Paleoceanographic significance of sediment color on western North Atlantic drifts: I. Origin of color

Abstract

Reflectance spectra collected during ODP Leg 172 were used in concert with solid phase iron chemistry, carbonate content, and organic carbon content measurements to evaluate the agents responsible for setting the color in sediments. Factor analysis has proved a valuable and rapid technique to detect the local and regional primary factors that influence sediment color. On the western North Atlantic drifts, sediment color is the result of primary mineralogy as well as diagenetic changes. Sediment lightness is controlled by the carbonate content while the hue is primarily due to the presence of hematite and Fe²⁺/Fe³⁺ changes in clay minerals. Hematite, most likely derived from the Permo-Carboniferous red beds of the Canadian Maritimes, is differentially preserved at various sites due to differences in reductive diagenesis and dilution by other sedimentary components. Various intensities for diagenesis result from changes in organic carbon content, sedimentation rates, and H₂S production via anaerobic methane oxidation. Iron monosulfides occur extensively at all high sedimentation sites especially in glacial periods suggesting increased high terrigenous flux and/or increased reactive iron flux in glacials.

Introduction

Color, which is a readily observable physical property of rocks or sediments, has long been considered of diagnostic and correlative value in geology. It has been acknowledged that the sediment chromatic characteristics are generally given by its iron-rich minerals: oxyhydroxides, sulfides, iron-rich clay minerals (e.g., Potter et al., 1980). Carbonate and opal increase the sediment brightness while organic matter decreases it (e.g., Balsam and Deaton, 1991, Balsam and Deaton, 1996, Mix et al., 1995). Other mineralogical constituents may also exert secondary or local influence on color reflectance (e.g., Mix et al., 1995, Balsam and Deaton, 1996). In fresh sediments, color is an ephemeral characteristic, changing rapidly after exposure to oxidizing conditions, and due to drying after recovery.

For geological studies, the Munsell Color system has been traditionally used to describe sediment color. The system is a qualitative ordinal color scale dependent on visual comparison of the sample to color chips (Goddard et al., 1948). The color chips describe a three-dimensional space of hue, value, and chroma. The hue characterizes the property of being yellow, red, green, blue, purple, or any binary mixture thereof. The value describes the color lightness relative to a gray scale ranging from white to black, while the chroma corresponds to the degree of difference between a color and a gray of the same value or lightness. Data recorded in this system cannot be easily manipulated mathematically and vary in quality as observers and observational conditions change. Also color differences are too subtle and abundant for detailed work to be accomplished only by visual inspection.

Alternative quantitative approaches to the Munsell system that have been used to describe the sediment color include CIE La*b* chromaticity space (e.g., Nagao and Nakashima, 1992, Saito, 1995, Weber, 1998) and CIE x-y-Y color space (e.g., Merrill and Beck, 1996), which are both reductions of the reflectance spectra taken in the visible wavelength domain. The CIE La*b*, used frequently in ODP color estimation, is a three-dimensional color space which approximates a cylinder (Billmeyer and Saltzman, 1981). Its vertical axis L (lightness) is similar to the Munsell value, while a* and b* are the Cartesian coordinates that define a quasi-circular hue-chroma space (i.e., a normal section of the cylinder) at any given L. If one uses polar coordinates for the La*b* space, the hue (H) is more naturally defined as an angle measured counterclockwise from east, while chroma (C) is the length of the radius within a section of the cylinder. In this presentation, an angle of 45° corresponds roughly to a red hue, 90° to yellow, 180° to green, and 240° corresponds to blue. Chroma increases to the exterior of the cylindrical color space. Simple transforms are used to convert La*b* coordinates to LCH coordinates:

$$\text{C=((a}{}^{\mathrm{*}}\text{)}{}^2\text{+(b}{}^{\mathrm{*}}\text{)}{}^2\text{)}{}^{\mathrm{1/2}}$$

$$\text{H=}\text{tan}{}^{\mathrm{-1}}\text{(b}{}^{\mathrm{*}}\text{/a}{}^{\mathrm{*}}\text{) (}\text{degrees}\text{),}\text{0°≤H≤360°}$$

Other geological applications utilized directly the reflectance spectrum data in the visible range (sometimes extended to near-ultraviolet and near-infrared; e.g., Balsam and Deaton, 1991, Mix et al., 1995, Harris et al., 1997, Ortiz et al., 1999). By using the spectral signatures of iron oxides, calcium carbonate, opal, clay minerals, organic matter, and other sedimentary components, it is possible in some cases to empirically quantify their presence in sediments (e.g., Balsam and Deaton, 1991, Mix et al., 1995). Factor analysis and multiple regression have been employed in deconvolving the sediment reflectance spectra. The possibility of rapid estimation of sediment composition at a high resolution is the main advantage of using sediment reflectance spectra in paleoceanographic studies. However rigorous quantitative calibrations are difficult to obtain since the spectral characteristics of individual components are subjected to a matrix effect (Balsam and Deaton, 1991). Also the original color of the sediments at the time of deposition is likely to differ from their color after the early diagenesis or the final color of the lithified rock.

In this paper, we present results from a study that investigated the origin of color in sediments drilled on two sediment drifts in the western North Atlantic: Blake–Bahama Outer Ridge and Northeast Bermuda Rise (Fig. 1; referred to as BBOR and BR in following discussion). Knowledge of processes responsible for setting the color in sediments is necessary in order to evaluate color as a potential paleoceanographic proxy (see the companion paper of Giosan et al., 2002). Sediments recovered during Leg 172 in the western North Atlantic had a variegated colorful character. Similar to other sites in the Atlantic basin (e.g., Schneider et al., 1995), a strong correlation between luminance (L) and carbonate content has been observed for the entire column of Pliocene to recent sediments. The presence of hematite (and possibly other iron hydroxides) imparted a reddish hue to sediments occurring mostly in glacial–deglacial intervals especially at deeper sites on BBOR and BR. They are the so-called ‘brick red lutites’ (Fig. 1), which are considered to be characteristic tracers of northern sediment sources (Hollister and Heezen, 1972, Piper et al., 1994). Thought to originate in the Canadian Maritime Provinces, which have extensive Permo-Carboniferous red beds exposed (Fig. 1), the red lutites were one of the first direct pieces of evidence for the existence of a deep western boundary current along the eastern continental margin of North America (Needham et al., 1969). Other sediments exhibited a greenish hue suggesting that iron is present in a reduced state in clay minerals (e.g., König et al., 1999). Also related to reduction, bluish-black metastable iron monosulfides occurred at Leg 172 deep sites (Keigwin et al., 1998). Therefore in order to understand the color variations at Leg 172 sites and interpret them as paleoceanographic changes, we need to know both the role of the primary detrital and pelagic sediment delivery on sediment color as well as the role of diagenesis in promoting secondary changes in sediment color.