Earth's climate history contains clues about the global climate system and its natural range of variability that are crucial for assessing future climate change. Two frequencies of variability are of interest: orbital and sub-orbital. Although the idea that changes in the Earth's orbit affect climate in a predictable way has been a central tenet of climate change theory, this idea has been recently challenged by dating of the Penultimate Deglaciation, the timing of which appears to conflict with orbital predictions. Sub-orbital climate variability has been documented in many locations, particularly the Greenland ice cores and the North Atlantic, but it is not clear how significant these events are from a global perspective. Because sea level is the complement to global continental ice volume, sea-level changes are sensitive indicators of global climate. A detailed sea-level reconstruction is theoretically possible through uranium-thorium dating of corals. However, this potential has not been realized, primarily because of significant problems with coral dating. The source of excess 234 U in fossil corals and its relationship to uranium-series age determinations has been an outstanding problem in geochronology for more than 20 years, and it has become increasingly apparent that a substantial fraction of coral isotope ratios cannot be explained by closed-system decay. Here, a set of decay equations is derived from first principles that accounts for this open-system behavior, permitting the calculation of open-system coral ages. This new approach dramatically increases the accuracy and resolution of sea-level reconstructions. For the first time, it is possible to create a high-resolution sea-level curve for the last 250,000 years from published isotope ratio data. Two fundamental conclusions can be drawn from this sea-level reconstruction: (1) Significant suborbital sea-level variability seems to be a persistent feature of the record. (2) In spite of this complexity, the signature of orbitally driven sea-level change is clear. For the last 250,000 years, except 135-150 ka for which sea level data are lacking, each peak in Northern hemisphere summer insolation is followed by a peak in sea level.
