Marine sediment components: identification and dispersal assessed …€¦ · seismic-facies analysis; and geologic evolution of passive margins. He has published more than 55 papers

Int. J. Environment and Health, Vol. 1, No. 3, 2007 403

Copyright © 2007 Inderscience Enterprises Ltd.

Marine sediment components: identification and dispersal assessed by diffuse reflectance spectrophotometry

William Balsam* and John E. Damuth Department of Earth and Environmental Sciences, University of Texas at Arlington, Arlington, TX 76019, USA Fax: +1 817 272 2628 E-mail: [email protected] E-mail: [email protected] *Corresponding author

Bobby Deaton Department of Physics, Texas Wesleyan University, Fort Worth, TX 76105, USA Fax: +1 817 531 4202 E-mail: [email protected]

Abstract: Analysis of deep-sea sediments presents a number of exceptional challenges. Generally, the amount of material available is small and has to be shared among investigators. Also, deep-sea sediment contains a diverse range of components from a variety of sources including lithogenous, biogenous and hydrogenous. Finally, large number of samples is frequently analysed making rapid analysis a necessity. Given these limitations, a technique that is rapid, non-destructive or uses a very little sample, and is capable of identifying a large number of sediment components with a single analysis would be ideal. Diffuse Reflectance Spectrophotometry (DRS) has the potential to meet many of these criteria. In this paper, we discuss the application of DRS to deep-sea sediments from a variety of marine settings. We examine the modern sediment distribution in the Atlantic Ocean, Gulf of Mexico and Argentine Basin and compare the distribution of modern with the last glacial maximum sediment in the Atlantic Ocean. DRS has also proven useful in times series studies. DRS is especially useful for producing time series from Ocean Drilling Program (ODP) holes that require thousands of samples to produce a single time series. In this paper, time series from both the Atlantic and Pacific Oceans are examined.

Keywords: Argentine Basin; Atlantic Ocean; atmospheric dust; Diffuse Reflectance Spectrophotometry; DRS; Gulf of Mexico; iron oxides; marine sediment dispersal; marine sediment distributions; Ocean Drilling Programme; ODP; paleoclimatology.

Reference to this paper should be made as follows: Balsam, W., Damuth, J.E. and Deaton, B. (2007) ‘Marine sediment components: identification and dispersal assessed by diffuse reflectance spectrophotometry’, Int. J. Environment and Health, Vol. 1, No. 3, pp.403–426.

404 W. Balsam, J.E. Damuth and B. Deaton

Biographical notes: William Balsam is a Professor in the Department of Earth and Environmental Sciences at the University of Texas at Arlington. He has co-authored more than 60 research papers in a variety of fields including paleontology, marine geology, paleoclimatology, geoarchaeology, and oil and gas exploration. Since the early 1990s, his primary source of research data has been reflectance spectroscopy. He has used diffuse reflectance spectra to map ocean bottom sediment composition, determine thermal alteration of conodonts and sediments under basalt flows, trace the movement of eolian dust into the oceans, and help interpret paleoclimatic variations in the Chinese loess sequence.

John E. Damuth is an Adjunct Professor in the Department of Earth and Environmental Sciences at the University of Texas at Arlington. He has previously worked both in academic research (Lamont-Doherty Geological Observatory of Columbia University) and in industry (Mobil Research and Development Corp,). His research emphasises the study of deep-water clastic depositional systems including the morphology and depositional processes of deep-sea deposits and their bed forms; development of submarine fans and related features; glacio-eustatic control of deep-sea sediment deposition; seismic-facies analysis; and geologic evolution of passive margins. He has published more than 55 papers in refereed journals. His most recent major research is the Gulf Intraslope Basins Project (GIB), which is sponsored by industry, and studies late Quaternary depositional processes in intraslope basins in order to develop more reliable depositional analogues for hydrocarbon reservoirs.

Bobby C. Deaton is a Professor of Physics at the Texas Wesleyan University. His current research interests are concentrated on the study of geological materials using spectrophotometric analyses, primarily in the visible region of the electromagnetic spectrum. He is the author of over 50 publications in the fields of solid state physics, plate tectonics and applications of total reflectance to a wide variety of geological problems.

1 Introduction

Marine sediments present a special set of analytical problems. Typically, little material is available for analysis; the amount of sample available can range from < 1 to 6 or 7 g. As a result, techniques that are non-destructive or require little sample for analyses are favoured. In addition, the typical marine sediment is a mixture of sediments from many sources, although this mixture is frequently dominated by a few types of clay minerals transported from land and by biogenic carbonate or silica produced at or near the sea surface. Hence, any technique that seeks to determine the composition of marine sediments must be able to identify not only dominant components, but also ancillary components that are frequently present in low concentrations and may contain significant information about sediment source, transport and dispersal. Finally, when analysing marine cores, large numbers of samples, frequently in the thousands, may have to be analysed. As a result, analytical techniques must be rapid to accommodate the large number of samples commonly measured. In this paper, we describe the results of a technique that meets many of these criteria, Diffuse Reflectance Spectrophotometry (DRS).

Marine sediment components: identification and dispersal assessed by DRS 405

Spectrophotometry can be performed in a number of ways. Chemists commonly use spectrophotometry to identify substances in a liquid through absorption. With absorption, spectrophotometry, a light beam of known characteristics, is passed through a cuvette and compared with a reference light beam. Absorption is recorded as attenuation of light as a function of wavelength. This technique is useful because absorption wavelengths are frequently unique to a particular substance. For geological materials, absorption spectrophotometry is not generally used because of the difficulties involved in getting many materials into solution. As a result, silica and many silicate minerals would have to be dissolved in HF to be thoroughly dispersed in a solution. Instead of absorption spectrophotometry, geologists and soil scientists have tried a different approach, DRS. With DRS solid surfaces are analysed. These surfaces may be natural, as in the case of a rock, soil or deep-sea core or they may be produced by grinding a sample and placing it on a glass microslide. Hence, because DRS can analyse a solid surface it has the potential to be non-destructive.

As with any technique, DRS has a number of advantages and disadvantages. Advantages include modestly priced instrumentation, speed, the ability to identify both crystalline and amorphous materials and the ability to identify some materials at extremely low concentrations. Many NUV/VIS/NIR (Near Ultra-Violet/Visible/Near Infrared) spectrophotometers, either bench top or portable, can be purchased for ~$20,000. Analysis is reasonably fast, a few seconds per sample for a portable machine to about 2 min per sample for a bench top machine. Unlike X-Ray Diffraction (XRD) that depends upon internal reflections from crystalline surfaces, DRS only requires that the materials analysed be solids. Solid materials may include both minerals and amorphous materials such as glass and some organic compounds. Materials that are strongly coloured, like the iron oxide/oxy-hydroxide minerals hematite and goethite, can generally be identified at very low concentrations. For example, hematite can be identified at a concentration as low as 0.01% by weight (Deaton and Balsam, 1991).

DRS has two main disadvantages. First, NUV/VIS/NIR DRS is not particularly sensitive to materials that have a spectrum with very little structure, that is, a flat spectrum. In practice this means white, grey or black materials, although some of these materials may have unique features in the NUV or NIR. Secondly, spectra are influenced by the mixture of materials they contain. As with XRD spectra, one has to be aware of this ‘background effect’ when analysing for a specific compound. Umixing a material’s spectral signal from the background can be difficult and may require special statistical techniques such as factor or cluster analysis.

In this paper, we demonstrate the utility of DRS for identifying various components in marine sediments, especially deep-water sediments, and review two different types of analyses. First, we examine studies that have utilised DRS to identify and map sea-bed sediment components within a single time horizon, primarily modern sediment distributions. From these modern sediment distributions, it is usually possible, using a distinctive sediment from a point source, to track sediment movement within an ocean basin. Secondly, we examine temporal patterns in changing sediment composition by analysing a series of down-core deep-sea sediment samples. The changing patterns observed may be related to sediment provenance and/or climate, especially climate changes associated with glacial interglacial cycles.


2 Previous work

Several techniques other than DRS have been used to map and identify marine sediment components (see for example Biscaye, 1965; McCoy, 1983) and suggest causes for sediment distributions and identify possible dispersal paths. Techniques include smear slides, XRD, and carbonate analysis. Smear slide analysis has the advantage of being rapid and requiring very little material, but lacks precision because sediment components are usually classified as to their relative abundance (rare, present, common, abundant, etc.). In addition, whereas smear slide analysis is useful for identifying various biogenic components, it provides limited information about the lithogenous fraction other than grain size, with the exception that larger particles such as sand grains or foraminifer tests are commonly underrepresented. Because the lithogenous fraction may provide important information about sediment dispersal, this is an important shortcoming. Smear slide analysis has been used to map modern sediments in the Pacific Ocean (McCoy, 1983), Atlantic Ocean (Balsam and McCoy, 1987) and Gulf of Mexico (Balsam and Beeson, 2003). In addition to mapping modern sediments, smear slide analysis is routinely used by sedimentologists on Ocean Drilling Program (ODP) ships to classify sediments. When applied to ODP cores, smear slide analysis provides a rapid, gross description of changes in sediment composition and has been routinely used by shipboard sedimentologists during expeditions of the Integrated ODP (IODP) and its predecessors (ODP; Deep Sea Drilling Project, DSDP).

For many years, researchers have been analysing marine sediments by XRD (see for example seminal articles by Pinsak and Murray, 1958; Griffen, 1962; Biscaye, 1965). These studies have generally concentrated on clay minerals. Results indicate that the dominant clay mineral in the marine environment is illite, but that climate plays an important role in the distribution of clay minerals in the marine environment. Where physical weathering dominates, chlorite is common; where chemical weathering dominates, kaolinite is common. Chlorite is transported to the oceans mainly by glacial action and is input over a large area, so it is of limited usefulness in tracing sediment dispersal. However, in some areas, the Gulf of Mexico (Balsam and Beeson, 2003) and the Rio Grande Rise in the South Atlantic Ocean (Jones, 1984), kaolinite may be a useful tracer of fluvial influx because it is introduced from a limited number of sources.

Most calcium carbonate in deep-sea sediment is biogenic in origin and was produced by planktonic foraminifera and coccoliths slowly raining down onto the ocean floor. Although productivity has some control on carbonate distributions, the primary controls are

1 dilution by terrigenous sources

2 dissolution by deep water charged with CO2.

If dissolution can be factored out, carbonate content can provide a rough measure of dilution by terrigenous sediment components. Maps show carbonate content is generally highest (frequently >80% by weight) on ridges, rises and seamounts isolated from continents, and lowest in deep basins and continental margins (continental slope and rise, abyssal plains, deep-sea fans; Balsam and McCoy, 1987) where it is effected by both dilution and dissolution. Along continental margins, much of the sediment is a mixture of terrigenous material of fluvial origin and carbonate of pelagic origin (depending upon depositional processes), and is commonly termed hemipelagic sediment. Although most


carbonate in hemipelagic sediment appears to be of pelagic origin, at least one study (Balsam and Williams, 1993) suggests that by analysing the O18/O16 of the <63 µm carbonate fraction, it is possible to identify carbonate derived from land. This is possible because carbonate from terrestrial sources has been modified by dissolution and reprecipitation and has very low O18/O16 values when compared with marine plankton. On land, carbonate is usually dissolved and transported in solution. However, in environments where physical erosion dominates, for example glacial erosion, carbonate may be transported as individual grains to the deep-sea. Hence, by mapping the O18/O16of the < 63 µm carbonate fraction of marine sediment it should be possible to

trace the dispersal path of glacially derived carbonate sediment.

3 Methodology

3.1 Instrumentation

The studies described in this paper employ both bench-top and portable spectrophotometers utilised in a total reflectance mode. With portable instruments like the Minolta CM-2002 (Konica Minolta, Mahwah, NJ), the surface of a sample may be scanned without actually disturbing or removing any sample. The CM-2002 operates only in the visible part of the EM spectrum, that is from 400 to 700 nm. This instrument has been used by ODP and in some of the studies reported here conducted in conjunction with ODP. For bench top instruments, only a small amount of sample, as little as 0.1 g, is required for analysis. The bench top instrument we used for the studies described here is a Perkin–Elmer Lambda 6 (Perkin–Elmer Corp., Norwalk, CT) NUV/VIS/NIR spectrophotometer with a wavelength range of 250–850 nm. With this laboratory grade instrument, it is possible to analyse a sample about every 2 min; for a portable spectrophotometer, like the Minolta CM-2002, samples can be analysed as fast as the instrument can be positioned on the surface. (For a detailed comparison of portable and laboratory spectrophotometers, see Balsam, Damuth and Schneider, 1997, and Balsam and Damuth, 2000).

3.2 Sample preparation

For the portable spectrophotometer, little sample preparation is required. The sample surface must be smooth, so that the instrument aperture can be placed on it with minimal light leakage. In addition, if the surface is wet, it should be covered with a neutral (absorbs little light and absorbs uniformly across the spectrum) plastic film like Glad Cling Wrap™ (Balsam, Damuth and Schneider, 1997). Although samples can be analysed wet, water content affects the results, primarily by muting the spectral signal compared with dry samples (Balsam, Deaton and Damuth, 1998).

For analyses using a laboratory (bench top) spectrophotometer, samples are usually measured dry. This is done by taking a small amount (as little as 0.1 g, but preferably ~0.25 g) of sample ground to < 38 µm and preparing a slurry on a glass microslide with distilled water. The preparation should be thick enough so that light cannot shine through it. The glass microslide with the sediment slurry is then allowed to dry in a low temperature (< 40°C) oven or at room temperature. The resulting slide resembles a thick XRD slide. Grinding the sample and making the slide takes about 5 min, not counting the drying time. If the sample contains a significant amount of organic matter or smectite


clays that absorb water, we use spectroscopic grade propanol instead of distilled water to prevent cracking as the sediment dries and to disperse the clays and organic material. Details of sample preparation for the laboratory grade spectrophotometer can be found in Balsam and Deaton (1991).

3.3 Spectral data analysis

Reflectance spectrophotometers usually record information as percent reflectance relative to some pure white standard. For both the Lambda 6 and Minolta CM-2002, we use barium sulfate as a standard. The Minolta CM-2002 portable instrument uses a xenon-pulsed light source, whereas the Perkin Elmer Lambda 6 uses a deuterium light source for wavelengths from 250 to 350 nm and a tungsten filament light source for wavelengths from 350 to 850 nm. Differences in light sources could potentially bias results, although calibration to a standard should theoretically remove differences. The CM-2002 has a minimum wavelength sampling interval of 10 nm, whereas the Lambda 6 can sample at 1 nm intervals. For the CM-2002, we used the 10 nm sampling interval for data processing; for the Lambda 6 spectral data were recorded at 1 nm intervals but, for most of the studies reported here, processed at 10 nm intervals. Data were all saved to magnetic media for later processing.

There are no standard methods for analysing DRS data. In this paper, we base much of our discussion on a method that involves the calculation of first-derivative values. Within the wavelength range we analysed, raw or untransformed percent reflectance data produce spectral curves that generally rise monotonically toward longer wavelengths (Figure 1A and B). As a result, these curves are difficult to compare because they are smooth and change gradually. One method for making the curves easier to analyse and compare is to calculate changes in the slope of the curve by taking the first-derivative of the percent reflectance curve (Barranco, Balsam and Deaton, 1989; Balsam and Deaton, 1991; Deaton and Balsam, 1991). First-derivative values are high wherever the slope of the percent reflectance curve changes rapidly. For much of the data presented in this paper, we calculated first-derivative values at a 10-nm sample interval from 400 to 700 nm for the CM-2002 and 250–850 nm for the Lamba 6 instrument. The resulting values obtained from the analysis are then plotted at the midpoint of each 10-nm interval (i.e. the 400–410 value is plotted at 405 nm, etc.) and are expressed as percent per nm (Figure 1A and B). Barranco, Balsam and Deaton (1989), Deaton and Balsam (1991) and Balsam and Deaton (1991, 1996) have shown that the position of peaks on the first-derivative curves is indicative of sediment composition and mineralogy.

Although first-derivative curves provide substantial compositional information about many samples, they are not useful for samples with uniform reflectance across the wavelength range of the analysis. For example, white materials such as carbonate, opaline silica or quartz routinely exhibit high reflectance in the entire VIS (that is why the eye perceives these materials as white). First-derivative curves for these white materials exhibit little change in the VIS, although reflectance changes for these materials in the NUV and NIR (Gaffey, 1985; Herbert, Tom and Burnett, 1992) are diagnostic, that is, allow their identification. Similarly, very dark compounds that exhibit consistently low values throughout the VIS and therefore exhibit few first-derivative peaks are unlikely to produce curves diagnostic of those compounds. While DRS is a useful technique for identifying many materials, it is not likely to be a useful technique for distinguishing materials that are predominantly white, grey or black.


Figure 1 Diagram showing the relationship between percent reflectance and first-derivative curves

Note: how the first-derivative curves contain peaks whereas the percent reflectance curves do not. These peaks have been related to different minerals (Deaton and Balsam, 1991; Balsam and Deaton, 1996) making the first-derivative curves easier to interpret. (A) V22-177 is outside the African dust plume and contains no hematite; (B) V23-100 was taken from within the African dust plume in the Atlantic Ocean and clearly contains hematite based on the position of the major first-derivative peak (see Figure 2).

4 Case studies

Here, we employ the case study approach to demonstrate how DRS can be used in the marine environment. These case studies are divided into time-slice and time-series studies. Time-slice studies display the distribution of sediment types at one time, that is at a single synoptic level. The most commonly used time-slice is the present, and a number of maps of modern sediment distribution based on DRS will be presented. The Last


Glacial Maximum (LGM) is another common time-slice and we present one map of LGM sediment distribution. Time-slices provide the information necessary to map sediment distributions and to compare changes in sediment distributions at different times. However, they provide no information on changes between the time-slices. Time-series studies provide a temporal record of changes at one geographic locality. These temporal records often hold clues as to the reasons for changes, especially when combined with time-slice studies. Time-series and time-slices are not mutually exclusive. Time-slices other than today have to be based on numerous time-series studies in which stratigraphy is used to established a particular time level.

4.1 Time-slice studies/mapping sediment distribution

4.1.1 Surface sediment distribution

4.1.1.1 Atlantic Ocean

The first and only ocean to be mapped in its entirety using DRS is the Atlantic Ocean. In 1991, Balsam and Deaton published a paper showing the distribution of sediment types as determined from DRS. In that paper, factor analysis was used to unmix spectra from 178 surface samples distributed throughout the North and South Atlantic Oceans. They extracted six factors from analysis of VIS spectra and related those factors (in order to decrease importance) to:

1 pelagic vs. terrigenous influx (largely carbonate content)

2 goethite

3 hematite

4 smectite clays

5 sediment organic content

6 chlorite.

It is important to note that these results do not indicate importance in terms of weight concentration in the sediments; rather, they indicate the importance of the components relative to their effect on the VIS spectra. For many spectrally identified sediment components, hematite for example, their effect on the spectrum is far greater than their weight concentration would suggest. Only a small amount of hematite (<0.1% by weight) can colour a sediment red (Deaton and Balsam, 1991).

A good example of how these factors are interpreted and assessed is shown by factor 2, the goethite factor. First, the factor pattern (loading) curve which displays how important each wavelength is in each factor is plotted (Figure 2A). Then the factor pattern curve is compared with a similar curve of a known material. Because the factors were calculated from first derivatives they are interpreted by comparison to first-derivative curves of a known compound. The loading curve for factor 2 exhibits a pattern which is similar to 0.5% goethite in a matrix synthesised to mimic typical North Atlantic deep-sea sediments (Figure 2A). Finally, the factor scores, which describe how important a factor is in each sample, are plotted. For time-slices, these plots are geographically distributed, thereby producing a map of the factor. For factor 2 this map (Figure 3) exhibits high values off the Amazon River. This is consistent with the high humidity


conditions known to produce goethite as opposed to hematite (Schwertmann, 1971, 1988; Balsam, Ji and Chen, 2004). On the other hand, hematite (factor 3, Figure 2B) appears to indicate input from the Sahara Desert (discussed below under LGM sediment distribution), a dry, windy region and is characterised by a large lobe of high values off the coast of northwest Africa. Eolian transport of hematite from the Sahara has been extensively discussed by Balsam, Otto-Bliesner and Deaton (1995) and Balsam et al. (2007).

Figure 2 Examples from the Atlantic illustrating how factors are interpreted

Note: (A) The factor 2 loading curve from Atlantic core tops compared with the first-derivative curve of 0.5% goethite in a matrix of material designed to simulate typical western North Atlantic sediment (Deaton and Balsam, 1991). The similarity between the factor loading curve and the first-derivative curve suggests that this factor is recording primarily goethite. (B) The factor 3 loading curve from Atlantic core tops compared with the first-derivative curve of 0.03% hematite in a complex matrix containing clays, quartz and feldspar. The similarity between the factor loading curve and the first-derivative curve suggests that this factor is recording primarily hematite.

Source: After Balsam and Deaton, (1991)


Figure 3 Map showing the distribution of factor 2 scores

Note: The dashed pattern indicates where goethite is important in this negative factor. This map is consistent with the conditions known to influence goethite formation, that is the warm humid conditions of the Amazon River drainage basin (see text for details).

Source: After Balsam and Deaton, (1991)

4.1.1.2 Argentine Basin

Using 236 samples, Balsam and Wolhart (1993) performed a detailed spectral analysis of Argentine Basin surface sediments. They extracted four factors from the first derivatives of the spectral data. Those four factors were interpreted as:

1 goethite

2 pelagic vs. terrigenous sedimentation

3 sediment organic content

4 chlorite.

Two of the factors, goethite and chlorite, appeared useful for identifying sediment sources and tracing sediment dispersal paths. Chlorite, probably produced by physical weathering in the Antarctic, was transported into the Argentine Basin by Antarctic Bottom Water. Secondary sources of chlorite and the primary source of goethite include the Falkland Plateau and Argentine margin. On the Argentine margin, it is clear that both goethite and chlorite are related to river transport across the shelf. From these areas goethite and chlorite differ in their methods of transport. Chlorite is dispersed normal to


the slope with one lobe going from the Falkland Plateau into the centre of the Argentine basin, probably as a result of gravitational processes or iceberg transport. On the other hand, goethite is transported subparallel to the slope and is found around the margins of the Argentine Basin suggesting transport by contour currents (Figure 4).

Figure 4 Mapped distribution of factor 1, goethite, in the Argentine Basin

Note: High values of the goethite factor along the Argentine margin (dashed pattern) suggest that goethite dispersal is related to river transport across the shelf and contour currents in deeper waters

Source: After Balsam and Wolhart, (1993)

4.1.13 Gulf of Mexico

The Gulf of Mexico is an ideal locality to use DRS to trace sediment movements. It is a small basin into which a large number of rivers transport sediment, sediment that can be identified spectrally and much of which appears to have a unique and limited source area. As a result, even though Gulf sedimentation is dominated by the Mississippi/Atchafalaya Rivers, there are many other sources that can be identified and sediment transport paths can traced from the shelf into deeper parts of the Gulf (Balsam and Beeson, 2003; Ellwood, Balsam and Roberts, 2006). Balsam and Beeson (2003) undertook a DRS study of Gulf surface sediments using 186 core top and grab samples distributed throughout the Gulf, with the exception of the Yucatan Shelf where they lacked samples. From the first derivatives of the spectral data, seven factors were extracted. In order to their relative importance these factors are interpreted as:

1 marl and calcareous clay

2 glauconite

3 kaolinite

4 organic matter


5 phosphorite

6 hematite

7 goethite.

Of these factors, the two most interesting in terms of sediment transport and dispersal are kaolinite and hematite. Kaolinite is generally produced by intense chemical weathering. These conditions are met in the southeastern USA which is warm and humid. As a result, rivers that drain the southeastern USA input kaolinite to the Gulf of Mexico. However, it is only the kaolinite from the Alabama/Tombigbee and Yellow/Escambia Rivers that appears to be transported across the shelf and into deeper portions of the Gulf (Figure 5). Other rivers, for example the Sabine and Trinity Rivers which drain humid areas of east Texas, transport kaolinite onto the shelf, but not into deeper portions of the Gulf. Some kaolinite also appears to be transported by the Rio Grande (Figure 5), possibly derived from the volcanic terrain in the southwestern USA and Mexico drained by this river. This kaolinite is transported across the shelf and into deeper portions of the Gulf where it joins kaolinite coming from the southeastern USA to form a continuous band of moderate and high kaolinite values in deeper portions of the Gulf.

Figure 5 Distribution of the kaolinite factor in the Gulf of Mexico

Note: The kaolinite source is mainly from the Alabama/Tombigbee and Yellow/Escambia Rivers in the northern Gulf and from the Rio Grande in the western Gulf.

Source: After Balsam and Beeson (2003).

The Rio Grande and possibly the Colorado River (of Texas) appear to be significant sources of hematite that is transported into deeper portions of the Gulf (Figure 6). Once in the marine environment, hematite appears to be transported across the shelf off the Rio Grande and then eastward, downslope along the base of the Sigsbee Escarpment. By using hematite, kaolinite and clays, Balsam and Beeson (2003) were able to propose sediment transport and dispersal paths in the Gulf of Mexico (Figure 7). Their data indicate two main paths that sediment follows into deeper portions of the Gulf of Mexico.


One path, indicated by kaolinite and clays from the Mississippi, is down the Desoto and Mississippi Canyons and onto the Mississippi Fan. The second main path, indicated by hematite and kaolinite, is down the Rio Grande Slope with transport parallel to the Sigsbee Escarpment toward the Mississippi Fan. In the main, this work is supported by Ellwood, Balsam and Roberts (2006) in their analysis of magnetic properties of the same samples used by Balsam and Beeson (2003).

Figure 6 Distribution of the hematite factor in the Gulf of Mexico

Note: Darker patterns (higher factor scores) indicate higher concentrations of hematite. The hematite source is mainly from the Rio Grande in the western Gulf.


Figure 7 Recent sediment dispersal paths as indicated by DRS and smear slide analysis (for colours see online version)



4.1.2 LGM sediment distribution – Atlantic

The only large scale map of a DRS-identified mineral for a time period other than the present is Balsam, Otto-Bliesner and Deaton’s (1995) map showing the distribution of hematite in the Atlantic Ocean at the present (Figure 8A) and during the LGM (Figure 8B). This map shows the input of hematite-rich dust from the Sahara Desert into the Atlantic Ocean off northwest Africa. It also shows an area of high hematite content off the east coast of North America. During the LGM (Figure 8B), marine sediments off eastern North America were, at least in part, derived from Permo-Carboniferous red beds in the Canadian Maritime Provinces. Sediments from these red beds coloured the marine sediments off eastern North America red, forming what has been called ‘Brick Red Lutites’. These are described in detail below.

Figure 8 Maps of a factor interpreted as hematite in the Atlantic Ocean contrasting (A) the modern (core top) and (B) Last Glacial Maximum distribution (LGM) of this mineral

Note: The cross-hatched pattern indicates high concentrations of hematite with darker patterns indicating higher concentrations; the dashed pattern indicates lack of hematite. Hematite exhibits two main regions of accumulation in the Atlantic Ocean, one off northwest Africa as a result of aeolian input and a second off eastern North America from Brick Red Lutites (BRL). Significant aeolian input off northwest Africa is present in both modern and LGM sediments but expands during glacial time. BRL results primarily from ice transport into the western North Atlantic during glacial and late glacial times from a Permo-Carboniferous red bed source located in the maritime provinces of Canada (see text for details).

Source: After Balsam, Otto-Bliesner and Deaton (1995).


4.2 Time-series studies/local time-slice studies

The division of time-series studies into piston core and drill core (ODP holes) may, at first seem unnecessary and arbitrary. However, whereas time-series can be produced from both piston cores and drill cores, the length of time involved in drill core studies generally is significantly longer than for piston core studies. For piston cores, time-series generally encompass a maximum of several hundred thousand years compared with millions of years for drill cores. This difference in time leads to differences in the number of samples analysed, especially for drill cores that span a million years or more.

4.2.1 ‘Brick Red Lutites’

The ability of DRS to identify iron oxides at low concentration has played a major role in several time-series studies. Barranco, Balsam and Deaton (1989) used DRS to trace the distribution of hematite-rich sediment (Brick Red Lutite, BRL), derived from the Maritime Provinces of Canada and transported into the western North Atlantic from late glacial time to the present (Figure 8A and B). They noted that the maximum input of BRL was during late glacial time (Figure 8B) and that the deposition of BRL had significantly decreased by the mid-Holocene (Figure 8A). BRL was also used to document the position of the Western Boundary Undercurrent (WBUC), a current flowing along the bottom off the east coast of North America. Once transported into the marine environment through the Cabot Straits, BRL was then transported to the south and west, at least as far as Cape Hatteras, by this deep boundary current. The locus of flow was at about 4,000 m off the Cabot Straits, possibly deepening to about 5,000 m off Cape Hatteras (Figure 8B).

4.2.2 Saharan dust transported to the Atlantic Ocean

Iron oxides were also used to trace the movement of dust, derived from the Sahara Desert, and transported by wind into the North Atlantic Ocean (Balsam, Otto-Bliesner and Deaton, 1995). As noted for centuries, dust derived from the Sahara has a red colour, a colour imparted by the mineral hematite. Evidence of this dust is clearly displayed on both the modern and LGM (Figure 8A and B) maps of the hematite factor. Although these two time-slice maps exhibit some minor differences, time-series studies show that the maximum extent of hematite transport into the Atlantic Ocean was not the LGM, but instead about 16 ka. At this time, not only did the geographic area affected by Saharan dust increase, but also the locus of deposition shifted to the south.

4.2.3 Asian dust transported to the Pacific Ocean

Asian dust differs from Saharan dust in that the concentration of hematite is lower and the concentration of goethite is higher. In order to determine the source of dust in core V21-146 taken off Japan, Balsam, Ellwood and Ji (2005) used DRS to analyse the iron oxides in the core and compared them with iron oxides on the Chinese Loess Plateau (CLP). The DRS signal from the iron oxides in V21-146 and CLP were nearly identical (Figure 9). In addition, Magnetic Susceptibility (MS) in V21-146 matched the MS trends in the Chinese loess sections. Balsam, Ellwood and Ji (2005) concluded that at least some of the sediment collecting at the site of V21-146 was dust derived from the CLP and transported by wind to the site of V21-146. However, the lack of correlation between iron


oxide concentrations and mass accumulation rates suggests multiple sources for the sediment accumulating at the site of V21-146 and, of these sources, dust from the CLP is likely a minor one.

Figure 9 A comparison of spectral patterns from marine core V21-146 taken in the Pacific off Japan (factor 1) and dust from several sites on the Chinese Loess Plateau (CLP), Luochuan, Huanxian and Yanchang (for colours see online version)

Note: This figure indicates that at least some of the sediment at the site of V21-146 is eolian and is derived from the CLP.

Source: After Balsam, Ellwood and Ji (2005).

4.3 Studies of Ocean Drilling Programme sediment cores: evaluation of shipboard – vs. laboratory-based analysis

During participation on ODP Leg 150, one of us (JED) discovered that ODP had deployed a hand-held Minolta CM-2002 spectrophotometer onboard the JOIDES Resolution for use during shipboard core description. The Minolta instrument was placed onboard ship primarily to assist sedimentologists in determining sediment colour during core description. Leg 150 sedimentologists found that the Minolta produced colour values that were too numerous and complex (i.e. fractional Munsell values) to easily relate to the standard colour charts, and the use of the Minolta for colour determination was discontinued early in the leg. However, on Leg 154 shipboard scientists realised that the instrument could also be used to measure optical lightness (i.e. brightness or L*; Curry et al., 1995), a potential proxy for carbonate content in marine pelagic sediments (Balsam, Deaton and Damuth, 1999). Carbonate curves are important for stratigraphy and as proxies for climate curves, especially during the Quaternary (e.g. Damuth, 1975; Cline and Hayes, 1976; Prell and Hayes, 1976 and references therein).


Upon completion of the leg, the present authors examined the specifications for the Minolta CM-2002 and found that it was, indeed, an actual hand-held spectrophotometer with the capability of measuring spectral data between 400 and 700 nm at 10 nm intervals. We decided that because one of us (JED) was going to participate in ODP Leg 155 in the near future, we would collect spectral data aboard ship using the Minolta instrument from all the Leg 155 cores to determine whether the Minolta data were comparable in quality to data generated by our shore-based laboratory grade spectrophotometers.

4.3.1 Leg 154 –Ceara Rise

Harris and Mix (1999) used DRS to analyse two Leg 154 ODP Sites, Holes 925 and 926 (Curry et al., 1995) on the Ceara Rise, a topographic high east of the Amazon Fan. They used a specially designed instrument to scan the cores and analysed the wavelength range from 250 to 950 nm (NUV/VIS/NIR). This instrument was first used on Leg 138, making it the first spectrophotometer to be deployed on an ODP vessel (Mix et al., 1992). By factor analysing VIS wavelengths they identified two iron oxide factors, a hematite factor and a goethite factor. Using the spectral data from Deaton and Balsam (1991) they produced equations to estimate the concentration of hematite and goethite in the ODP samples. Presuming the hematite and goethite deposited on the Ceara Rise was transported by the Amazon River they proceeded to use changes in the concentration of these minerals to monitor climate changes in the Amazon River Basin.

From Harris and Mix’s point of view, there are two main sources of sediment on the Ceara Rise, the fluvial outflow from the Amazon River and locally generated biogenic sediment. They reject the possibility of an eolian source for some of the sediments on the Ceara Rise, citing Sarnthein et al.’s (1982) suggestion that the locus of deposition of the African eolian plume is at about 20°N. However, when the extreme sensitivity of DRS is used to identify iron oxides in equatorial and tropical Atlantic cores (Balsam, Otto-Bliesner and Deaton, 1995), it is clear that the African dust plume does affect the Ceara Rise. Further, when the distribution of hematite in the Atlantic is mapped (Figure 8B), there does not appear to be increased influx of hematite from the Amazon River during the LGM; rather, the locus of hematite deposition appears to shift south toward the Ceara Rise. Cores from the Ceara Rise (Balsam, Otto-Bliesner and Deaton, 1995 and unpublished data) exhibit similar temporal patterns of hematite concentration to cores further east, suggesting that hematite on the Ceara Rise is not monitoring Amazon Basin aridity, but instead eolian sediments from Africa.

4.3.2 Results of Leg 155 – Amazon Deep-Sea Fan

During ODP Leg 155 to the Amazon Deep-Sea Fan the Minolta CM 2002 spectrophotometer was routinely used to measure spectral data (as opposed to colour data) down each core at approximately 5–10-cm intervals (Schneider et al., 1995). We compared this shipboard Minolta data set, measured on wet Amazon Fan sediments, with spectral data derived from dry Amazon Fan sediment samples using our laboratory-grade Perkin–Elmer Lambda 6 spectrophotometer (Balsam, Damuth and Schneider, 1997). This comparison yielded surprisingly good results for each set of data when the first-derivative curves are used for comparison, especially if wavelengths < 430 nm were eliminated, due to a bias introduced by covering the wet cores on shipboard with plastic food wrap


(Saran Wrap™) to prevent damage to the Minolta. (After testing several brands of plastic food wrap, we subsequently recommended that Glad Cling Wrap™ be used to cover wet cores because it was the only tested brand that transmitted light uniformly across the VIS and had a minimal effect on spectra (Balsam, Damuth and Schneider, 1997)). However, absolute values of the first derivatives derived from wet percent reflectance curves are smaller and produce first-derivative curves in which variations are muted compared with curves from the dry measurements.

Factor analysis of both the shore-based dry samples and the shipboard wet samples demonstrated that similar factors were present in both data sets, but the shore-based data set produced at least seven interpretable factors, whereas the wet core data yielded only four factors. These four comparable factors from each data set can be interpreted in terms of sediment composition and represent hematite and goethite, clay minerals, organic matter and carbonate (Balsam, Damuth and Schneider, 1997). Factor scores plotted downhole for each drill site appear, in some intervals, to reflect temporal relationships related to fluctuations in fan sedimentation caused by glacio-eustatic sea-level fluctuations and are related to various fan units and deposits (e.g. levees, HARP units, mass-transport deposits, and hemipelagic sediments; Balsam, Damuth and Schneider, 1997).

4.3.3 Results of Leg 164 – Blake Outer Ridge

This leg provided an opportunity to study sediments from the Blake Outer Ridge, a giant contourite deposit shaped by the WBUC and the continental margin of the Carolinas. We used DRS to identify variations in transport of sediment (e.g. BRL, see above) from the Cabot Straits southward to the Blake Ridge by identifying hematite as a marker (Barranco, Balsam and Deaton, 1989). The hematite signatures downhole in the Leg 164 drilling sites on the Blake Outer Ridge indicate that significant quantities of hematite are present only in the top 2 m of the sediment column drilled at the crest of the Blake Ridge. This suggests that the WBUC was as shallow as these sites only during the very latest Pleistocene to Holocene (~14–5 ka; Balsam and Damuth, 2000).

4.3.4 Leg 174A – New Jersey Continental Slope

During ODP Leg 174A, a series of sites were drilled on the continental shelf off New Jersey as a part of the New Jersey Mid-Atlantic Sea-Level Transect, a major initiative dedicated to study the history of sea-level change along this continental margin (Mountain et al., 1994; Christie-Blick, Austin and Malone, 2003). A relatively continuous and complete upper Pleistocene section was recovered from the upper slope at Site 1073 (Austin et al., 1998). McHugh and Olson (2002) determined the 18O stratigraphy for this interval (0–500 mbsf) and calibrated it with 14C chronology and biostratigraphy. The resultant 18O curve was then correlated with the SPECMAP oxygen-isotope stratigraphy for the late Quaternary (Imbrie et al., 1984) which showed that this interval in Site 1073 spans Marine Isotope Stages (MIS) 1 (partial), 2–4, 5 (partial), and 8 through 18 (~5–700 ka; McHugh and Olson, 2002). Therefore, this section represents an important, highly expanded hemipelagic Pleistocene section that responds to fluctuations in terrigenous sediment on the upper slope throughout numerous glacial/interglacial cycles.

DRS analysis of dry samples with our laboratory based Lambda 6 produced results that assisted in determining the effects of sea-level change in this region. Factor 3 is


interpreted as hematite with subordinate amounts of goethite, both of which may form from alteration of siderite based on the downhole correlation of Factor 3 with siderite. Hematite and goethite identified here are not grain coatings, but are detrital microscopic grains produced during pedogenesis and transported along with clay minerals. Both Factor 3 and siderite exhibit excellent inverse correlations with the 18O curve suggesting that erosion and transport of soils, and large amounts of sediments in general, to the slope and rise occurred during periods of glacio-eustatic sea-level lowering. Factor 4 is interpreted as organic matter and shows a fairly good inverse correlation with 18Ofluctuations. This is also consistent with erosion and transport of detrital sediments from land and the exposed continental shelf to the slope and rise during periods of glacioeustatic sea-level lowering. Finally, DRS appears to be a more sensitive method of determining glauconite (Factor 2) downhole than smear-slide analysis because spectral analysis can identify much finer grains of this material than is possible using microscopy. The lack of correlation of glauconite downhole with other minerals typically derived from erosion on land suggests that the glauconite represented here is authigenic, rather than detrital.

4.3.5 Leg 188– Antarctic Continental Margin Contourites and Trough-Mouth Fan

ODP Leg 188 drilled sites on the continental shelf and margin seaward of Prydz Bay, offshore Antarctica (O’Brien et al., 2001). We measured and analysed NUV/VIS/NIR spectral data from dry, processed core samples recovered at ODP Sites 1165 from the Wild Drift, a large contourite drift deposit on the continental rise seaward of Prydz Bay, and Site 1167 from the Prydz Channel Trough Mouth Fan on the continental slope seaward of Prydz Channel, using our laboratory grade Perkin–Elmer spectrophotometer to help determine temporal mineralogical changes downhole (Damuth and Balsam, 2004).

We generated plots of factor scores downhole at the two sites to determine trends in temporal changes. At Site 1165, Factor 1 (goethite and chlorite) dominates above ~540 mbsf, Factor 2 (organic matter) is important above ~280 mbsf, and Factor 3 (clay minerals, possibly montmorillonite and illite) is high from 25 to 350 mbsf. Factor 4 (maghemite) exhibits two zones of high values from 20 to 50 and 540 to 750 mbsf. Factor 5 (hematite) is important from 0 to 95 mbsf. At Site 1167A, Factor 1 is important only in the top 5 mbsf, whereas Factor 2 is important throughout the entire hole. Factor 3 is important only from 210 to 227 mbsf, Factor 4 exhibits a few high values in the top 30 mbsf, and Factor 5 is most important from 30 to 210 mbsf. These variations are related to the input of glacial sediment in a complex manner.

4.3.6 Expedition 303 – Subpolar North Atlantic

One of the stated goals of IODP Expedition 303 was to examine climate changes on a sub-Milankovitch time scale. Data from geological sources clearly demonstrate that climate change has been strongly influenced by Milankovitch cycles, cyclic changes in Earth’s orbital parameters, most frequently cycles of about 20,000, 40,000 and 100,000 years. Sub-Milankovitch climate variability has only begun to be investigated and has produced some surprising results with the identification of periods of rapid jumps in the state of climate, that is, periods of abrupt climate change (Clark, Webb and Keigwin, 1999).


Expedition 303 acquired cores from the subpolar North Atlantic, one of the most climatically sensitive regions on Earth and a region that is prone to jumps in climate mode. Geologically, these abrupt jumps in climate-state are marked by Heinrich Events, the sudden influx of glacially transported sediment (especially detrital carbonate) in marine cores, and by the influx of Hematite-Stained Grains (HSG), both on quasi-periodic time-scales (Bond et al., 2001). HSG was one of the proxies that Bond et al. (2001) employed to document that these quasi-periodic, abrupt events, were related to changes in solar irradiance. Identification of HSG, as done by Bond et al. (2001) is both slow and time consuming. Samples are washed and sieved to extract the coarse (~sand) fraction. The coarse size fraction is then examined under a microscope and the number of HSG counted relative to all the grains in the sample. Because, as described above, DRS is extremely sensitive to hematite it can be used to identify sediment layers rich in HSG. In addition, because the technique is rapid and uses small samples, more closely spaced samples can be analysed in a shorter time then with the method used by Bond et al. (2001). Therefore, because of the suspected presence of HSG, we analysed samples from Expedition 303, Site 1308 taken from the ice-rafted detritus belt.

Figure 10 First-derivative value at 575 nm vs. modified composite depth in meters for Hole 1308

Note: The first-derivative value at 575 nm is a proxy for hematite content because hematite exhibits a major peak at this wavelength. The height of peaks seen in this diagram suggests a hematite concentration in excess of 0.1% by weight for the largest peaks considering the composition of the matrix (Deaton and Balsam, 1991). The largest peaks occur at about a 100,000 year interval. The smallest peaks record sub-Milankovitch cycles. Samples were taken at about a 500 year interval to insure that cycles as short as 1,500 years would be fully defined.


Our data from Site 1308 (Figure 10) clearly show long-term Milankovitch cycles and shorter-term cycles superimposed on the longer cycles. Our final interpretation awaits the development of a stratigraphy capable of producing millennial-scale and finer correlations. Expedition 303 participants are in the process of generating a paleointensity-assisted chronology that will allow correlations at a sub-Milankovitch scales. Once this is completed, we will be able to place our Site 1308 data into a detailed time framework.

5 Conclusions

In this study, we have shown that DRS is useful for identifying a variety of compounds in deep-sea sediments. Some of these compounds, chlorite for example, are easily identified by other means, specifically XRD. Other sediment components, hematite and goethite for example, are in such low concentration that they are not easily identified by other techniques. In addition, unlike XRD, DRS is capable of identifying compounds that are not crystalline like organic material. We have demonstrated the utility of DRS in a number of ways including using DRS to map the distribution of six sediment components in the Atlantic Ocean and using the DRS of one of these components, hematite, to trace the movement of eolian sediment from Africa into and across the tropical Atlantic. In addition, when applied to the Gulf of Mexico, DRS demonstrated the ability to trace sediment dispersal paths using kaolinite input along the Alabama/Mississippi shelf and hematite from the Rio Grande River.

Time-series studies from the Atlantic and Pacific demonstrate that DRS is capable of analysing a large number of small samples rapidly and producing consistent results. Some of these results, notably those from V21-146 taken in the Pacific off Japan, clearly reveal changes in Milankovitch bands. Because of the small amount of sample needed and rapid analysis times, DRS also holds the promise of producing results at sub-Milankovitch frequencies.

Acknowledgements

Parts of this research were supported by grants from NSF to WLB and JED. Specific grant numbers can be found in the cited literature. We thank Dr. B. Ellwood of the Louisiana State University for a thoughtful review of this manuscript. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the sponsors. We would also like to thank the core curators at Lamont-Doherty Earth Observatory and ODP for assistance in obtaining samples.

References

Austin, J.A. Jr., Christie-Blick, N., Malone, M.J., Berné, S., Borre, M.K., Claypool, G., Damuth, J.E., Delius, H., Dickens, G., Flemings, P., Fulthorpe, C., Hesselbo, S., Hoyanagi, K., Katz, M., Krawinkel, H., Major, C., McCarthy, F., McHugh, C., Mountain, G., Oda, H., Olson, H., Pirmez, C., Savrada, C., Smart, C., Sohl, L., Vanderaveroet, P., Wei, W. and Whiting, B. (1998) ‘Site 1073’, Paper presented in the Proceedings of the ODP, Init. Repts., 174A, College Station, TX (Ocean Drilling Program).


Barranco, F.T. Jr., Balsam, W.L. and Deaton, B.C. (1989) ‘Quantitative reassessment of brick red lutites: evidence from reflectance spectrophotometry’, Marine Geology, Vol. 89, pp.299–314.

Balsam, W.L. and Beeson, J. (2003) ‘Sea floor sediment distribution in the Gulf of Mexico’, Deep-Sea Research, Part I, Vol. 50, pp.1421–1444. (doi:10.1016/j.dsr.2003.06.001).

Balsam, W.L. and Damuth, J.E. (2000) ‘Further investigations of shipboard vs shore-based spectral data: Implications for interpreting Leg 164 sediment composition’, in C.K. Paull, R. Matsumoto, P.J. Wallace and W.P. Dillon (Eds), Paper presented in the Proceedings of theODP Sci. Results (Vol. 164, pp.313–324), College Station, TX.

Balsam, W.L. and Deaton, B.C. (1991) ‘Sediment dispersal in the Atlantic Ocean: evaluation by visible light spectra’, Reviews in Aquatic Sciences, Vol. 4, pp.411–447.

Balsam, W.L. and Deaton, B.C. (1996) ‘Determining the composition of late quaternary marine sediments from NUV, VIS and NIR diffuse reflectance spectra’, Marine Geology, Vol. 134, pp.31–55.

Balsam, W.L. and McCoy, F.W. Jr. (1987) ‘Atlantic sediments: glacial/interglacial comparisons’, Paleoceanography, Vol. 2, pp.531–542.

Balsam, W.L. and Williams, D. (1993) ‘Transport of carbonate sediment in the western North Atlantic: evidence from oxygen and carbon isotopes’, Marine Geology, Vol. 112, pp.23–34.

Balsam, W.L. and Wolhart, R. (1993) ‘Sediment dispersal in the Argentine Basin: evidence from visible light spectra’, Deep-Sea Research, Vol. 40, pp.1001–1031.

Balsam, W., Arimoto, R., Ji, J., Shen, Z. and Chen, J. (2007) ‘Aeolian dust in sediment: a re-examination of methods for identification and dispersal assessed by diffuse reflectance spectrophotometry’, Int. J. Environment and Health.

Balsam, W.L., Damuth, J.E. and Schneider, R.R. (1997) ‘Comparison of shipboard vs. shore-based spectral data from Amazon Fan cores: implications for interpreting sediment composition’, in R.D. Flood, D.J.W. Piper, A. Klaus and L.C. Peterson (Eds), Paper presented in the Proceedings of the ODP Sci. Results (Vol. 155, pp.193–215), College Station, TX.

Balsam, W.L., Deaton, B.C. and Damuth, J.E. (1998) ‘The effects of water content on diffuse reflectance spectrophotometry studies of deep-sea sediment cores’, Marine Geology, Vol. 149, pp.177–189.

Balsam, W.L., Deaton, B.C. and Damuth, J.E. (1999) ‘Evaluating optical lightness as a proxy for carbonate content in marine sediment cores: implications for marine sedimentation’, Marine Geology, Vol. 161, pp.141–153.

Balsam, W.L., Ellwood, B.B. and Ji, J. (2005) ‘Direct correlation of the marine oxygen isotope record with the Chinese Loess Plateau iron oxide and magnetic susceptibility records’, Palaeogeography, Palaeoclimatology, Palaeoecology, Vol. 221, pp.141–152.

Balsam, W.L., Ji, J. and Chen, J. (2004) ‘Climatic interpretation of the Luochuan and Lingtai loess section, China, based on changing iron oxide mineralogy’, Earth and Planetary Science Letters, Vol. 223, pp.335–348, (doi:10.1016/j.epsl.2004.04.023).

Balsam, W., Otto-Bliesner, B. and Deaton, B. (1995) ‘Modern and last glacial maximum eolian sedimentation patterns in the Atlantic Ocean interpreted from sediment iron oxide content’, Paleoceanography, Vol. 10, pp.493–507.

Biscaye, P.E. (1965) ‘Mineralogy and sedimentation of recent deep sea clay in the Atlantic Ocean and adjacent seas and oceans’, Geological Society of America Bulletin, Vol. 76, pp.803–832.

Bond, G., Kromer, B., Beer, J., Muscheler, R., Evans, M., Showers, W., Hoffman, S., Lotti-Bond, R., Hajdas, I. and Bonani, G. (2001) ‘Persistent solar influence on North Atlantic Climate during the Holocene’, Science, Vol. 294, pp.2130–2136.

Christie-Blick, N., Austin, J.A., Jr. and Malone, M.J. (Eds.) (2003) Paper presented in the Proceedings of the ODP, Sci. Results, 174A [CD-ROM]. Available at: Ocean Drilling Program, Texas A&M University, College Station, TX 77845-9547, USA.

Clark. P.U., Webb, R.S. and Keigwin, L.D. (1999) ‘Mechanisms of global climate change at millennial time scales’, Geophysical Monograph, Vol. 112, p.394. Washington, DC: American Geophysical Union.


Cline, R.M. and Hayes, J.D. (1976) ‘Late quaternary climatic record in western Equatorial Atlantic sediments’, in R. Cline and J.D. Hays (Eds), Investigations of Late Quaternary Paleoceanography and Paleoclimatology (Vol. 145, pp.165–200). Boulder, CO: Geological Society of America, Memoir.

Curry, W.B., Shackleton, N.J., Richter, C., Backman, J.E., Bassinot, F., Bickert, T., Chaisson, W.P., Cullen, J.L., deMenocal, P., Dobson, D.M., Ewert, L., Grützner, J., Hagelberg, T.K., Hampt, G., Harris, S.E., Herbert, T.D., Moran, K., Murayama, M., Murray, D.W., Pearson, P.N., Raffi, I., Schneider, D.A., Tiedemann, R., Valet, J-P., Weedon, G.P., Yasuda, H. and Zachos, J.C. (1995) Proc. ODP, Init. Repts., 154: College Station, TX (Ocean Drilling Program).

Damuth, J.E. (1975) ‘Quaternary climate change as revealed by calcium carbonate fluctuations in western Equatorial Atlantic sediments’, Deep-Sea Research, Vol. 22, pp.725–743.

Damuth, J.E. and Balsam, W.L. (2004) ‘Data report: spectral data from Sites 1165 and 1167including the HiRISC Section from Hole 1165B’ [CD-ROM], in A.K. Cooper, P.E. O’Brien and C. Richter (Eds), Paper presented in the Proceedings of the ODP, Sci. Results (Vol. 188, pp.1–49). College Station, TX: Ocean Drilling Program.

Deaton, B.C. and Balsam, W.L. (1991) ‘Visible spectroscopy – a rapid method for determining hematite and goethite concentration in geological materials’, Journal of Sedimentary Petrology, Vol. 61, pp.628–632.

Ellwood, B.B., Balsam, W.L. and Roberts, H. (2006) ‘Gulf of Mexico sediment sources and sediment transport trends from magnetic susceptibility measurements of surface samples’, Marine Geology, Vol. 230, pp.237–248.

Gaffey, S. (1985) ‘Reflectance spectroscopy in the visible and near-infrared (0.35–2.55 micron): applications in carbonate petrology’, Geology, Vol. 13, pp.270–273.

Griffen, G.M. (1962) ‘Regional clay-mineral facies – product of weathering intensity and current distribution in the northeastern Gulf of Mexico’, Geological Society of America Bulletin,Vol. 73, pp.737–768.

Harris, S.E. and Mix, A.C. (1999) ‘Pleistocene precipitation balance in the Amazon Basin recorded in deep sea sediments’, Quaternary Research, Vol. 51, pp.14–26.

Herbert, T.D., Tom, B.A. and Burnett, C. (1992) ‘Precise major component determinations in deep-sea sediments using Fourier Transform Infrared Spectroscopy’, Geochimica et Cosmochimica Acta, Vol. 56, pp.1759–1763.

Imbrie, J., Hays, J.D., Martinson, D.G., McIntyre, A., Mix, A.C., Morley, J.J., Pisias, N.G., Prell, W.L. and Shackleton, N.J. (1984) ‘The orbital theory of Pleistocene climate: support from a revised chronology of the marine 18O record’, in A. Berger, J. Imbrie, J. Hays, G. Kukla and B. Saltzman (Eds), Milankovitch and Climate (Pt. 1) (pp.269–305). Dordrecht, Holland: D. Reidel.

Jones, G.A. (1984) ‘Advective transport of clay minerals in the region of the Rio Grande Rise’, Marine Geology, Vol. 58, pp.187–212.

McCoy, F.W. (1983) ‘Seafloor sediment: map of the circum-Pacific region, northwest quadrant (1:10,000,000)’, in K.J. Drummond (Ed.), Circum-Pacific Map Project of the Circum-Pacific Council for Energy and Mineral Resources. Tulsa, OK: American Association of Petroleum Geologists.

McHugh, C.M.G. and Olson, H.C. (2002) ‘Pleistocene chronology of continental margin sedimentation: new insights into traditional models, New Jersey’, Marine Geology, Vol. 186, pp.389–411.

Mix, A.C., Rugh, W., Pisias, N.G. and Veirs, S. (1992) ‘Leg 138 Shipboard sedimentologists, and Leg 138 Scientific Party, Color reflectance spectroscopy: a tool for rapid characterization of deep-sea sediments’, in N.G. Pisias, L.A. Mayer, T.R. Janecek, A. Palmer-Julson and T.H. van Andel (Eds), Proc. ODP, Init. Repts., 138 (pp.67–77). College Station, TX: Ocean Drilling Program.


Mountain, G.S., Miller, K.G., Blum, P., Alm, P-E., Aubry, M-P., Burckle, L.H., Christensen, B.A., Compton, J., Damuth, J.E., Deconinick, J-F., de Verteuil, L., Fulthorpe, C.S., Gartner, S., Guèrin, G., Hesselbo, S.P., Hoppie, B., Katz, M., Kotake, N., Lorenzo, J.M., McCracken, S., McHugh, C.M., Quayle, W.C., Saito, Y., Snyder, S.W., ten Kate, W.G., Urbat, M., Van Fossen, M.C. and Vecsea, A. (1994) Paper presented in the Proceedings of the ODP, Init. Repts. (Vol. 150). College Station, TX: Ocean Drilling Program.

O’Brien, P.E., Cooper, A.K., Richter, C., Barr, S.R., Bohaty, S.M., Claypool, G.E., Damuth, J.E., Erwin, P.S., Florindo, F., Forsberg, C.F., Grützner, J., Handwerger, D.A., Januszczak, N.N., Kaiko, A., Kryc, K.A., Lavelle, M., Passchier, S., Pospichal, J.J., Quilty, P.G., Rebesco, M.A., Strand, K.O., Taylor, B., Theissen, K.M., Warnke, D.A., Whalen, P.A., Whitehead, J.M. and Williams, T. (2001) Proc. ODP, Init. Repts., 188 [CDROM]. Available at: Ocean Drilling Program. College Station, TX: Texas A& M University.

Pinsak, A.P. and Murray, H.H. (1958) ‘Regional clay mineral patterns in the Gulf of Mexico’, Paper presented in the Proceedings of the Seventh National Conference on Clay and Clay Minerals, Washington (pp.162–177).

Prell, W.L. and Hayes, J.D. (1976) ‘Late Pleistocene faunal and temperature patterns of the Colombia Basin’, in R. Cline and J.D. Hays (Eds), Investigations of Late Quaternary Paleoceanography and Paleoclimatology (Vol. 145, pp.201–220). Boulder, CO: Geological Society of America, Memoir.

Sarnthein, M., Theide, J., Pflaumann, U., Erlenkeuser, H., Futterer, D., Koopmann, B., Lange, H. and Seibold, E. (1982) ‘Atmospheric and oceanic circulation patterns of northwest Africa during the past 25 million years’, in U. Von Rad, K. Hinz, M. Sarnthein and E. Seibold (Eds), Geology of the Northwest African Continental Margin (pp.545–604). Heidelberg, Germany: Springer.

Schneider, R.R., Cramp, A., Damuth, J.E., Hiscott, R.N., Kowsmann, R.O., Lopez, M., Nanayama, F., Normark, W.R. and Shipboard Scientific Party (1995) ‘Color-reflectance measurements obtained from Leg 155 cores’, in R.D. Flood, D.J.W Piper and A. Klaus (Eds), Paper presented in the Proceeding of ODP, Init. Repts. (155, pp.697–700), College Station, TX: Ocean Drilling Program.

Schwertmann, U. (1971) ‘Transformation of hematite to goethite in soils’, Nature, Vol. 232, pp.624–625.

Schwertmann, U. (1988) ‘Occurrence and formation of iron oxides in various pedoenvironments’, in J.W. Stucki, B.A. Goodman and U. Schwertmann (Eds), Iron in Soils and Clay(pp.267–308). Berlin, Germany: NATO Science Series C, Springer.

Documents

Marine sediment components: identification and dispersal assessed …€¦ · seismic-facies analysis; and geologic evolution of passive margins. He has published more than 55 papers