Seafloor sediments, ice cores, and Milankovitch theory: paleoclimate in harmony

Eccentricity, obliquity, and precession: the longest seasons on Earth

At least since the days of Isaac Newton and Johannes Kepler, astronomers have recognized that the Earth does not orbit the sun along a perfectly circular path. Nonetheless, these pioneering physicists developed elegant laws to describe that orbit mathematically. Using these laws, one could calculate precisely where the Earth would be in relationship to the sun six months in the future, for example, or even eight months in the past. Today, these same laws aid in space exploration to launch satellites to the edge of our solar system, landing them perfectly in orbit with planets like Saturn or on the surface of Mars.

In addition to recognizing that Earth’s orbit was slightly elliptical, astronomers also observed that gravitational interactions with neighboring planets could affect our course. The internal dynamics of our spinning globe, moreover, affected the extent to which Earth’s axis tilted toward the sun (the phenomenon responsible for annual seasons). Centuries after Newton and Kepler, a mathematician by the name of Milutin Milankovitch would combine this complex set of physical interactions into an elegant theory that changed forever how we study Earth history. Milankovitch proposed that over long time scales, on the order of tens of thousands of years, Earth’s orbit would change shape slightly, but in a predictable and cyclical fashion.

Why is this important? Simply put, the shape of Earth’s orbit determines the amount of energy it receives from the sun. It determines the mean position of the equator, the temperature gradient from equator to pole, and the climatic disparity between winter and summer. At one extreme, Earth would experience abnormally hot summers and cold winters in the northern hemisphere, for example; at the other, relatively cool summers and warm winters. According to Milankovitch’s hypothesis, these long-term effects should therefore drive climatic changes all across the Earth, similar to modern seasonal cycles. And just like Kepler’s laws of planetary motion, we could predict exactly when in Earth history each ‘season’ would have dominated.

With the advent of modern geochronology and paleoclimatology, geologists began to reconstruct Earth’s climate over time, and Milankovitch’s ideas were put to the test. Utilizing all sorts of proxies for ancient variations in temperature, sea-level, vegetation, and ice sheets, study after study (e.g. Hays et al., 1976) would confirm that Earth’s climate varied predictably on regular timescales. Milankovitch’s hypothesis thus became Milankovitch Theory, and geology forever changed.

Milankovitch cycles are perhaps best known in cores from ocean sediments and glacial ice. The stable-isotope chemistry of both marine fossils and ice layers, for example, are excellent proxies for temperature and weather patterns, as well as ocean circulation. These data fluctuate on the same periodicity as solar insolation, as predicted by Milankovitch, providing multiple independent lines of supporting evidence. The image below depicts how even greenhouse gases spike every ~100,000 years or so, causing ice ages to be interrupted by a warm period.

Greenhouse gas and oxygen-isotope data from the Vostok Ice Core in Antarctica, alongside temperature variations (red). Bottom line (orange) depicts fluctuations in solar insolation (amount of energy reaching the Earth from the Sun). Image from Wikimedia commons.

Orbital tuning: how Milankovitch cycles help calibrate our clocks

Geologists have developed several methods by which to date ocean sediments and ice cores, but none are perfect and all have their limitations. Let’s begin with ocean sediment cores. Deep-ocean sediments accumulate at very low rates: only a few centimeters every 1,000 years, on average. We can determine how fast these sediments accumulate either by 1) observing the modern rate and extrapolating back in time, or 2) dating sediment layers directly and dividing the depth of those layers by their age. The first approach works fine, so long as the rate has not varied considerably over time and we can estimate how sediments are compacted with burial. This assumption is safe in most cases, due to the fact that deep oceans are far removed from influencing factors like rivers, which discharge tons of sediment into the water. Deep ocean sediments consist mainly of microfossils (e.g. foraminifera, radiolarians, and diatoms), organic matter, and fine particles of silt and clay, all of which can take hundreds of years to settle out of the water column. Only major perturbations to ocean circulation or Earth’s climate will have much impact, therefore, on how fast these particles accumulate.

Since ocean sediments contain abundant organic matter, they can be dated by the radiocarbon method at least back to ~50,000 years. This allows geologists to determine how much sedimentation rate has varied in response to the transition from last ice age to the current warm period (a major climate shift). In many cases, the rate is relatively constant, justifying the use of what is called a Constant Sedimentation Rate (CSR) age model. Geologists use CSR age models to predict the age of underlying layers, based on the assumption that sedimentation rate was similar in the past.

 At this point, many young-Earth creationists would be quick to highlight the word assumption. But as I’ve stated before, assumptions by themselves are not detrimental to historical science—rather, they are a necessary part of the process. Geologists likewise are not satisfied by resting their conclusions in an unverified assumption, even if they have good reason to make it. But we can test the assumption of constant sedimentation rate using other tools employed to date ocean sediments.

Layers of tephra regularly occur in ocean sediments, due to large volcanic eruptions periodically dumping tons of ash into the oceans. These ash layers can be dated directly using the Argon-Argon (Ar-Ar) method, or at least ‘fingerprinted’ geochemically and correlated to known volcanic deposits on land (which themselves are dated by the Ar-Ar method). Tephrochronology allows paleoceanographers to construct precise chronologies through a completely independent method, providing either verification or falsification of the Constant Sedimentation Rate age model (e.g. Ross et al., 2012). More importantly, the Ar-Ar method can be applied to much older marine sediments, extending chronologies back many millions of years. In the absence of tephra layers, ocean sediments can be dated directly using isotopes of beryllium (Bourles et al., 1989), with varying degrees of success.

Another test utilizes known reversals in Earth’s magnetic field. These polarity reversals are recorded in ocean basalts on the seafloor, which can be dated directly by several radiometric techniques.  Every time the magnetic alignment of particles changes in a sediment core, we can identify the exact age of that layer. If the paleomagnetic age matches the age predicted by the CSR age model, then we know that sedimentation rate was relatively constant. Otherwise, the CSR model is falsified for that particular sediment core, but in return, we gain a better estimate of long-term sedimentation rate.

Finally, Milankovitch Theory helps determine and/or refine the age of ocean sediments, in which cyclic variations in chemistry have been recorded. If we analyze marine microfossils, for example, the ratio of magnesium to calcium (Mg/Ca) and stable isotopes of oxygen (δ18O) shift in response to sea level and temperature. Since sea level and temperature vary with Milankovitch cycles (called the “pacemaker of the ice ages”), we can predict when these ratios should be either high or low. Just as we can use Kepler’s laws to determine how many months ago the northern hemisphere experienced the peak of winter, causing thermometers to plummet, we can use Milankovitch Theory to determine how long ago Earth experienced glacial conditions, causing geological thermometers to plummet. Applying this technique (called orbital tuning), paleoceanographers have constructed a timeline of ocean δ18O extending back 5.3 million years (Lisiecki and Raymo, 2005).

Compilation of ocean sediment cores, in which δ18O (vertical axis) was analyzed in benthic foraminifera. The δ18O value is related to global ice volume (on land), which fluctuates in response to solar energy reaching the Earth. These records are combined through rigorous statistical analysis, to prevent artificial correlation of random patterns. Image credit: figure 2 from Lisiecki and Raymo (2005).

By combining dozens of records, Lisiecki and Raymo (2005) did not simply have to assume a constant sedimentation rate. Instead, they used a far more precise timekeeper: variations in Earth’s orbit due to Milankovitch cycles. The fact that all sediment records yield the same pattern in δ18O confirmed that each was recording a global shift in ocean chemistry. If the sedimentation rate varied significantly at any one site for a given interval, then it would ‘stand out’ among the other records and could be calibrated to them. Therefore, the authors were able to reconstruct how the global average sedimentation rate varied over time. Note that over the past 5.3 million years, the sedimentation rate did not deviate far from ~4 centimeters/1,000 years. So it turns out, the assumption behind CSR age models for these data would almost be justified.

From Lisiecki and Raymo (2005); reconstructed sedimentation rate, averaged between all sediment cores.

Telling time in a room of imperfect clocks

Hopefully, this brief but incomplete overview provides you with a better sense of how the last few million years of Earth history are recovered from the deep ocean, as well as the limitations of those efforts. Paleoceanographers have worked tirelessly over the years to provide climatologists with better correlations and improved age models for their data. As a result, our understanding of Earth’s climate during the Pliocene and Pleistocene has only been refined. Despite that each dating method carries inherent challenges, the abundance of clocks aids geologists in getting all the ‘ticks’ and the ‘tocks’ to line up just right. But how do we know which clock is correct, and could they all be wrong? In the next post, I’ll consider these questions in light of a direct challenge to the independence of geological dating methods, which have been used to construct age models for ocean sediment cores.

 

 

Featured image reproduced from EAPS at MIT.