INTRODUCTION AND BACKGROUND

To infer paleoceanographic information from physical and chemical characteristics of sediment it is necessary to obtain the most continuous and accurate measurements possible. Standard analytical methods for discrete samples are noncontinuous, time consuming, and expensive. Relatively fast core logging methods and certain high-resolution downhole measurements can now obtain continuous data at much finer scales (down to millimeter scale) than are practical for individual sampling methods. In addition, logging methods are nondestructive and, in the case of downhole measurements, record data continuously over nonrecovered and recovered intervals of the borehole. These enormous advantages are important for the relatively long Ocean Drilling Program (ODP) cores and especially for critical boundaries, which were major objectives of Leg 165 (Fig. 1A).

One especially interesting and important paleoceanographic episode sampled at two sites during Leg 165 is known as the Late Paleocene Thermal Maximum (LPTM) (Kennett and Stott, 1991; Zachos et al., 1993), which occurred approximately 55 Ma in the late Paleocene Epoch. Figure 1B shows the location of drill sites in the Caribbean of the Paleocene. The LPTM represents a relatively short but most pronounced event of global warming and is associated with dramatic changes in the biosphere and ocean circulation. Oxygen isotope records indicate a rapid (<10 k.y.) warming of high-latitude surface and deep waters (Kennett and Stott, 1991; Shackleton, 1986; Zachos et al., 1993). The large negative excursion of ¹⁸O is coeval with a large, short-term decrease in ¹³C in both marine and terrestrial records and a major global extinction of benthic foraminifers (Kennett and Stott, 1991; Pak and Miller, 1992; Thomas and Shackleton, 1996). The event is believed to have been associated with a temporary change in dominant deep-water sources from high to low latitudes (Pak and Miller, 1992; Eldholm and Thomas, 1993). The resulting decrease in dissolved oxygen content of warmer deep waters is likely the major cause of the mass extinction of benthic foraminifers (Kennett and Stott, 1990, 1991; Thomas and Shackleton, 1996). Isotope excursions and the benthic faunal extinction occurred rapidly (<10⁵ yr) (Kennett and Stott, 1991; Aubry et al., 1996), and isotope values and species richness returned to pre-excursion levels in ~50,000 yr (Kennett and Stott, 1991; Thomas and Shackleton, 1996). Prior to Leg 165, sediments documenting the LPTM were identified in deep sea sequences of the Southern Oceans (Kennett and Stott, 1991), the Indian Ocean (Zachos et al., 1992), the equatorial Pacific (Bralower et al., 1995), and the Atlantic (Pak and Miller, 1992; Thomas and Shackleton, 1996).

The long-term warming of the late Paleocene climate may be related to elevated levels of atmospheric CO₂, perhaps caused by the voluminous CO₂ degassing of the effusive eruptions from the North Atlantic igneous province (Eldholm and Thomas, 1993). However, it appears that only high latitudes warmed significantly, whereas equatorial areas remained at much the same temperature as today (Shackleton and Boersma, 1981; Bralower et al., 1995). Bralower et al. (1997) proposed that a circum-Caribbean volcanic episode, documented by abundant ash layers found interbedded within rocks containing the LPTM layer of Sites 999 and 1001, may have resulted in short-term atmospheric cooling preferentially at low latitudes. This preferential cooling at low latitudes would act to decrease the difference between high- and low-latitude sea-surface temperatures (SSTs), perhaps triggering a change of deep-water sources from high- to low-latitude areas. The consequent warming of deep water could have resulted in the dissociation of gas hydrates, fueling further climatic warming and the dramatic transformation within the global carbon cycle, as characterized by the large negative ¹³C excursion at the LPTM (Dickens et al., 1995, 1997).

In this paper we demonstrate that our data uniquely defines the lithologic, physical, and chemical properties across the two Caribbean LPTM sections by continuous downhole logging and shore-based measurements of gamma-ray attenuation porosity evaluator (GRAPE) density, magnetic susceptibility, and chemical intensities of the cores by different core scanning (i.e., core logging) methods.

There are a number of core logging methods that are now routinely utilized both aboard the JOIDES Resolution and in shore-based laboratories, such as Minolta™ color scanning and physical properties logged by a multisensor track (MST) (e.g., Curry et al., 1995; Sigurdsson, Leckie, Acton, et al., 1997). These types of conventional continuous core-log methods indirectly define parameters which are important for paleoceanographic interpretations (e.g., carbonate contents are derived from color data, sediment densities are calculated from attenuation of gamma-rays, terrigenous input is interpreted by variations in the magnetic susceptibility).

In addition to standard logging methods of cores and in the borehole, we applied a relatively new method of geochemical core logging. In contrast to conventional core log measurements, the X-ray fluorescence (XRF) core scanner is able to directly measure several parameters that are important for paleoceanographic interpretations, such as Ca (representing "carbonate") and Fe (representing "terrigenous" and/or "volcanic"). The XRF core scanner measurements are not affected by the length of time since the core was cut, which can affect color measurements, and/or reduction diagenesis within the core, which can alter magnetic susceptibility. These are factors that can have an enormous influence on the quality and continuity of conventional core log data and, therefore, limit paleoceanographic interpretation.

Continuous rotary coring rarely results in continuous core recovery, and Leg 165 drilling was no exception. Downhole measurements still provide the only continuous record of the borehole wall. We will show that detailed correlation of high-resolution downhole measurements and continuous core log data allow exact core log integration. The creation of composite sections was essential for sampling plans to verify and define the LPTM by its typical negative shift in carbon isotope values (Bralower et al., 1997), and the correlation of both adjacent holes and distant sites. Furthermore, the data provide thickness estimates of the LPTM claystone, time estimates of the onset and duration of the LPTM, and constraints on the occurrence and distribution of associated ash layers. Together these observations further define the dramatic oceanographic, climatic, and paleoenvironmental changes associated with the LPTM event.