A primary initiative for ODP is to understand the causes and consequences of millennial-scale climate change. Such changes, including the well-known "Heinrich" events at ~5- to 10-k.y. intervals and "Dansgaard-Oeschger" events of ~1- to 3-k.y. periods, are well documented in rapidly accumulating North Atlantic sediments (e.g., Labeyrie and Elliot, 1999; Clark et al., 2000). This regional expression of rapid climate change has led to hypotheses that millennial-scale changes are driven either by instabilities in Northern Hemisphere ice sheets that surround the North Atlantic region (MacAyeal, 1993) or by oscillations in the formation of North Atlantic Deep Water (Broecker et al., 1990) or in some cases perhaps by rhythmic solar forcing that is expressed in the changing climate of the North Atlantic (Bond et al., 1997, 2001). Similar millennial-scale climatic oscillations have been detected in the northeast Pacific (e.g., Hendy and Kennett, 1999, 2000), suggesting the transmission of rapid climate change events from their North Atlantic origins into the Pacific (Mikolajewicz et al., 1997).

An alternative (but perhaps not mutually exclusive) hypothesis to explain the occurrence of millennial-scale climate change is that rapid climate oscillations may emanate from the tropics, where they originate as unstable responses to insolation (McIntyre and Molfino, 1996). A likely tropical source of rapid climate variability may be modulation of interannual-to-decadal climate changes of the eastern tropical Pacific (Cane and Clement, 1999), a region well known for El Niño Southern Oscillation (ENSO) events. The tropical hypothesis is plausible, given lake and ice core records from South America that suggest long-term changes in the mean state of ENSO events (Rodbell et al., 1999; Thompson et al., 1998), glacial–interglacial sea-surface temperature changes that mimic spatial patterns of change associated with modern La Niña events (Pisias and Mix, 1997; Mix et al., 1999b; Beaufort et al., 2001), and model results suggesting the sensitivity of long-term average oceanic condition to changes in El Niño frequency forced by orbital insolation (Clement et al., 1999). It remains unclear how changes in the high latitudes and low latitudes interact, however.

On an interannual to decadal scale, Liu and Huang (2000) argue that based on modern heat budgets, warming in the equatorial Pacific over the past 50 yr is related to reduction of the wind-driven advection of cool water from the eastern boundary current and that such effects that dominate the eastern tropical Pacific extend even to the western Pacific and perhaps elsewhere. On longer (glacial–interglacial) timescales, a similar linkage between the equatorial Pacific and the eastern boundary current seems to be important (Pisias and Mix, 1997; Feldberg and Mix, 2002).

If the tropical hypothesis is true, we expect that millennial-scale climate events will be clearly recorded in sites with high sedimentation rates from the eastern equatorial Pacific, particularly near the Galapagos Islands, a region sampled at Site 1240. If the connection to the eastern boundary current is important in driving such changes, we might also expect to find similar patterns of variability in sites to the south, including Sites 1233, 1234, and 1235.

We already know, based on continental paleoenvironmental records in midlatitude Chile such as glacier advances in the southern Andes correlated to Northern Hemisphere climate events (Lowell et al., 1995; Denton et al., 1999), vegetation changes during both Termination 1 and marine isotope Stage 3 (MIS 3) (Heusser et al., 1999; Moreno et al., 2001), and weathering activity in the Chilean Norte Chico (27°S) (Lamy et al., 2000), that this region is very sensitive to rapid climate change. However, so far it has proven difficult to compare such changes unambiguously to those of the Northern Hemisphere because of limitations in dating.

Short marine cores from the region reveal that surface ocean properties also varied significantly within the southern Peru-Chile Current. Sea-surface temperature reconstructions based on the U^k'₃₇ method reveal pronounced millennial-scale oscillations during both the late glacial period (Kim et al., in press) and the Holocene (Lamy et al., 2002). Similar variations in sea-surface salinity (SSS) reconstructed for the Holocene at Site 1233 indicate close links to continental climate changes, as SSS within this region is strongly influenced by freshwater input in the southern Chilean fjord region (Lamy et al., 2002).

Recent data, especially from the northeast Pacific (Behl and Kennett, 1996; Mix et al., 1999a; Lund and Mix, 1998) also affirm the importance of understanding Pacific deep and intermediate water circulation on millennial timescales. Broecker (1998) points to the existence of a "bipolar seesaw" effect, in which millennial-scale changes in Antarctic temperature are out of phase with (leading) Northern Hemisphere events. This inference is buttressed by data from the Southern Ocean (Charles et al., 1996) and by comparison of ice core ¹⁸O or ²H data from both hemispheres synchronized with the methane record (Blunier et al., 1998). Because the patterns of change are significantly different in the Northern and Southern Hemispheres, we can use the pattern of ventilation events on the millennial scale as a "fingerprint" of their source.

In the Pacific, Northern Hemisphere sites such as Site 893 from the Santa Barbara Basin provide compelling evidence of millennial-scale events of enriched oxygen content (i.e, unvarved sediments containing oxic benthic foraminiferal assemblages), which are approximately correlated with cold events in the North Atlantic (Kennett and Ingram, 1995; Behl and Kennett, 1996; Cannariato et al., 1999). Off Oregon, Site 1019 (980 m water depth) suggests that regional productivity effects could contribute to variations in the oxygen minimum zone off California (Mix et al., 1999a) and that ventilation of intermediate water masses is stronger during warm (e.g., Bølling-Allerød) climate events than it is during cold (e.g., Younger Dryas) climate events. At greater water depths, a possible Southern Hemisphere connection is found, as variations in benthic ¹³C appear linked to the Antarctic Cold Reversal, which precedes Younger Dryas cooling of the Northern Hemisphere (Blunier et al., 1997). With these data alone, the processes responsible for millennial-scale climate changes in the Pacific remain unclear.

Leg 202 sites will contribute to the understanding of millennial-scale climate changes by providing records of high sedimentation rates from sites off central Chile (Sites 1233, 1234, and 1235), within the equatorial cold tongue (Site 1240), and on the Costa Rica margin (Site 1242).

The sites drilled along the Chilean continental margin at 41°S (Site 1233) and ~3°S (Sites 1234 and 1235) (Fig. F29) provide sediment sequences with the potential for unprecedented ultra high resolution records of surface and deep-ocean conditions as well as continental paleoclimates during the last two glacial–interglacial cycles.

Sites 1233–1235 are particularly well located for assessing variations in the strength of AAIW through time (Fig. F30). Site 1235 monitors the boundary between the Gunther Undercurrent (the poleward undercurrent, a relatively low-oxygen water mass), and AAIW (a relatively high-oxygen water mass). Site 1233 is located roughly in the core of AAIW and thus should provide the best available record of AAIW properties relatively close to its source in the Antarctic Subpolar Front. Finally, Site 1234 monitors the deeper boundary of AAIW in its zone of mixing with Pacific Central Water (a relatively low-oxygen water mass).

Together, these records will fill a crucial gap in the paleoceanographic archive because there are no existing high-resolution cores capable of monitoring the intermediate waters that upwell at low latitudes and influence productivity and carbon outgassing to the atmosphere (Broecker and Peng, 1982). The temperature changes deduced from benthic foraminiferal oxygen isotopes at Sites 1233–1235 should provide a sensitive monitor for assessing the effect of intermediate-depth temperature changes on tropical climate over millennial to decadal timescales. Thus, Sites 1233–1235 afford an excellent opportunity to examine the role of chemical changes in high-latitude surface waters that can be transferred to the tropics via AAIW (e.g., Oppo and Fairbanks, 1989; Ninnemann and Charles, 1997).

Shipboard biostratigraphic data indicate that the bases of Sites 1233–1235 are younger than 260 ka. Additional tentative stratigraphic markers are derived from analyses of paleomagnetic variations, such as the Laschamp Excursion at ~41 ka, which is particularly pronounced at Sites 1233 and 1234 (see "Biostratigraphy, Magnetostratigraphy, Age Models, and Sedimentation Rates," above). These data indicate that sedimentation rates are generally >50 cm/k.y. reaching >1 m/k.y. in major parts of the recovered sequences, particularly at Sites 1233 and 1235. Such high sedimentation rates at key locations provide an outstanding opportunity to explore atmospheric and oceanographic variability in the Southern Hemisphere on the relatively short timescales from millennia to centuries and perhaps even decades.

Sediments at all three sites are dominated by lithologically homogeneous fine-grained terrigenous material with very high magnetic susceptibilities (Fig. F31). Minor amounts of biogenic components, mostly well preserved for paleoceanographic studies, are present throughout nearly the entire recovered sequences. The high sedimentation rates can be explained by extremely high terrigenous sediment supply due to enhanced fluvial discharge in response to heavy continental rainfall (Lamy et al., 2001). Probably in response to northward decreasing precipitation, sedimentation rates appear to be lower at Sites 1234 and 1235 when compared to Site 1233, particularly during MIS 2 to 3. However, this pattern might have been different in the older parts of the records, either indicating a different contribution of syndepositional sediment focused within the local basins or different rainfall patterns during MIS 5 and 6.

The magnetic susceptibility records of all sites document pronounced variability both on Milankovitch and sub-Milankovitch timescales. Higher magnetic susceptibility appears to be characteristic of glacial intervals, which might indicate increased terrigenous sediment input due to higher rainfall during regional cold phases. This is consistent with findings from marine sediment cores off central and northern Chile (Lamy et al., 1998, 1999).

Even as the Leg 202 cores were being analyzed at sea, we were able to develop preliminary age models based on correlation of magnetic susceptibility data at Site 1233 and in a ¹⁴C-accelerator mass spectrometry (AMS)-dated gravity core from the site that spans the last ~8000 yr (Lamy et al., 2001). In addition, shipboard paleomagnetic data revealed a major excursion of Earth's magnetic field, which we correlated to the well-known Laschamp Excursion at ~41 ka. The shipboard chronologies do not yet allow us to link specific variations in magnetic susceptibility at Sites 1233–1235 to climate changes recorded in either the Northern Hemisphere or Southern Hemisphere ice sheets. Nevertheless, it is already clear that such linkages are likely to be found. For example, within the interval of the last ~45 k.y., Site 1233 displays strong sub-Milankovitch variability with general patterns similar to those observed in polar ice cores (Fig. F32). By analogy to variations within the Holocene (Lamy et al., 2001) and to longer-period glacial–interglacial variations in the region, we expect that the variations in magnetic susceptibility at Site 1233 are related to changes in continental rainfall, which probably reflect latitudinal shifts of the southern westerly winds.

Shipboard data on diatom abundance and assemblage composition reveal pronounced sub-Milankovitch variability, possibly reflecting changes in upwelling and biological production at Site 1233. Furthermore, significant changes in continental freshwater input seem to be indicated by the abundance of freshwater diatoms.

In the tropics, Site 1240 is ideally located to assess rapid variations in equatorial upwelling, perhaps related to long-term stability of ENSO events (Clement, 1999; Beaufort et al., 2001). Here, sedimentation rates are near 10 cm/k.y., suggesting that climate changes on the scale of centuries may be well recorded here.

Site 1242 will provide for analyses of sea-surface salinities in the Panama Basin, related to excess precipitation relative to evaporation near the intertropical convergence. This region of the Panama Basin is noted for its extreme warmth (often >30°C), exceptionally low salinity (near 32), and a strong, shallow pycnocline (typically centered near 20–40 m depth). These features all reflect high rainfall relative to evaporation (Magaña et al., 1999), which stabilizes the water column and diminishes vertical mixing of heat and other properties. A portion of the net freshwater flux to the Panama Basin originates in the Atlantic or Caribbean (Jousaumme et al., 1986), so low salinities here partially reflect the transport of freshwater from the Atlantic to Pacific Basins via the atmosphere. The dynamics of this transport are important because this relatively small transport of freshwater helps to maintain the relatively high salinity of the Atlantic Ocean—a key parameter in maintaining the global thermohaline "conveyor belt" circulation dominated by North Atlantic Deep Water production (Zaucker et al., 1994; Rahmstorf, 1995). Thus, a millennial-scale record of sea-surface salinity variation at Site 1242 will help to test the tropical hypotheses of rapid climate change and will evaluate a possible mechanism that would link climate changes of the tropics via the hydrological cycle to those of the high latitudes through the control of oceanic salinity budgets on deepwater formation.

Variations in paleomagnetic secular variation and NRM intensity observed at Sites 1233–1235 and 1240 will likely facilitate synchronization of these important Southern Hemisphere sedimentary records to high-resolution records from the Northern Hemisphere, and this will provide a test of the phasing of millennial-scale climate changes between the two hemispheres.