which today is close to freezing, was warm between 100 million and 70-million-years ago. This is inferred from the isotopic composition of foraminifera that then lived on the deep-seafloor. Paleoceanographic research suggests mechanisms that may have caused the refrigeration of the deep-sea since that time. Geologists are investigating the effects of those thermal changes as well as related aspects of ocean evolution on time scales that range from thousands to billions of years.
Although the history of seawater is an important subject in its own right, it also serves as an indicator of processes that have shaped the outer parts of the Earth through time. Limits for the composition of ancient oceans are determined from the mineralogy and chemistry of marine evaporites, the sediments formed by the evaporation of seawater, but even these indicators leave a wide range of uncertainty. The most useful technique currently available to define the major compositional variation of seawater over the past 600-million-years requires extraction and analysis of brines trapped in the rock salt found within marine evaporite deposits. In many instances these brines appear to have suffered little, if any, alteration. Their composition is not that of seawater, but the mass compositional parameters of the parent seawater can be reconstructed from the brine chemistry by correcting for the effects of evaporation and for the precipitation of limestone, gypsum, and rock salt. The results of the analysis of more than 100 inclusion fluids from marine evaporites covering the past 550-million-years of earth history suggest that the chemical composition of seawater has not changed greatly. This observation has come as something of a surprise, because the isotopic compositions of sulfur, strontium, and carbon in seawater have varied significantly. During the next few years the chemical evolution of seawater should be defined much more precisely, and we anticipate gaining a clearer understanding of the mechanisms that have controlled the composition of seawater.
We know less about patterns of circulation for modern oceans than about those of the modern atmosphere because of the logistical difficulty of gathering oceanographic data. This deficiency limits the accuracy of paleoceanographic modeling. However, conditions within ancient oceans can be reconstructed by using patterns of modern oceanic circulation to reassemble the thermal structure and dominant currents in ancient oceans and by selecting especially important physical, chemical, and biological indicators in the geological record to plot distributions. Nowhere has this approach been undertaken more effectively than in the Climate: Long-Range Investigation, Mapping, and Prediction (CLIMAP) project and its successors, broad international initiatives inaugurated in 1971 to recreate the ice age world of the past million years.
Although CLIMAP's broad goal was to investigate global climates for the past million years, its crowning achievement was the production of a climatic map of the world as it existed 18,000 years ago. This was the time of the most recent glacial maximum. In the overall strategy of CLIMAP, the most important element was reconstruction of sea-surface temperatures for the time frame of 18,000 years ago. Fossil occurrences of living marine species were used to chart the geographic distribution of ancient temperatures.
The most general conclusion drawn from the CLIMAP model was that 18,000 years ago the average sea-surface temperature was 2.3°C cooler than it is today. The high spatial resolution of the analysis permitted many more specific results. The equatorial Atlantic and Pacific oceans did not cool as much. Waters near the sea surface were generally cooler and more mixed than they are today, with a less pronounced thermal contrast between surface and deep waters. Ice floes extended to much lower latitudes in the North Atlantic—the Gulf Stream flowed eastward toward Spain, not Great Britain. And in the North Pacific, radiolarian species that today are restricted to cool waters from northern California to Washington ranged at least 1,000 km farther south. From other evidence we know that glacial expansion took place primarily in the north, with ice caps centered in Hudson Bay, Greenland, and Scandinavia, but marked climatic changes occurred in the Southern Hemisphere as well.
Reconstructing ocean temperatures and current patterns for earlier times is more difficult. Nevertheless, certain striking oceanographic changes that occurred tens and even hundreds of millions of years ago are clear. Deep-sea conditions changed drastically over geological time in response to profound global changes in shallow marine thermal regimes and in terrestrial climates. Today, throughout the globe the deep sea remains only slightly above freezing because its waters are derived from polar regions. At those high latitudes, surface currents cool so severely that they become much more dense than the underlying water. The chilled water sinks to the bottom and spreads along the deep seafloor to equatorial latitudes, forming a cold basal layer in all the oceans.
Fifty-five-million years ago, many regions of the Earth were much warmer than they are today. At