early 1900s, when the 11 -year sunspot cycle was found to be recorded in the rings of trees in the southwestern United States. Major advances have been made since then, particularly within the last few decades. Many problems associated with interpretation of the rings (e.g., their causal relationship to climate) have been overcome, and the records have been extended from several hundred years to several thousand years or longer. Tree rings are now invaluable proxy indicators, because of their continuity and remarkable precision. In a similar vein, though more recently, the study of annual rings in massive corals is now yielding information regarding marine climates.
Many of the other proxy indicators, such as isotopes, fossil assemblages, and lake levels, had been used in the past to provide geological evidence in paleoclimate and paleoceanographic studies; they represent one of the standard tools in the geologist's arsenal. The relatively low resolution of the typical geological record had restricted their use chiefly to the study of long-time scale phenomena, such as glacial cycles. However, the desire to answer questions demanding higher-resolution data (for instance, whether the rapid climate excursions observed during the last deglaciation were anomalous, or were characteristic of a major climate transition) spurred improvements in the methods of recovery, analysis, and interpretation, which have yielded higher-resolution data. Improved collection techniques, such as long coring and better drilling methods, have made it possible to acquire the larger samples needed with the higher-resolution deposits; more natural data banks, such as ice caps and high deposition sediments, have been identified; and our understanding of both natural recorders and appropriate extraction techniques has increased, permitting higher precision with smaller samples.
All these developments are beginning to contribute to our knowledge of natural climate variability on decade-to-century time scales. Indeed, proxy indicators are now producing some of the most exciting and valuable records of variability to date. Furthermore, given the relatively embryonic state of the science, they have great potential for contributing to our understanding of the modern climate, particularly over longer time scales. New indicators are constantly being evaluated. For example, the use of biological ecosystems as proxy indicators of climate is demonstrated by McGowan (1995) and Reifsnyder (1995), both of which appear in this chapter. Numerous other proxy recorders—lake sediments, cave deposits, marshes—can now provide significant new insights. These are reviewed more thoroughly in Bradley and Jones (1992). In addition, the tremendous quantity of material in the historic records that is pertinent to climate is becoming increasingly useful, thanks to recent cataloguing that includes important metadata.
Numerous proxy indicators exist in the ocean, representing a wide spectrum of different oceanic and climate variables and spanning a wide range of time scales. Indicators residing in the sea-floor sediments may provide a nearly continuous record, often spanning tens of millions of years. These include pollen, faunal, and floral assemblages that have settled and accumulated, the isotopic compositions of skeletal material and tests from bottom-dwelling or floating organisms, and certain geological deposits and sediment types or compositions.
These ocean proxy records are accessible through sediment coring, drilling, and, in the case of certain bottom dwellers such as isolated corals, dredging. While the degrees of accuracy and resolution available vary, the information that can be extracted is staggering. For example (see the papers of Ruddiman and McIntyre, 1973; CLIMAP, 1981; and Imbrie et al., 1992 for more details), ocean proxy indicators have yielded regional information on the following characteristics: sea-surface and bottom temperature (from fossil assemblage composition), continental and landlocked ice volume and sea-level height (from oxygen isotope ratios), the partitioning of carbon between the land and oceans (from carbon isotope ratios), alkalinity of local water (from fossil preservation indices), deep-water circulation (from relative isotope compositions), surface productivity (from vertical isotope gradients), water-column stability (from radiolarian abundances), major front locations (from fossil, ice-rafted debris, and sediment-type distributions), deep-water temperature changes and surface salinity (from relative isotope compositions), vertical gradients of water-mass properties like temperature or salinity or of water-mass distributions (from analyses of sediments from different depths), deepwater velocity changes and source information (from sediment distributions near restricted passages), and predominant wind directions and intensities (from sediment compositions).
At typical sedimentation rates in the deep ocean basins (about 3 cm per 1000 years), sediment mixing serves as a low-pass filter, limiting the resolution of these proxy indicators to time scales of thousands of years. Regions in which the sediments accumulate faster offer the potential for resolving variations over significantly shorter time scales, but such areas often occur along continental margins where the interpretation of the sediment column is notoriously difficult, due to processes such as mass wasting. Consequently, until recently, these oceanic proxy indicators were used primarily to document and study climate variability on millennial or greater time scales. However, as Lehman (1995) explains in this chapter, areas with high deposition rates and relatively clean sediment records have now been located and sampled. These sample data are providing useful infor-