29. DATA REPORT: COLOR RECORDS FROM THE CALIFORNIA MARGIN: PROXY INDICATORS FOR SEDIMENT COMPOSITION AND CLIMATIC CHANGE¹

INTRODUCTION

The growing importance of high-resolution paleoceanographic studies in recent years has increased the need for methods that allow fast and easy compilation of long, detailed, and continuous time series. Sediment color is one property of the sediment that reflects chemical composition and, to some extent, physical properties, and as such can be used as a proxy for paleoceanographic studies. Composition of the sediment flux, and its preservation at the sediment/water interface, is controlled by climatic factors (e.g., weather, current patterns, surface water productivity, and water column stratification). Thus color variations offer a proxy to variations in the external factors that control sedimentation and thus a monitor of regional and (or) global climate change. Deep-sea sediments are particularly suitable for such an approach because their composition is relatively simple. This allows color changes to be correlated directly with variations in one or at most a few sediment components, when color measurements are calibrated against discrete geochemical and sedimentological analyses.

Quantitative color measurements are now routinely produced during cruises of the Ocean Drilling Program (ODP), to generate color time series that provide a means for correlating parallel holes. The Oregon State University spectral reflectometer (Mix et al., 1995) or the Minolta spectrophotometer allow the collection of discrete measurements that form an average over some surface area of sediment. In contrast, a much higher stratigraphic resolution can be obtained from digital images of the core surface, as each pixel in the image represents one measurement. The actual resolution depends on the camera and its distance from the sediment, but an effective resolution of one measurement per 100 µm, or better, can easily be obtained with modern cameras and computers. To test different color-data collection systems for suitability for use on board, a video-camera system was used during ODP Leg 167 for continuous imaging of all cores that were drilled (Lyle, Koizumi, Richter, et al., 1997). The ultra-high resolution time series that can be obtained from such images have been used to unravel the climatic history on annual to centennial scales represented in annually laminated sediments (Merrill and Beck, 1995; Schaaf and Thurow, 1995, 1998).

The video system used during Leg 167 is essentially the same as that described by Merrill and Beck (1995), who tested the system and developed software for image capture and processing. The system was brought aboard for further testing, to determine whether it warranted inclusion in the standard shipboard analytical tools. Video images of discrete, overlapping segments of sediment potentially offer a method for archiving of sedimentary structures, but in addition, high-resolution time series of color variation can be derived from these images. Automated processing and mosaicing of individual images containing a 20-cm segment of sediment allowed the compilation of a color time series while keeping up with the core flow (Lyle, Koizumi, Richter, et al., 1997). However, it was found that the procedures used to convert video color to continuous time series were not able to deal adequately with several practical problems that arose during the leg, the main two being:

1. The standard method (Rogers, 1985) that was used to convert the red, blue, and green (RGB) in the video images to color-coordinates more suitable for presentation (L*a*b* system defined by the Commission Internationale de l'Éclairage [CIE]) was found not to be adequate. Recalibration of the camera during the drilling of Site 1019 produced a change in calculated color output, resulting in an offset in values for sections drilled before and after the recalibration (Fig. 1).
2. Correction for nonuniform distribution of the illuminant was based on the pattern observed in images of a neutral gray card that was scanned regularly during the cruise. However, distribution of especially L* (lightness of the sediment) within an image is highly dependent on the reflectivity of a surface, and correction curves calculated from the gray card were too flat to handle the more steeply curved color data produced by wet (more reflective) sediment surfaces (Fig. 2).

It was therefore desirable to reprocess the images to obtain better calibrated color time series. Here we describe modified methods for processing the image data and present the updated color records.

¹Lyle, M., Koizumi, I., Richter, C., and Moore, T.C., Jr. (Eds.), 2000. Proc. ODP, Sci. Results, 167 [Online]. Available from World Wide Web: <http://www-odp.tamu.edu/publications/167_SR/167sr.htm>. [Cited YYYY-MM-DD]

²Department of Geological Sciences, University College London, Gower Street, London WC1E 6BT, United Kingdom. Correspondence author: j.thurow@ucl.ac.uk

³Ocean Drilling Program, Texas A&M University, 1000 Discovery Drive, College Station TX 77845-9547, USA.

⁴Present address: St. Lawrence University, 22 Romoda Drive, Canton NY 13617, USA.

Date of initial receipt: 28 October 1998
Date of acceptance: 19 April 1999
Ms 167SR-236