystic minerals and their proportions. For example, if by visual inspection the rock contains 1% (by volume) phenocrysts, we precede the rock name (i.e., basalt) with the term "aphyric." If the phenocrysts are more abundant, we used the term "phyric" preceded by the names of phenocrystic minerals, arranged in order of decreasing abundance from left to right. Finally, we denoted the overall abundance of phenocrysts by the first modifiers, where "sparsely" corresponds to 1%–2% phenocrysts, "moderately" indicates 2%–10% phenocrysts, and "highly" refers to >10% phenocrysts. Therefore, a unit classified as a sparsely plagioclase-olivine phyric basalt contains <2% (by volume) phenocrysts with plagioclase in greater abundant than olivine.

The contacts between units were identified by a significant change in phenocryst proportion and assemblage, grain size, and/or texture. The contacts may either be preserved or not preserved. The forms of the contacts are described either as planar, irregular, angular, gradational, or not preserved. The following terms are used to describe the type of contact relationship preserved: extrusive, intrusive, depositional and erosional as defined by chilled zones and contacts, intercalated volcaniclastic, volcanic and sedimentary rocks, intrusive sills and dikes, and brecciated or scoriaceous zones.

Unconsolidated sediments on the East Greenland continental shelf are dominantly glaciomarine in origin and occur as overburden of variable thickness on the volcanic basement. Sites were chosen in order to minimize this sedimentary overburden. Because the diamond-bit coring system requires high flushing rates to keep the drill bit clean, the drilling strategy was designed to minimize recovery of the glaciomarine overburden. Consequently, undisturbed sedimentary units were rarely cored and recovered. Most often the soft mud matrix was washed out, leaving the coarser fractions and clasts. Given the variability in recovery, we devised a simple working classification scheme for shipboard description of the glaciomarine overburden. This classification scheme differs from the more comprehensive scheme used by ODP (e.g., Mazzullo and Graham, 1988), in that only four major types of unconsolidated material were distinguished: mud, diamicton, diamicton clasts, and basaltic gravel. We define these unconsolidated materials in the following ways:

• Mud: Sediment dominated by particles ranging in grain size from clay to silt.
• Diamicton: Sediment consisting of "sand and/or larger particles dispersed through a muddy matrix" (Flint et al., 1960).
• Diamicton clasts: Accumulation of pebble- to cobble-sized clasts representing different rock lithologies (including petrographically distinguishable basaltic material) derived from a muddy matrix and/or sand washed from the hole.
• Basaltic gravel: Accumulation of petrographically indistinguishable basaltic clasts.

The term "gravel" is used in its broadest sense to refer to a range of particle sizes from sand to pebble. Basaltic gravel may represent scree derived from mechanical weathering of the local basement or may be of unknown origin. The use of the phrase "basaltic gravel" acknowledges the sedimentary origin of the material ("S" unit) but offers the shipboard scientists the opportunity to describe the material in terms of its igneous features. This information was very useful in site selection and evaluation.

In addition to classification of sediments using the above scheme, the size, shape, and degree of sorting of rock fragments were also recorded. We adopt the Udden-Wentworth grain-size scale convention (Table T2). The roundness and sphericity of grains and clasts are described using the descriptive terms of Powers (1953), where "roundness" is very angular, angular, subangular, subrounded, rounded, or well-rounded, and "sphericity" is either low or high. The sorting nomenclature of Folk (1974) is used (Table T3).

Thin sections prepared on board were described using the terminology for the VCDs. A glossary of common terms is provided in Table T1. Thin section descriptions were initially recorded on Thin Section Description forms (Fig. F10) and entered into the ODIN database. Among the observations are the type, proportion, size, and habit of phenocrysts, and, where recognizable, the groundmass mineralogy and grain size. The proportion of mesostasis was also estimated, along with secondary minerals and the overall state of alteration. Thin section descriptions may differ from the VCD, and no attempt was made to revise the VCD forms once thin sections became available.

Each igneous unit recovered during the Norskald drilling in 1998 was analyzed for selected elements by XRF on the ship using a portable X-MET 920 probe. The probe was placed directly onto the core, either on flat, cut surfaces or on rounded core surfaces. Ti, Fe, Zr, Sr, Cr, and Ni were analyzed. Single-element intensities were measured for 300 s and calibrated against a collection of 20 basalt slabs from East Greenland on which XRF major element data from the GEUS XRF laboratory in Copenhagen, Denmark (Sørensen, 1975; Kystol and Larsen, 1999), were available (Table T4). The calibration converted major element intensities to oxide concentrations in weight percent (i.e., TiO₂ and FeO [all iron as FeO]) and trace element intensities to element concentrations in parts per million (i.e., Zr, Sr, Cr, and Ni). Calibrations used to estimate elemental concentrations from X-ray intensity are given in Table T5. Calibrations for Ni and Cr are not considered reliable. Drift was negligible as determined by repeated analysis of two East Greenland basalts (412481 and 404214). These same working standards were used to estimate analytical precision, which was found to be better than 11% for TiO₂, 4% for FeO, 9% for Zr, and 22% for Sr. Cr and Ni are poorly determined because of an inhomogeneous distribution of Cr- and Ni-rich phases and absorption of CrK radiation by Ni.

Shipboard paleomagnetic studies were undertaken onboard the Norskald in 1998 but not onboard the Aranda in 1999. Natural remnant magnetization (NRM) was measured on discrete specimens (10.8 cm³) subsampled from the working half of the core with a Molspin spinner magnetometer. The magnetic stability of selected specimens was determined by stepwise alternating-field (AF) demagnetization using a Molspin two-axis tumble demagnetizer. The AF demagnetization was performed in a stepwise manner with at least 10 steps to a maximum field of 100 mT. Both Zijderveld plots and equal-area stereographic projections were used to interpret the results. The characteristic remnant magnetization (ChRM) was isolated using principal component analysis (Kirschvink, 1980). No thermal demagnetization was performed. The results of shore-based NRM measurements for core recovered in both 1998 and 1999 are also presented in individual site chapters of this volume (see "Shore-Based Scientific Procedures").

Because the azimuth of the drill core is unknown, only the inclination and intensity of the remnance are meaningful. For each site, the inclination of the ChRM and NRM were plotted against depth and the sequence divided into magnetozones. Ideally, a magnetozone is defined by at least two successive igneous units for which the inclination clearly indicates normal/reversed polarity. The stability of NRM was confirmed by AF demagnetizations. Shifts in polarity that are observed only in single igneous units or units having intermediate inclination are ascribed to short-time geomagnetic excursions. Whenever possible, the magnetostratigraphy is combined with chemical stratigraphy for cross-correlation and as an aid for intrasite correlation. Because of the lack of other information besides the magnetic lineation of the ocean floor, a direct correlation of the magnetozones to the geomagnetic polarity timescale (Cande and Kent, 1995) is tenuous.