In the work discussed here, Adkinset al.1 demonstrate that measurements of 230Th, a uranium decay product, can be used to account for changes in sediment deposition rate between the orbitally derived ‘mile posts’ for the beginning and end of the last interglacial period.

Uranium is soluble and of nearly constant concentration in sea water, whereas thorium adsorbs rapidly to sedimenting detrital particles. Decay of 234U to 230Th within sea water therefore produces a flux of 230Th to the sea floor that depends only on water depth. After accounting for some production of 230Th in sediments (from sedimentary U, which is easily measured) and some loss due to decay (230Th is itself radioactive), one can back-calculate the amount of 230Th at the time of deposition, which however requires independent estimates of age. These are, of course, lacking between the mile posts. But precise Th and U measurements in radiocarbon-dated Holocene sediments of the Bermuda Rise13,14 have shown that age- and U-corrected 230Th values were often considerably higher than expected solely from local seawater production, indicating that regionally deposited 230Th had been swept up by bottom currents and preferentially redeposited at the Bermuda Rise.

The amount of this sediment ‘focusing’ can be calculated simply as the excess 230Th measured and corrected for decay since the time of deposition divided by that expected from production in overlying sea water in a given increment of time13. At the Bermuda Rise this ‘focusing factor’ (F) varies inversely with the percentage by weight of biogenic carbonate, presumably because the same climatic forces that promote delivery of terrestrial detritus to the deep western Atlantic and, consequently, dilution of the near-constant surface production of carbonate, also act to invigorate recirculation at depth.

Whatever the cause, Adkins et al. realized they could estimate F from carbonate proxy data, and then use their 230Th results to calculate the time elapsed between measurements — which in this case were made at astonishingly dense 2-cm intervals. Their technique is still heavily reliant on the accuracy of the orbitally derived mile posts, but this way of refining the timescale will be very useful even if those posts are eventually moved.