Sedimentary cycles reflect climatic oscillations that are ultimately controlled by the Earth's orbital cycles. Therefore, sedimentary cycles can be used to construct astronomical timescales. Using this method, astronomical timescales have been established for the last 12 million years that are more acurate and have a much higher resolution than previous geological timescales. Such timescales are fundamental to an increasing number of applications in many disciplines of Earth science. Researchers are now working to extend the astronomical timescale to earlier time intervals and establish similar time- scales for the continental record.
Time is an indispensable tool for Earth scientists striving to understand all kinds of processes and determine rates of change. More than a century ago, just before the invention of radiometric dating, G. K. Gilbert realized that astronomically forced cyclicity in marine sedimentary archives can be used to estimate the duration of parts of the geological record. His estimates suggested that our planet was much older than the 100 million years (or even 20 m.y., by some accounts) that had been calculated by a thermodynamic cooling model of the Earth.
Gilbert linked his sedimentary cycles to perturbations in the Earth's orbit and rotation axis that are caused by gravitational interactions of our planet with the Sun, the Moon, and the other planets. These interactions give rise to cyclic changes in the eccentricity of the Earth's orbit, with main periods of 100,000 and 413,000 years, and in the tilt (obliquity) and precession of the Earth's axis with main periods of 41,000, and 21,000 years, respectively.
Eccentricity is a measure of the degree of elongation of the orbit and varies between nearly zero (circular orbit) and 0.06 (slightly elliptical). Obliquity describes the angle between the Earth's axis of rotation and the orbital plane and varies between 22° and 25°. Axial precession is the slow movement of the rotation axis around a circular path with one revolution completed every 26,000 years. Due to the opposite movement of the eccentric orbit itself, the precession of the equinoxes—also called climatic precession—completes one full cycle in about 21,000 years.
These perturbations in the Earth's orbit and rotation axis are climatically important because they affect the global, seasonal, and latitudinal distribution of the incoming solar insolation. They are held responsible for the Pleistocene ice ages but also affect low-latitude climatic systems such as the monsoon. Orbitally forced climatic oscillations are recorded in sedimentary archives through changes in sediment properties, fossil communities, and chemical characteristics.
While Earth scientists can read these archives to reconstruct paleoclimate, astronomers have formulated astronomical solutions that include both the solar-planetary system and the Earth-Moon system. With these astronomical solutions they compute the past variations in precession, obliquity, and eccentricity. As a logical next step, sedimentary archives can be dated by matching patterns of paleoclimatic variability with patterns in the computed astronomical curves. This astronomical tuning of the sedimentary record results in timescales that are independent of radioisotopic dating and are tied to the recent times through a direct match with astronomical curves. A major breakthrough came only during the last decades, when studies directed at establishing such astronomical timescales yielded unprecedented accuracy and resolution for the last 15 million years.
Initially, research focused mostly on the inferred orbital forcing of the Pleistocene ice ages, resulting in the astronomical calibration of late Pleistocene paleoclimatic records that mainly reflect glacial cyclicity. Following the early tuning of the late Pleistocene, the astronomical timescale is now firmly established for the last 5.3 million years—the entire Pliocene-Pleistocene—using paleoclimatic records from Ocean Drilling Project (ODP) sites in the eastern equatorial Pacific and North Atlantic [Shackleton et al., 1990] and sedimentary cycle patterns in marine successions exposed on land in the Mediterranean [Hilgen, 1991] . In the Mediterranean area, well-known sections such as as Capo Rossello, Eraclea Minoa, Singa, and Vrica were used to construct the timescale. These sections show a conspicuous cyclic bedding that is dominantly controlled by precession and eccentricity. The sections consist of carbonate cycles or sapropels(brownish-colored layers enriched in organic carbon) (Figure 1).
Fig. 1. The classical section of Punta di Maiata on southern Sicily, Italy. Small-scale quadripartite color cycles (grey-white, beige-white) reflect precession, larger-scale indurated (white) beds reflect 100- and 400-k.y. eccentricity cycles. (Photo courtesy of J. Dinarés)
Recently, Lourens et al. [1996] slightly modified the earlier astronomical timescale of Hilgen [1991] using different astronomical solutions and a more realistic target curve (insolation). Details in the sedimentary cycle patterns—in particular those caused by interference between precession and obliquity—allowed them to determine which astronomical solution is the most accurate from a geological point of view. This appeared to be the solution formulated by Laskar, with close to present-day values for the dynamical ellipticity of the Earth and the tidal dissipation by the Moon (both parameters are important because they affect the Earth-Moon system and change when entering a glacial).
This conclusion is consistent with recent theoretical modeling studies that show dynamical ellipticity will not change enough during glacials to significantly change the precessional period from its present value. The history of the astronomical timescale hit a turning point when magnetic reversals were dated independently of radioisotopic techniques. From 1982 onward, the astronomical polarity timescale (APTS) started to deviate from polarity timescales that were based on K/Ar dating. The new timescales revealed discrepancies of 5–10%, and the astronomical ages were consistently older than the K/Ar ages. Since then, the validity of the APTS has been confirmed by radioisotopic dating using the new 40Ar/39Ar (single crystal) laser fusion technique. The initial discrepancies are now explained by the difficulty of completely degassing the alkali feldspar sanidine using the conventional K/Ar method.
Today, the accuracy of 40Ar/39Ar dating is limited by uncertainty about the ages of neutron fluence monitors, or mineral dating standards. Ages currently used for these standards are slightly but significantly different and result in a 1.0–1.5 % error as compared with a 0.3 % precision of modern 40Ar/39Ar techniques. Ages of dating standards were recently calibrated independently of absolute isotopic abundance measurements; researchers compared 40Ar/39Ar and astronomical ages of geomagnetic polarity reversals over the last 3.5 m.y. [Renne et al., 1994] . This indirect approach via interpolated ages of magnetic reversal boundaries awaits confirmation, which could be accomplished by directly comparing astronomical and 40Ar/39Ar ages of ash layers intercalated in cyclic sedimentary sequences that have been astronomically dated. The intercalibration of radioisotopic and astronomical time aims to eliminate uncertainties about the correct ages of the dating standards.
Independent support of the APTS came from a study by Wilson [1993] , who showed that the astrochronology resulted in a more consistent and steady history of seafloor spreading rates. Wilson suggested only minor refinements to the astronomical timescale. In particular, he suggested that the age of 2.60 Ma for the Gauss/Matuyama boundary was too old by 10-20 kyr. Our detailed reversal record of the G/M boundary showed that this indeed was the case.
Application of the APTS reduced the discrepancy—from 6% to a marginally significant 1.5%—between averaged plate velocities calculated from space geodetic data for the last decade and from the Nuvel-1 global plate motion model for approximately the last 3 m.y. This outcome is consistent with the increased steadiness of plate motion inferred by Wilson [1993] and led to a recalibration of Nuvel-1 to Nuvel-1A [DeMets et al., 1994] .
Application of the APTS also provides better age constraints on the dating of short magnetic reversal excursions and reversals, tectonic pulses and rotations, evolutionary changes, biochronology and faunal turnovers, and long-term variations in lithology such as third-order sequences. A major reason for developing astronomical timescales, however, is their applications in paleoclimatic and paleoceanographic studies. In particular, such studies are directed at deciphering the intricate relationships between astronomical forcing and climate response, and their recording in sedimentary archives.
The next goal was to extend the APTS into the Miocene. On the basis of records from ODP Leg 138 in the eastern equatorial Pacific, a preliminary and partial astronomical tuning was established by Shackleton and coworkers for the interval between 6 and 10 Ma, while we obtained an astronomical duration for late Miocene polarity sequences on Crete (Greece) by multiplying the number of sedimentary cycles by the average period of the precession cycle.
Next we studied the sedimentary cycle patterns in the Cretan sections as well as those in partly older sections on Gavdos and Sicily and calibrated the cycle patterns to the astronomical curves [Hilgen et al., 1995, Figure 2] . For this purpose, we used the same phase relations (between sedimentary cycles and astronomical cycles) that were employed earlier in the Plio-Pleistocene. One of the results of the late Miocene APTS is an age of 7.24 Ma for the Tortonian-Messinian boundary. This infers that the Messinian lasted 1.91 m.y., since the Miocene-Pliocene boundary has an astronomical age of 5.33 Ma. The astronomical ages of the polarity reversals are consistently older than the ages in the most recent geomagnetic polarity timescale. Similar discrepancies are found with astronomically derived ages obtained from the orbital tuning of ODP Leg 138 records. Unpublished work on ODP Leg 154 sediments (N. J. Shackleton, personal communication, 1995), however, confirms that the late Miocene ages from Leg 138 are too young.
Fig. 2. Astronomical calibration of late Miocene sapropels to eccentricity. Large clusters (bold lines) correspond to 400-k.y. eccentricity maxima, small clusters to 100-k.y. maxima.
The Mediterranean APTS for the Miocene now spans from early Messinian (6.8 Ma) to late Serravalian(12.2 Ma). Since the previous timescale covered the last 5.3 m.y., this results in a "Messinian Gap" from 5.3 to 6.8 Ma. This gap is explained by the less favorable sediments (for example, diatomites, evaporites) deposited during the so-called Messinian 'salinity crisis' and the notoriously complex depositional history of the Mediterranean during this time interval. These less favorable sediments of the Messinian also display a cyclicity, however.
Closing the Messinian gap is one of the major goals of current research; this accomplishment will help solve many intriguing problems concerning the climatic and tectonic history of the Messinian. In addition, it will solve existing discrepancies regarding the age of Chron C3A. The astronomical ages of this Chron are of special interest because a major phase of plate reorganization and associated changes in spreading rates are attributed to the latest Miocene.
Open marine sequences were used to develop the late Neogene APTS, but the continental record clearly must be included if we are to determine the evolution of continental climate and its intercalibration with the marine realm. In fact, the terrestrial sedimentary record seems to be the logical place to search for Milankovitch cycles because, in the absence of oceanographic processes with their complicated, nonlinear feedback mechanisms, a more direct registration of orbitally induced changes in climate may be expected there. Potential and serious drawbacks of using continental successions, however, are the usual lack of accurate time control and the occurrence of hiatuses resulting from tectonic activity, base-level changes, autocyclic processes, and intermittent erosion. Nevertheless, regular subsidence balanced by continuous sedimentation does occur in continental settings as shown by classical studies of Milankovitch cycles in terrestrial successions [e.g., Olsen et al., 1996] . These examples, however, all date from remote periods in the Earth's history and lack the necessary first-order time control.
Our continental research efforts in the Mediterranean focus on fluviolacustrine cyclicity in late Miocene and Pliocene basin fills of Greece, and on middle to late Miocene red-bed sequences in Spain. Preliminary results from Spain suggest that a prominent cyclic bedding is related to the 100-kyr eccentricity cycle, while smaller-scale cycles may reflect the influence of precession. The construction of an integrated stratigraphic framework will allow detailed, cyclostratigraphic "bed-to-bed" correlations to marine sequences both within and outside the Mediterranean area. In addition, there are plans to extend the existing APTS back to 17 Ma.
Preliminary results of the lacustrine successions of the Ptolemais area in northwestern Greece suggest that cyclic lignite-marl alternations reflect precession-controlled variations in regional climate. The successions are being correlated with the previously studied marine sequences, which will help study of the influence of astronomically induced climatic changes along terrestrial-marine environmental gradients in the eastern Mediterranean. Such integrated marine-continental studies will provide important constraints for climate modeling experiments aiming at a better understanding of the astronomical forcing of climate.
Astronomical timescales that are anchored to the recent era via direct calibration to computed astronomical target curves will be extended through the Paleogene back into the Mesozoic. For these older time intervals, the 400-k.y. eccentricity cycle is particularly suitable for establishing first-order correlations to the astronomical record. A first attempt was made to establish an anchored astronomical timescale for the K/T boundary interval [Ten Kate and Sprenger, 1993] , even though high-precision 40Ar/39Ar datings of the boundary suggest that their astronomical correlation is too young by 400 k.y.
Also, pelagic successions of Cretaceous to Paleocene age from the southern Atlantic and the northern Apennines [e.g., Fischer et al., 1991] and lacustrine successions from the Triassic Newark rift basin [Olsen et al., 1996] could be used to extend the anchored astronomical timescale; they are assumed to be continuous and clearly reflect the influence of longer-term eccentricity cycles. For these older time intervals, the fine details of astronomical target curves cannot be employed because the accuracy of the astronomical solution is limited by the chaotic motion inherent to our solar system. These details can be derived from the geological record, however, and may help refine astronomical models and solutions.
Summarizing, the development and application of pre-Pliocene astronomical (polarity) timescales is still in its infancy despite the considerable progress made recently. The validity and success of the orbital tuning approach, however, is clearly demonstrated by the adoption of the astronomical timescale as a standard for the Plio-Pleistocene in the most recent timescales. Future research aims to refine the anchored APTS for Neogene marine and continental successions, extend it through the Paleogene into the Mesozoic, develop floating astronomical timescales for parts of the Mesozoic and Paleozoic, and intercalibrate radioisotopic and astronomical time. The APTS will be increasingly applied to research into seafloor-spreading histories, magnetic reversals and excursions, biochronology and evolution, and paleoclimate and paleoceanography, including the important link to climate modeling.
We thank the coworkers of the MIOMAR and CoMCoM research projects for their contributions to this article. The Miocene Marine Archives Reading project received financial support from the EU Human Capital and Mobility Program. The Continental Marine Correlations in the Mediterranean Programme and continental research in Spain is supported by the Netherlands Geosiences Foundation with financial aid from the Netherlands Organization of Scientific Research.
DeMets, C., R. G. Gordon, D. F. Argus, and S. Stein, Effect of recent revisions to the geomagnetic reversal timescale on estimates of current plate motion, Geophys. Res. Lett., 21 , 2191, 1994.
Fischer, A.G., T.D. Herbert, G. Napoleone, I. Premoli Silva, and M. Ripepe, Albian pelagic rhythms (Piobbico core), J. Sediment. Petrol., 61 , 1164, 1991.
Hilgen, F. J., Extension of the astronomically calibrated (polarity) timescale to the Miocene/Pliocene boundary, Earth Planet. Sci. Lett., 107 , 349, 1991.
Hilgen, F. J., W. Krijgsman, C. G. Langereis, L. J. Lourens, A. Santarelli, and W. J. Zachariasse, Extending the astronomical (polarity) time scale into the Miocene, Earth Planet. Sci. Lett., 136 , 495, 1995.
Lourens, L. J., A. Antonarakou, F. J. Hilgen, A. A. M. Van Hoof, C. Vergnaud-Grazzini, and W. J. Zachariasse, Evaluation of the Plio-Pleistocene astronomical timescale, Paleoceanography, 11 , 391, 1996.
Olsen, P. E., D. V. Kent, B. Cornet, W. K. Witte, and R. W. Schlische, High-resolution stratigraphy of the Newark rift basin (early Mesozoic, eastern North America), Geol. Soc. Am. Bull., 108 , 40, 1996.
Renne, P. R., A. L. Deino, R. C. Walter, B. D. Turrin, C. C. Swisher, T. A. Becker, G. H. Curtis, W. D. Sharp, and A. R. Jaouni, Intercalibration of astronomical and radioisotopic time, Geology, 22 , 783, 1994.
Shackleton, N. J., A. Berger, and W. R. Peltier, An alternative astronomical calibration of the lower Pleistocene timescale based on ODP Site 677, Trans. R. Soc. Edinburgh Earth Sci., 81 , 251, 1990.
Ten Kate, W. G. H. Z., and A. Sprenger, Orbital periodicities above and below the Cretaceous/Paleogene boundary at Zumaya (N Spain), Agost and Relleu (SE Spain), Sediment. Geol., 87 , 69, 1993.
Wilson, D. S., Confirmation of the astronomical calibration of the magnetic polarity time scale from sea-floor spreading rates, Nature, 364 , 788, 1993.
Back to Science for Everyone
![Figure 2]](#eos96336f2.gif){kind=link}