**********************************************************************************
 

1 Introduction

Studies of the role of the tropics in global climate during modern times have vastly improved the scientific communities understanding of the key oceanic and atmospheric processes.  However, the modern climate system and its variability may already be influenced by the injection of fossil fuel CO2 into the atmosphere.  Understanding the nature and variability pre-industrial age global climate require the use of proxy records, and corals are the most promising proxy to study tropical ocean-atmosphere component of the global climate system.
 

2 Corals and Global Climate

2.1 Overview of Modern Tropical Climatology

2.2 Corals as Proxies of Modern Tropical Climate

Coral records with seasonal to annual resolution can contribute to resolving key uncertainties in our knowledge of tropical climate.  Continuous coral records spanning several centuries are revealing the natural limits and behavior of the tropical ocean-atmosphere, at time scales that are relevant to society.  The aragonite skeletons of reef-building corals carry a diverse suite of isotopic and chemical indicators that track water temperature, salinity, and isotopic composition as well as site-specific features including turbidity, runoff, and upwelling intensity.  By documenting the natural behavior of these systems, we can assess their sensitivity to various forcings including natural phenomena, such as solar and volcanic changes, and anthropogenic inputs such as increasing greenhouse gas concentrations and land-use changes.  A high priority in coral research is to produce quantitative indicators of specific aspects of climate that can be integrated with other high-resolution paleoclimate data, including tree rings, ice cores, and varved sediments, and compared with climate model output to help assess model performance.

Massive corals from the tropical ocean are the only paleoclimate archive that offer both the annual resolution and multicentury record length needed for quantification of seasonal-centennial changes in the tropical surface ocean.  Worldwide, 15 to 20 coral-based climate records extending back to at least the mid 1800s AD are currently available or nearing completion.  The published records have provided new information on environmental changes in surface ocean conditions over the past several centuries in many regions of the tropics (Patzold, 1984; Druffel and Griffin, 1993; Dunbar et al., 1994, 1996; Heiss, 1994; Linsley et al., 1994; Charles et al. 1997; Crowley et al., 1997; Lough and Barnes, 1997; Isdale et al., 1998; Quinn et al., 1998).  An additional 20 to 30 records extend back 50 to 100 years from the present.  Since reliable instrumental records of past SST variability in the tropics rarely extend prior to about 1950, these coral records are highly useful for studies of decadal climate variability and linkages among climate systems at interannual time scales.

Most reef corals live at depths of < 20 m and grow continuously at rates of 6-20 mm yr-1, producing annual density bands that provide time markers for the development of long chronologies (Knutson et al., 1972).  Density bands provide an inexpensive, fast, and precise chronology of skeletal growth with many coral records having absolute annual chronologies.   However, some corals produce complicated or ambiguous density patterns and some coral records are assigned age accuracies of about 1%.  Where banding is absent or poorly defined, the seasonal cycling of detailed oxygen or carbon isotope records (Fairbanks and Dodge, 1979; Cole et al., 1993; Gagan et al., 1996; Evans et al., 1998) can be used to either fill gaps or even to establish relatively long chronologies.  Application of cross-dating and multiple age-specific tracers should allow most coral records to achieve true annual chronologic precision.

Many coral studies have utilized oxygen isotopic measurements because they are readily available and relatively straightforward to interpret.  Oxygen isotope ratios (d18O) of coral aragonite provide a useful history of environmental variability.  In oceanic settings where the oxygen isotopic composition of seawater is constant, coral skeletal d18O records SST variability, usually according to the standard paleotemperature relationship for carbonates (Epstein et al., 1953).   The isotopic composition is offset by a biological non-equilibrium component that appears to be stable through time, as long as a consistent, maximum growth axis is sampled within a coral colony (Weber and Woodhead, 1972; Dunbar and Wellington, 1981; Patzold, 1984; McConnaughey, 1989; Winter et al., 1991; Shen et al., 1992; Gagan et al., 1994; Leder et al., 1996; Swart et al., 1996; Wellington et al., 1996). When seawater d18O varies in response to changes in the balance between precipitation, evaporation, and water advection, the coral d18O changes accordingly (Swart and Coleman, 1980; Dunbar and Wellington, 1981; Cole et al., 1993; Gagan et al., 1994; Linsley et al., 1994; Fairbanks et al., 1997).  Long records of coral d18O have been used to develop precipitation reconstructions from sites where seawater d18O correlates with rainfall (Cole et al., 1993; Linsley et al., 1994).

Six of the eight longest published coral d18O records from equatorial and near-equatorial sites in the Pacific and Indian Oceans show a long-term warming / freshening trend generally beginning during, or before, the 19th century (Cole, 1996).  If the shift in d18O is due entirely to warming of the surface ocean, it is equivalent to 0.5 to 2oC since the early 1800s.  Superimposed on this long-term trend are abrupt, spatially coherent shifts in the coral d18O values.  Three widely separated records (Panama, Vanuatu, Aqaba) display an abrupt shift to warmer / wetter conditions at around 1850-1860.  Another significant shift to warmer / wetter conditions occurs between 1925 and 1940 in three records from the far western Pacific and Indian Ocean regions (Cebu, Seychelles, Aqaba).  In the most recent records from equatorial sites, the trend in coral d18O leads up to the 1980s, the warmest decade in the past two centuries.  This trend was identified from the much shorter instrumental record but dismissed as spurious, linked to possible biases in the data (Terray, 1995).
Intriguing as the long-term and abrupt shifts in coral d18O may be, they must be interpreted with caution.  Potential biological causes of long-term variability in coral d18O have yet to be determined via observation or experiment.  However, there are no a priori reasons to expect a strictly biological aging effect that would produce spurious results at multi-decadal to century time scales.  A non-biological cause of long-term variability in coral d18O may result from the coral surface growing at shallower water depths (by several meters) as the coral ages.  This may expose the coral growth surface to waters of slightly different temperature, salinity, and light intensity.  It is noteworthy that the Galapagos coral d18O record shows little change over the last 400 years, and is the only record produced by horizontal sampling of the coral colony.  However, the other seven records were produced from colonies whose upper growth surfaces terminated in a range of water depths (1-7 m) and all show similar long-term trends.  Moreover, the timing of the abrupt shifts in coral d18O (ca 1850-60 and 1925-40) is remarkably consistent among widely separated records, suggesting they may be responding to global climate forcing.

The other interesting aspect of these records is the large magnitude of variability they reveal at decade to century time scales, both in terms of SST and rainfall.  In the eastern Pacific, a comprehensive network of coral climate reconstructions is developing (Druffel, 1981; _Druffel et al., 1990; Linn et al., 1990; Shen, 1996; Shen et al., 1991, 1992; Dunbar et al., 1994, 1996; Linsley et al., 1994, 1999; Wellington and Dunbar, 1995; Wellington et al., 1996; Carriquiry et al., 1994).  The 370-yr coral d18O record of SST from the Galapagos reveals linkages among interannual through multi-decadal time scales.  The 280-yr d18O record from a rainfall-sensitive site near Panama indicates decadal periods in the strength / position of Intertropical Convergence Zone (ITCZ) precipitation superimposed on a long-term increase in either rainfall or SST.  The decade-century variance in the Galapagos and Panama records appears to correlate inversely, consistent with the southward displacement of the ITCZ during warmer periods in the Galapagos (Dunbar et al., 1996).  In the western Pacific, decadal variability of SST or rainfall is implicated at Vanuatu by a persistent 14-yr period in the d18O record (Quinn et al., 1993).  The 3-century record of d18O and D14C from Abraham Reef (Great Barrier Reef) has been interpreted as reflecting changes in circulation off northeast Australia (Druffel and Griffin, 1993).   SST reconstructions from Indian Ocean corals reveal ENSO frequencies, a biennial cycle and substantial variability on decadal time scales, which is poorly documented by the very limited instrumental SST record.  In the Seychelles coral, decadal patterns of variation correspond with Indian monsoon rainfall indices, suggesting that long-term regional rainfall variability may originate at least in part from the ocean (Charles et al., 1997).
Some regional isotopic events and trends can be explained in terms of variability in key climatic systems of the Pacific.  For example, the driest / coolest interval at Tarawa (early 1950s) corresponds with a very dry / cool period at Vanuatu (southwest Pacific) and a warm / wet period at Cebu (Philippines).  These observations are consistent with a westward contraction of the warm pool and consequent local anchoring of the Indonesian Low in the far western Pacific.  Cook (1995) discusses another multi-decadal inverse relationship between SSTs derived from the Galapagos coral record (Dunbar et al., 1994), a Great Barrier Reef SST index derived from coral growth band thickness (Lough and Barnes, 1997), and the Abraham Reef d18O record (Druffel and Griffin, 1993).  These results are consistent with the ENSO warm phase, which raises eastern Pacific SSTs and causes slight cooling and drought in northeast Australia.  Thus it appears that ENSO-type variability may operate over long time scales.

Most of these long coral time series have only recently become available and this short discussion is not meant to forecast the results of a comprehensive synthesis effort.  However, the results thus far are intriguing enough to warrant further data acquisition and initial attempts at quantitative synthesis.

2.3 Corals as Proxies of Past Tropical Climate

• The diversity of geochemical tracers in corals potentially encompasses a broad array of environmental parameters.  In many cases, redundant indicators permit the cross-checking of results;
• Corals have the ability to track seasonal climate changes within the annual cycle allowing sea-surface conditions to be reconstructed with reference to specific seasons.  Seasonally-resolved paleoclimate data are particularly important for reconstructing dynamic climate systems such as the ENSO, monsoon systems, and seasonal oceanic upwelling; and
• The precise measurement of 230Th/234U ratios in corals yields extraordinarily precise dates for "floating" chronologies from fossil corals that are generally reliable to within a few percent of their age (Edwards et al., 1987).  Annual density bands, or annually varying geochemical tracers, can then be used to achieve annual chronological control within individual records.

Several outstanding issues in paleoclimate research could be particularly well investigated using fossil corals.  For instance, reconstructing past SSTs is a classical problem in paleoceanography.  The precise measurement of the ratio of Sr to Ca in coral aragonite by TIMS offers a promising thermometer for reconstructing SSTs of the distant past (Beck et al., 1992; Guilderson et al., 1994; Shen et al., 1996, Alibert and McCulloch, 1997; Gagan et al., 1998; Evans et al., 1998).  The applicability of this technique to fossil corals depends, to a large degree, on the stability of the Sr/Ca ratio of seawater through time.  This condition holds today because high-precision measurements of Sr/Ca in modern reef waters show little variability, equivalent to offsets of only 0.2oC in reconstructed SSTs (deVilliers et al., 1994; Shen et al., 1996).   However, recent models by Stoll and Schrag (1998) suggest that dissolution of Sr-enriched aragonite exposed on continental shelves during sea-level lowstands (i.e., the LGM) will increase the Sr/Ca ratio of glacial seawater, potentially producing "cool" artifacts of 1-2oC in reconstructed SSTs.  Nevertheless, the Sr/Ca ratio of seawater should remain sufficiently stable during interglacial sea-level highstands to provide paleo-SSTs reliable to &plusmn;0.5oC.

The correlation observed between coral Sr/Ca, U/Ca (Shen and Dunbar, 1995; Min et al., 1995), Mg/Ca ratios (Mitsuguchi et al., 1996) and d18O may also make it possible to determine sea-surface d18O, by removal of the temperature component of the coral d18O signal.  Maps of sea-surface d18O could be produced to estimate variations in the volume of the planetary ice caps.  It may also be possible to produce sea-surface salinity maps, if the strong correlation between seawater d18O and salinity holds through time (Craig and Gordon, 1965; Fairbanks et al., 1992; Fairbanks et al., 1997; Rohling and Bigg, 1998), which can be used to recover past patterns of rainfall and evaporation over the tropical oceans (Gagan et al., 1998).

2.3.1 Holocene

Reconstructing past climate for periods when the tropics may have been warmer than today are important for understanding the potential role of the tropics in global climate change.  Recent work has shown that even a small increase in tropical SST (on the order of 0.5oC) leads to a marked increase in oceanic evaporation and precipitable water in the atmosphere, both on the order of 20% (Flohn et al., 1990).  Coral records of the distant past could yield new insights into the links between the hydrological cycle and tropical SSTs.  Model simulations show that the tropical hydrological cycle and latitudinal gradients in SST may drive changes in the mid-latitude atmospheric circulation (BUSH AND PHILANDER, 1998; Rind, 1998).  Failure to find a convincing explanation for global temperature variations, particularly during the last millennium before the industrial revolution, suggests that the tropical hydrological cycle may be a key player in forcing climatic changes at higher latitudes.

Fossil coral paleotemperature data could also shed light on the debate about the potential for self-regulation of SSTs in the tropical warm pool regions.  Previous studies have suggested that the long-term mean SST in the tropics cannot warm beyond about 29.5oC because of negative feedbacks in the radiation balance of the surface-ocean and atmosphere (Ramanathan et al., 1989; Ramanathan and Collins, 1991; Waliser and Graham, 1993).  Coral SST reconstructions from periods when the earth may have been warmer than today could provide "hard evidence" on whether or not this apparent SST maximum can be exceeded.  Key time-slices that could shed light on this important question might include the last interglacial (125 ka), the mid-Holocene "climatic optimum" (9-5 ka), and perhaps the Medieval Warm Epoch (ca AD 1000 to 1300).

2.3.2 Deglacial

So far, the Barbados curve is the only one to encompass the whole deglaciation as it is based on offshore drilling of the reef crest species Acropora palmata.  The Barbados record suggests that the last deglaciation was characterized by two brief periods of accelerated melting (i.e., meltwater pulse, MWP) superimposed on a smooth and continuous rise of sea level with no reversals.  These two meltwater pulses, MWP-1A and MWP-1B centered at 14,000 yr BP and 11,300 cal yr BP, respectively, are thought to correspond to massive inputs of freshwater derived from melting continental ice (Fairbanks et al., 1992).  However, there are still debates concerning the general pattern of sea-level rise during the last deglaciation events, including the amplitude of the maximum lowstand during the Last Glacial Maximum (Peltier 1998; Fleming et al. 1998) and the occurrence of periods of accelerated sea-level rise (Okuno and Nakada 1999).  Indeed, the Barbados sea-level curve was obtained from three separate drowned reefs.  Macintyre (1972) mapped the two deeper reefs at Barbados and throughout most of the Caribbean islands, interpreting them as global still stands during the last deglacial.  Fairbanks (1989; 1990) reinterpreted Macintyre's (1972) still-stand model and estimated that rates of sea level rise reached 5 to 9 cm per year during the melt water pulses, drowning the Caribbean reefs, and reefs world-wide, by far exceeding the maximum growth rate of any constructional coral reef framework.  The presence of MWP-1B at Tahiti is doubtful and it is also puzzling that the sea-level jump observed in New Guinea is delayed by several centuries when compared to MWP-1B observed at Barbados.  Locker et al. (1996) have proposed additional sea-level steps during the early part of the deglaciation between 18,000 and 16,000 cal yr BP.

Compilations of local sea-level curves (Pirazzoli 1991) and numerical models of post-glacial isostatic readjustments (Lambeck 1993; Peltier 1994, 1998) demonstrate that local sea-level histories varied considerably around the world in response to ice-sheet unloading and to redistribution of water masses in the global ocean.  Reconstructing the last deglaciation at many sites around the globe is thus crucial to obtain quantitative constraints on the geophysical models of the volumes and deglaciation rates of change of the individual ice sheets, which partly covered North America and Europe during the LGM (Peltier, 1998).  In particular, new records may help to resolve the controversy about the ice sheet sources of MWP-1A which vary according to authors: from primarily the Laurentide (Peltier, 1994; Fairbanks et al., 1992), from the Fennoscandian ice sheet (Lindstrom and MacAyeal, 1993) and from Antarctica (Clark et al., 1996).

Glacial sea levels before the LGM are even more controversial and still largely unknown.  It would be extremely important to evaluate the average sea level during that time period and to determine if rapid sea-level changes occurred in synchrony with climatic fluctuations such as Heinrich events and/or Dansgaard-Oeschger cycles.  Only three Barbados samples were recovered by offshore drilling and dated by TIMS (Fairbanks, 1990; Bard et al., 1990b).  By contrast, numerous samples are available from the lowest terraces at Papua New Guinea that correspond chronologically to isotope stage 3 (Chappell et al. 1996).  This favorable situation is due to the rather high tectonic uplift rate that allows these low sea-level reefs to be emerged today.  The disadvantages of these samples are that they have been subjected to meteoric water alteration and that a large tectonic correction is needed to reconstruct their position relative to past sea levels.  Indeed, the long-term (100 kyr) average uplift rate at Huon Peninsula is around 2-3 m/kyr but there is still a debate concerning the 10-1 kyr scale fluctuations (Edwards 1995; Peltier 1995; Peltier, 1998).

Sites located far away from glaciated regions ("far field") are needed to constrain model estimates of the eustatic changes (Fleming et al., 1998) and to complement the information obtained for the site of Barbados.  It is also essential to select and study other reefs characterized by vertical tectonic movements that are small or regular within the investigated time span.

3 Corals and Radiocarbon Calibration

For the Holocene period it has been possible to find abundant fossil pines and oaks, and thus produce a high-resolution atmospheric 14C/12C curve by comparing 14C levels and tree ring counts on the same tree logs. Unfortunately, it has not been possible to pursue this so-called dendro-calibration much further because of the scarcity of trees during the Late Glacial period in North America and Europe.  Other types of records have been used to continue the calibration effort: annually laminated sediments and shallow corals from tropical islands which can be cross-dated by high-precision dating techniques: 14C by AMS (Nelson et al. 1977; Bennett et al. 1977) and 230Th/234U by TIMS (Edwards et al. 1987).  Since the first paper on this topic (Bard et al. 1990a) there has been a continuous increase of the data base of corals which were used in the 1993 calibration software (CALIB.3, Stuiver and Reimer 1993; Bard et al. 1993).  The calibration software was recently updated (INTCAL98) and extended for use back to 24,000 cal yr BP.  The revised software both refines the precision of the calibration and extends the curve farther back in time (see the new Calibration Issue of Radiocarbon vol 40, nb 3, 1998 and web site http://www.radiocarbon.org/Journal/v40n3/notice.html).  The INTCAL98 curve and software (Stuiver et al., 1998) are based on decadal tree-ring measurements back to nearly 12,000 cal yr BP (Kromer and Spurk 1998), a marine varve sequence back to 14,600 cal yr BP (Hughen et al., 1998) and coral dates between 12,000 and 24,000 cal yr BP (Bard et al., 1998; Burr et al., 1998).
Collecting and analyzing new corals to fill gaps in the calibration curve between 15,000 and 24,000-cal yr BP is highly desirable.  Probably more important and rewarding would be to obtain and analyze coral samples ranging between the LGM and about 45,000 yr BP, the practical limit of the 14C method. Bard et al. (1990, 1993, 1998) analyzed two coral samples in that time range: the first dated at 30,230 &plusmn; 160 cal yr BP was collected offshore of Barbados and the second comes from one of the lowest uplifted terraces at Huon Peninsula, Papua New Guinea, (previously dated by TIMS U-Th at 41,100 &plusmn; 500 cal-yr-BP by Dia et al. 1992).  The discrepancies between the 14C and U-Th ages are about 4000 and 5500 yr, respectively, but the proper assessment of diagenetic alteration is critical for old samples and these data on isolated samples need to be replicated.

It is worth pointing out that large age shifts of the order of 5000 yr have also been obtained by comparing mass spectrometric 14C and U-Th ages measured in an archeological site (Bishoff et al. 1994).  In addition, Vogel and Kronfeld (1997) applied conventional radioactive counting techniques to date old stalagmites from South Africa by U-Th and 14C.  Although the reliability of 14C and U-Th ages in speleothems is not as high as for corals (problems of unknown or variable initial 14C age and detrital Th contamination in speleothems), the main outcome of the work by Vogel and Kronfeld is that 14C dates between 35,000 and 45,000 cal-yr-BP are indeed 5000 yr too young.  Large differences between 14C and TIMS U-Th ages were also obtained recently on lacustrine inorganic aragonite (Schramm et al. 1996) and on speleothems from the Bahamas (Richards et al. 1997).

These five independent works comparing 14C and U-Th ages are in apparent contrast with the atmospheric 14C/12C reconstructed by using varved sediments from a Japanese lake (Kitagawa and van der Plicht 1998).  However, as acknowledged by these authors, their varve counts between 20,000 and 38,000 cal-yr-BP "should be considered as minimum ages" due to the high probability of missing varves.  Furthermore, beyond 38,000 cal yr BP, the 14C/12C record from the Japanese lake is only based on an extrapolation of the sedimentation rate and not on true annual counts.
It is interesting to have an historical perspective on the respective contributions of varved sediments versus coral absolute dating: all extensions of the calibration based on varves have been subsequently shown to be biased by underestimation of the true ages (the problem of "missing varves").  This has been the case for the varves from the Swedish chronology, from Soppensee in Switzerland, from Holzmaar in Germany and from annual counting of the Camp Century and Dye 3 Greenland ice cores.  Drilling and collecting old coral samples is thus a crucial task for improving the calibration, as it seems difficult to rely solely on varved sediments.

4 Call for ODP-based, Shallow-Water Drilling of Corals

5 References

**********************************************************************************

**********************************************************************************

1 Introduction

2 The Link Between Depositional Process and Stratal Architecture

 2.1  Understanding the origin of stratigraphic unconformities

If sampling is so important, isn't the ancient outcrop adequate?  In many cases the answer is yes, but the sediments buried directly beneath a modern system have three advantages: 1) processes that we can monitor in a modern shallow water system provide an informative context in which to interpret the record that is buried beneath it, a feature that is totally lacking in subaerial outcrops; 2) having not suffered the tectonic displacement and diagenetic overprinting that is imposed on outrcrop records, in-place marine sediments provide a less complicated history to unravel;  and 3) marine seismic reflection profiles provide data with quality, coverage, and ease of acquisition far more readily than can be accomplished with most land-based surveys. Consequently, the link between depositional process and stratal architecture can be made in shallow water much more effectively than it can on land.

The last three decades have witnessed huge advances in the ability to image this stratal architecture on and beneath the seafloor: high frequency side-scan sonar reveals shallow marine landscapes in breadth and detail hardly imagined possible just a few years ago (e.g., Goff et al., 1996); swept-frequency sonar penetrates tens of meters into the seafloor and reveals structures of decimeter resolution;  multichannel reflection profiles probe to several hundred meters, and more, with several-meter scale resolution (Austin et al., 1996; Fulthorpe et al., 1999).

2.2 Eustatic control - acknowledged, controversial, but not the ultimate goal

Hesitant to adopt a eustatic curve developed from seismic sequence stratigraphy, many researchers have employed other strategies to derive the history of eustatic change. Studies of reef terraces and atolls (e.g., Fairbanks, 1989) provide the best proxy for sea-level over the past 125 k.y., although these records have provided only limited resolution for the older record (e.g., Quinn et al., 1991).  The d18O record of deep-sea sediments provides a proxy for glacioeustasy over at least the past 33-42 m.y. (i.e., since the formation of the Antarctic icesheet in the middle Eocene; Browning et al., 1996).  Although d18O records provide good evidence for the timing of Cenozoic glacioeustatic changes, their amplitudes can only be coarsely estimated (Miller et al., 1987,1991).  Kominz et al. (1998) have shown remarkably similar residual water depths remain in wells along the New Jersey coastal plain after backstripping has removed the effects of compaction, loading, and tectonic subsidence.  They attribute this to eustatic changes in the range of 0.5-3 Ma with amplitudes of 10's of m that are roughly half those of Haq et al. (1987).  Clearly we need more studies based on margins and platforms with widely contrasting tectonic and sediment supply to enable eustatic signals to be disentangled from other process that generate the stratal architecture of shallow water deposits (Reynolds et al., 1991).  This requires an integrated approach based on seismic imaging, modeling, and drilling.

2.3 The role and needs of modeling

The power of these models to isolate and explain processes acting on the shallow water record, however, is weakened by our limited success in gathering samples.  At present, our ability to image stratal architecture has far outpaced our ability to recover samples and anchor our models to verifiable aspects of the rock record.  The accuracy of backstrip models, for example, critically depends on the precision of total subsidence which is the sum of basement cooling history, crustal flexure, isostatic loading by water and sediment, and sediment compaction; thus it is important to know rifting history, age, composition, and sediment porosities as well as possible (Steckler et al., 1999).  Deposition models are sensitive to the many agents that rework sediments such as storms, longshore currents, fluidized sediment failures, and more; without groundtruth along the entire transport profile from sediment source to sink these models are unable to account for the variety of marine erosional surfaces commonly found in shallow water sediments.  Physical models are hampered by the obvious reduction in size and acceleration in time that makes a lab duplication of geologic processes possible; determining the proper spatial and temporal scaling factors depends in part on understanding their relationships in the real world.  Shallow water drilling has the potential of contributing to each of these modeling efforts.   But as long as we are restricted to stratal images with spotty groundtruth samples, our understanding of fundamental dynamics at the land-sea interface will remain stalled.

 2.4 Clinoforms - a key to understanding the silciiclastic shallow water record

2.5 Facies and hardgrounds - keys  to understanding the carbonate shallow water record

3 The Link Between Terrestrial and Marine Climate Records

Another unique aspect of shallow water deposits is the degree to which they contain direct evidence for both marine and terrestrial environmental change.  Conditions on the adjacent landmass may be reflected in the rate and nature of sediment supply and in the floral remains washed into the coastal setting.  Marine conditions are reflected in the sediment grain size and composition, the marine floral and faunal remains, and the sediment geochemistry.  This co-occurrence of marine and terrestrial environmental tracers in a single archive provides a crucial link between land-based records of climate and environmental change (such as those from tree-rings, speleothems, lakes, borehole temperatures), and ocean-dominated records available from deep sea sediments.  Such reconciliation of terrestrial and marine records of climate variability on all time scales from interdecadal to glacial-interglacial represents an important step towards understanding the causes and consequences of global and regional climate change.

Extracting climatic information from shallow marine deposits is sometimes difficult because of inadequate age control.  High sedimentation rates and variable environmental conditions can result in sparse and discontinuous microfossil (e.g., foraminiferal) records.  Nonetheless, with good site selection, rapidly-deposited shelf deposits can provide data with which to examine variables such as sea surface temperature, marine productivity, sediment supply, glacial melt-water discharge, terrestrial paleobotany, to name but a few.   The temporal resolution of rapidly-deposited shelf deposits can rival that provided by the high-latitude ice cores.  However, in contrast to ice cores, and to some extent coral records, rapidly deposited shallow water systems are not restricted to any particular latitude.  Consequently, they have the potential to provide a uniquely global view of land-sea interaction.

4. Variations in Relative Sea Level and Solid Earth Geophysics

 The solution to future sea level forecasts requires a global sea level equation or model that takes into consideration the Earth's continuously deforming lithosphere and ocean surfaces.  Construction of a global sea level equation or model, such as that of Peltier (1998) for example, requires a vast array of relative sea level curves for iterative model development and testing.  Regional sea level models may prove to be locally useful for higher resolution studies, for example see Lambeck et al., (1996, 1997).
 Hundreds of relative sea level curves have been published, albeit with an uneven distribution of geographic coverage, resolution, and age span.   A disproportionate number of relative sea level curves come from measurements in emergent areas where sections may be readily and inexpensively cored on land.  High offshore drilling costs and expenses associated with offshore geophysical surveys and mapping have prohibited all but a few scientists from extending their relative sealevel curves offshore (see Wingfield, (1995) for a review).

Recently, members of this research community convened a conference in Albufeira Portugal (Feb. 13-18, 1999) entitled"Glacial-Interglacial Sea level Changes in Four Dimensions: Quaternary Sea Levels, Climate Change and Crustal Dynamics".  This was the third conference in a four-part series on "sea level", sponsored by the European Science Foundation.  Many countries were represented and this community voiced their unanimous endorsement for the Ocean Drilling Program to support shallow drilling platforms for sea level research. There was no shortage of ideas of what scientific questions could be answered with offshore drilling capabilities.  It was proposed that the final sea level conference in this four part series, include sessions on specific scientific benefits of offshore drilling.  For example, modelers at Albufeira championed offshore drilling transects needed to test differences in the radial viscosity structure of the mantle.  Geologists offered compelling reasons to extend onshore relative sea level records to offshore depths to constrain local isostatic adjustment rates and to refine regional isostatic models.  Needs ranged from placing the offshore sedimentologic features into a sea level framework for national economic, geotechnical, climate research, and other national and international applications.

This community is uniquely ready to join the Ocean Drilling Program in an international effort of sea level research requiring shallow water drilling platforms.  First, the science of this subject has longstanding international participation and requirements.  Second, interaction between geophysical theory and observation is in a mature phase with each group benefiting handsomely from the other.  Third, advanced sea floor mapping technologies are available locally for site survey work. Fourth, standard geotechnical industry wireline drilling and percussion sampling hardware is suitable and available for charter on a regional basis.  Continuously cased holes are required in these sandy environments.  Fifth, radiocarbon dating via accelerator mass spectrometry is routinely available and provides suitable age control in most cases.  Sixth, knowledge of past regional sealevel histories is necessary for future sea level model predictions.  Seventh, satellite geodesy measurements of the Earth's continuously deforming land and ocean surfaces and changes in the Earth's rotation, for example, require a sea level model based on the geologic history of sea level at a regional scale.

5 Strategies

5.1 Seismic Imaging

Because the shallow water processes can operate at a wide range of temporal and spatial scales, and the stratigraphic record concerned may span  theMesozoic - Recent time frame, the resolution and penetration of the seismic imaging system must be chosen to optimally characterize specific target sections.  More than one recording system, providing overlapping or "nested" scales of observation, may be required to image both the shallow subsurface and deeper section (Austin et al., 1996).
Drilling in shallow water where hydrocarbons may be encountered heightens concerns regarding safety and pollution.  Consequently, guidelines have been developed for shallow water, pre-drilling surveys (JOIDES, 1994) which stipulate the seismic system and survey layout that must be adopted for drilling in less than 200 m waterdepths. The guidelines are subject to revision by the JOIDES Pollution Prevention and Safety Panel (PPSP) and proponents should contact the JOIDES Office and/or PPSP when designing such surveys.

5.2 Drilling

1) The Mid-Atlantic Transect, off New Jersey involved drilling siliciclastic sediments sites from onshore the coastal plain, across the continental shelf to the slope and rise (Legs 150, 150X, 174A, and 174A  X).Leg 174A drilled sites on the shelf in wat erdepths of 80 m, but clearly demonstrated the limitations for such work bythe JOIDES Resolution, which drills uncased holes without mud circulation and is subject to heave.  Sands madeup much of the section but were minimally recovered.  Additional sites essential to the transect drilling strategy have been selected and surveyed in water depths of 30-50 m.  However, it is highly unlikely that the JOIDES Resolution or other dynamically positioned, open-circulation platform will be the technology of choice.

2) Great Bahama Bank: ODP Leg 166 drilled on the slope and a separately funded jack-up rig drilled two sites on the bank top.  Recovery by the jack-up rig was excellent, demonstrating the advantages of eliminating all heave in shallow-water drilling by not drilling from a floating platform.
Although proposed drilling is not complete, the timing of Neogene sea-level fluctuations along the New Jersey margin and the Bahamas bank appears similar. Great efforts, above and beyond that required for most ODP proposals, were made in support of both of these programs.  Both groups of proponents obtained independent funding for onshore drilling and comprehensive seismic surveys, including hazards surveys (New Jersey), and a jack-up rig (Bahamas).

Studies of regional variations in subsidence suggest that the process-architecture link is more complicated that previously thought (Karner et al., 1993), raising the concern that features must be more extensively mapped and sampled in three dimensions than previously thought.  Therefore, the simple transect approach initially adopted for sea-level studies, comprising a series of sites aligned in the dip (shore-perpendicular) direction, may be inadequate for what is now viewed as a more fundamental investigation into process-architecture links.

Multiple drill sites remain essentialfor several reasons:

1) Depositional facies vary within sequences in response to migrating shorelines and variations in both the composition and rate of supply of sediment. These changes are particularly pronounced near the clinoform breakpoint, or rollover, which marks the paleoshelf edge and a pronounced basinward increase in accommodation.

2) Estimates of eustatic amplitudes derived from backstripping multiple sites provide a valuable a check on consistency (Kominz et al., 1998). Furthermore, paleobathymetric data along a transect of sites can constrain positions of coastal onlap and shorelines, enhancing the reliability of seismic geometries in determining amplitudes (Moore et al., 1987; Greenlee and Moore, 1988).

3) Comparatively deep-water facies yield optimum age control because a) the hiatus associated with base level lowerings narrows in the seaward direction,  and b) microfossil abundance and diversity generally increases basinward. Therefore, sequences drilled in marginal marine settings to sample nearshore facies must also be traced basinward on seismic profiles and sampled in a mor edistal setting.

5.3 Age control

6 References