Plug samples of 1 in diameter and varying length (Tables T1, T2) were drilled from the working half of slabbed cores on board the JOIDES Resolution. Horizontal plugs were selected by S. Ehrenberg, and vertical plugs were selected by G. Eberli for their respective research projects. A further 23 whole-core samples were selected by S. Ehrenberg from Site 1196 and 1199 cores (Table T3). Sampling was limited to intervals with sufficient recovery of core having a coherent state of preservation suitable for drilling of a plug sample. Efforts were made to collect both horizontal and vertical plugs from nearby depths in as many places as possible, but many intervals were suitable for drilling only a single plug sample.
Plug samples from the platform top sites were analyzed at Reservoir Laboratories AS, Stavanger, Norway, and whole-core samples were analyzed at Reservoir Laboratories AS (RESLAB), Trondheim, Norway. Analytical techniques are described in reports provided by Reservoir Laboratories AS (see the "Supplementary Material" contents list). Porosity and grain density were calculated from sample weight, grain volume measured by helium injection using a Boyle's law porosimeter, and bulk volume calculated from length and diameter measured by caliper. Permeability was calculated using Darcy's law from the pressure decrease of flowing nitrogen across the plug length (20-bar confining pressure). A total of 79 Samples from slope and basin sites were measured at Terra-Tek, Salt Lake City, Utah, USA, using the same methods as for the platform samples. Permeability values reported by RESLAB as >50,000 Md have been set equal to 50,000 mD in Tables T1 and T2.
It was suspected that porosity values calculated using plug volume measured by caliper (PORC in Table T1) might be overestimated for a limited number of samples due to irregular plug shape. Some shape irregularities reflect poor lithification of granular limestone having little cement content, whereas other cases reflect the presence of larger vugs or rhodoliths. To evaluate the magnitude of such problems, porosity was recalculated for 46 horizontal plugs using plug volume measured by mercury displacement (PORM). The selected 46 plugs include 37 plugs with apparent shape irregularity or suspected error in PORC, 2 plugs missing PORC values, and 7 plugs with apparently good cylindrical shape (run as controls).
In general, PORM values are expected to be lower than "true" porosity values because the mercury will fill any indentations along plug walls larger than a certain minimum radius, and many such indentations will be part of the natural pore system of the sample. For the control samples (good cylindrical plug shape), it is assumed that PORC is equal to true porosity. These samples were measured by mercury displacement to provide an estimate of how much PORM needs to be increased to provide the best estimate of true porosity.
Plots of the results (Figs. F1, F2) show that PORC values of the control samples are roughly 7% higher than PORM (relative to PORM), with highest weighting given to the most porous samples. Based on this estimate, porosities of plugs with suspect or missing PORC values were corrected to 107% of the PORM value (except for five plugs having PORC < 107% of PORM, in which case the original PORC value was accepted). Corrected porosity values for the 34 plugs affected are shown in bold font in column "PORX" of Table T1.
Although this correction is clearly subject to major uncertainty, it is felt that the corrected values provide a more realistic estimate of the true porosities of these plugs than the original PORC values. An alternative available to anyone wishing to use this data set is to exclude the corrected porosity values. The 34 porosity values concerned are marked in bold font in Table T1 to facilitate this alternative. However, it is not recommended to use the original, uncorrected PORC values for these 34 samples.
After plug analyses were completed, thin sections were prepared from 139 horizontal plugs (25 mm x 47 mm glass slides) and all 23 whole-core samples (50 mm x 75 mm glass slides). Thin sections for these horizontal plugs were made by Independent Petrographic Services (United Kingdom). Thin sections for the vertical plugs were made by Spectrum Petrographics (Winston, Oregon, USA). Thin sections of horizontal plugs and whole-core samples were impregnated with blue epoxy and polished, and part of the area was stained for carbonate mineral identification by the method of Dickson (1966). A representative area (1.0 cm x 1.5 cm) of each horizontal plug and whole-core sample thin section was photographed at standard magnification using a digital camera mounted on a Leica MZ125 stereomicroscope (see "Appendix A" and "Appendix B").
Each sample was assigned to one of three carbonate mineralogic categories: limestone (L), dolostone (D), or mixed (M) partly dolomitized limestone (Tables T1, T2, T3), based on available information. For samples where thin sections were available, petrographic examination was used to evaluate the proportions of calcite and dolomite present. Samples with a ratio of dolomite/(dolomite + calcite) of 0.2–0.8 were are assigned to the "mixed" category. Samples for which thin sections were not available were assigned a mineralogic category based on the shipboard core descriptions and X-ray diffraction data, together with macroscopic examination of the sample.