RESULTS

Data used for the study presented here are given in Table T1. Spectral reflectance data measured on board for each hole are given in Tables T2, T3, T4, T5, T6, T7, T8, T9, T10, T11, T12, T13, T14, and T15.

For the investigations performed, the data quality was controlled by the ability of the CM 2002 to convert the spectral signal into color data (CIELAB parameters and/or Munsell color). Spectral distribution measurements that were not convertible to color values by the CM 2002, due to cracks or voids in the measured section, were removed from the data set. Also, all core catcher measurements were removed from the data set to gain a data set comparable to other physical properties data. Typically, <5% of the data was removed.

Results for Hole 1098C demonstrate that the spectrum of reflectance data (wavelength window = 7.0-6.5 µm) of samples documents systematic differences in signal absorption (Fig. F1); signal absorption is higher in the finely laminated intervals and lower in intervals reported to be turbidites.

As a second step, the variances (s²) of the reflectance data at wavelengths between 7.0 and 6.5 µm were calculated to get a proxy for the degree of signal absorption by using Equation 1, as follows:

(1)

Results show that high variance values reflect sediment intervals that are finely laminated and related to a low depositional energy level, whereas low variance values reflect turbiditic sediment intervals related to a high depositional energy level (Fig. F2). Differences in variance values are small, but they are consistent and correlate with the two end-members of the depositional environments. They are therefore able to characterize sediments deposited at different deposition energy levels vs. depth, as shown for Hole 1098C in Figure F2.

The reasons for these differences are not fully understood at the moment. Its known that finer grain sizes have a lower absorption than coarse grain sizes at wavelengths between 7.0 and 6.5 µm (e.g. Clark, 1999). It is also obvious that even small changes in the direction of lower porosity values will lead to a lower absorption, and vice versa.

In this case, two factors must be taken into account during intervals of high deposition energy levels (turbidite facies): (1) the sorting of turbidites is generally low, and (2) because of the higher deposition energy intervals, turbidite sequences consist of a broader grain-size spectrum and are likely to be more compacted during sedimentation, resulting in a slight reduction in porosity. Both factors will result in low variance values of the measured reflectance in the wavelength window (7.0-6.5 µm). On the other hand, for intervals of lower deposition energy levels (finely laminated facies), one would expect good sorting into certain grain-size fractions and a lower degree of compaction during the sedimentation process. As a result, variance values will be high.

In summary, diffuse spectral reflectance data provide valuable information about sediment deposition. This data analysis shows that on a high-resolution scale, it documents changes in the sedimentation environment as recorded in the two different facies types in Hole 1098C.