A new method of establishing deep-sea and surface temperatures across the last 5 million years provides "crucial" information about how the ice ages came about, according to scientists.

The study, published in the journal Nature, offers the first evidence that long-term trends in cooling and continental ice-volume cycles were not the same, while also providing new information on climate relationships that led to the development of ice ages over the last 2 million years.

"In fact, for temperature the major step toward the ice ages that have characterized the past two to three million years was a cooling event at 2.7 million years ago, but for ice-volume the crucial step was the development of the first intense ice age at around 2.15 million years ago," the researchers said in a statement. "Before these results, these were thought to have occurred together at about 2.5 million years ago."

Researchers from the University of Southampton, the National Oceanography Centre and the Australian National University collaborated on the study.

Our work focused on the discovery of new relationships within the natural Earth system. In that sense, the observed decoupling of temperature and ice-volume changes provides crucial new information for our understanding of how the ice ages developed," said study co-author Gavin Foster, from the school of Ocean and Earth Science at the University of Southampton.

"However, there are wider implications too," Foster said. "For example, a more refined sea-level record over millions of years is commercially interesting because it allows a better understanding of coastal sediment sequences that are relevant to the petroleum industry. Our record is also of interest to climate policy developments, because it opens the door to detailed comparisons between past atmospheric CO2 concentrations, global temperatures, and sea levels, which has enormous value to long-term future climate projections."

By recovering microscopic plankton fossils from the sea, the researchers were able to get a record of ancient water temperature by analyzing oxygen isotope ratios found in the fossils.

The region of the Mediterranean where the fossils were recovered is useful for this sort of analysis because the oxygen isotropic composition of the seawater is mainly determined by the flow of water through the Strait of Gibraltar, which is, in turn, sensitive to changes in global sea level.

During the ice ages, the continental ice sheet grew, and water flow through the Strait of Gibralter was reduced, thus causing a measurable increase in oxygen isotopes. The ancient plankton fossils have preserved this record, the researchers said.

"This is the first step for reconstructions from the Mediterranean records," said lead study author Eelco Rohling of Australian National University. "Our previous work has developed and refined this technique for Red Sea records, but in that location it is restricted to the last half a million years because there are no longer drill cores. In the Mediterranean, we could take it down all the way to 5.3 million years ago. There are uncertainties involved, so we included wide-ranging assessments of these, as well as pointers to the most promising avenues for improvement. This work lays the foundation for a concentrated effort toward refining and improving the new sea-level record."

Study co-author Mark Tamisiea from the National Oceanography Centre, Southampton added: "Flow through the Strait will depend not only on the ocean's volume, but also on how the land in the region moves up and down in response to the changing water levels. We use a global model of changes in the ocean and the ice sheets to estimate the deformation and gravity changes in the region, and how that will affect our estimate of global sea-level change."