Sediment cores are cylinders of mud, sand, and organic material pulled from the bottom of oceans, lakes, and wetlands, and they function as some of the most detailed natural archives on Earth. Layer by layer, particles settle and accumulate over time, trapping everything from microscopic fossils to volcanic ash to industrial pollutants. By drilling or pushing a tube into these deposits and extracting an intact column of sediment, scientists can read backward through centuries, millennia, or even millions of years of environmental history. The catch is that extracting a clean record from what is essentially a tube of wet dirt requires a surprising amount of technology, careful dating, and awareness of the many ways the signal can be distorted.
How Cores Are Collected
The simplest approach is a gravity corer, which is essentially a weighted tube lowered from a ship or platform and allowed to free-fall into the sediment. Gravity corers work well in soft muds and can recover several meters of material, but they tend to under-sample the deeper portions of a core because frictional resistance slows penetration as the tube goes deeper. Piston corers solve part of this problem by using an internal piston that creates suction, helping pull sediment into the tube and allowing recovery of longer sections. However, piston corers come with their own headaches. A study comparing four cores taken from the same spot on the Iberian Margin found that the recovered sequences differed in length by a factor of up to 2.7, entirely because of how each coring method distorted the stratigraphy. Cable rebound during piston coring caused the upper portions to be “over-sampled” (stretched), while gravity cores under-sampled the base. Heavier piston corers on longer cables made the problem worse, and cable rebound sometimes caused gravity corers to penetrate the same interval twice, creating repeated stratigraphic sections in the recovered core.1Marine Geology. Analysis and modelling of gravity- and piston coring based on soil mechanics
For deeper targets or harder substrates, scientific drilling programs use rotary drilling rigs mounted on specialized ships. The International Ocean Discovery Program and its predecessors have drilled hundreds of sites worldwide, sometimes penetrating over a kilometer below the seafloor. On land, similar rigs bore into lake beds and ancient sedimentary basins. The choice of tool shapes what the scientist ultimately sees in the lab, and an awareness of coring artifacts is baked into every serious study.
What Happens in the Lab
Once a core arrives at a research facility, it is typically split lengthwise. One half becomes the working half, available for sampling and destructive analysis; the other is archived. Before anyone takes a scalpel to it, though, non-destructive scanning techniques can extract a remarkable amount of information while the core is still intact.
X-ray fluorescence (XRF) scanning measures the chemical composition of the sediment at intervals as fine as half a centimeter, producing a detailed elemental profile down the entire length of the core. Computed tomography (CT) scanning goes further, generating three-dimensional images of the core’s internal structure at sub-millimeter resolution. When researchers applied both techniques to cores from a lagoon and a marsh in western Greece, they found that CT-derived density values correlated with the concentrations of land-derived elements measured by XRF, and three-dimensional reconstructions revealed microstructural details and fossil distributions that would have been invisible to the naked eye.2Revue de Micropaléontologie. CT scanning, X-ray fluorescence: Non-destructive techniques for the identification of sedimentary facies and structures The same pairing of methods has been used on ocean-drilling cores to distinguish between glacial and non-glacial sediment types, using elemental abundances alongside physical properties like color reflectance and magnetic susceptibility.3Geosphere. Multivariate modeling of glacimarine lithostratigraphy combining scanning XRF, multisensory core properties, and CT imagery: IODP Site U1419
Dating the Layers
A core without a timeline is just a tube of mud. Assigning ages to the layers is often the hardest part of the whole enterprise, and different time scales call for different tools.
For relatively recent sediments covering the last century or so, researchers commonly rely on two fallout radionuclides. Cesium-137 entered the environment through nuclear weapons testing in the 1950s through 1970s and from accidents like Chernobyl in 1986, creating recognizable spikes in the sediment record. Excess lead-210, a naturally occurring isotope with a half-life of about 22 years, decreases predictably with depth. The two are often combined to build a chronology for the past 150 years or so.4Earth System Science Data. A worldwide meta-analysis (1977–2020) of sediment core dating using fallout radionuclides including 137Cs and 210Pbxs In fast-accumulating settings like flood-control reservoirs, several lead-210 dating models can be tested against each other and cross-checked with cesium-137 peaks to pick the best fit.5Soil Science Society of America Journal. Dating Sediment in a Fast Sedimentation Reservoir using Cesium‐137 and Lead‐210
For older material, radiocarbon dating is the workhorse, but it comes with complications that non-specialists rarely hear about. Carbon in sediment arrives from many sources, and not all of it reflects the true age of the deposit. A study in Chesapeake Bay found that dating the total organic carbon in estuarine sediments gave largely unreliable results because much of the carbon, including coal, had washed in from upstream and was far older than the sediment itself. Mollusk shells and foraminifera gave more dependable ages, though reworking and burrowing by organisms could still cause problems.6Quaternary Research. Radiocarbon Dating, Chronologic Framework, and Changes in Accumulation Rates of Holocene Estuarine Sediments from Chesapeake Bay
Marine radiocarbon dating adds another layer of difficulty: the reservoir effect. Ocean water is isotopically “old” compared with the atmosphere because deep water has been out of contact with atmospheric carbon dioxide for centuries. The standard correction for modern surface waters is roughly 400 years, but simulations covering the past 50,000 years show that surface-ocean reservoir ages have varied between about 400 and 1,200 radiocarbon years at low to mid latitudes, reaching around 780 years at the Last Glacial Maximum and exceeding 2,000 years in polar oceans.7Geophysical Research Letters. Marine radiocarbon reservoir age simulations for the past 50,000 years Applying a single fixed correction across an entire core can seriously distort the timeline. At a site on the south coast of South Africa, researchers found that the reservoir offset changed through the Holocene as the connection between a coastal lake and the ocean shifted over time. They solved this by dating paired samples of wood and bulk organic sediment from the same depths, using the mismatch to calculate how the reservoir effect changed.8Quaternary Geochronology. The impact of changing reservoir effects on the 14C chronology of a Holocene sediment record from South Africa
For material older than radiocarbon’s useful range of roughly 50,000 years, researchers turn to methods like oxygen isotope stratigraphy, paleomagnetic reversals, and radiometric dating of volcanic ash or zircon crystals. These can be combined: a study of Late Pliocene marine sediments in central Japan used zircon uranium-lead dating alongside oxygen isotope curves and magnetic polarity reversals to build a composite age framework.9Newsletters on Stratigraphy. Oxygen isotope-magnetostratigraphy combined with zircon U–Pb dating for the Upper Pliocene composite marine succession in central Japan
Reading Past Climates From Microfossils and Chemistry
The shells of tiny marine organisms called foraminifera are among the most powerful climate recorders preserved in ocean sediments. Their calcium carbonate shells incorporate oxygen in a ratio of heavier to lighter isotopes that depends on both the temperature of the water and the global volume of ice locked up on land. Separating those two signals has been a long-standing challenge. Work on the mid-Pleistocene transition, the interval when ice-age cycles shifted from roughly 40,000-year to 100,000-year spacing, showed that both deep-ocean cooling and increasing ice volume contributed to the growing amplitude of oxygen isotope swings in foraminiferal records.10PubMed. Evolution of ocean temperature and ice volume through the mid-Pleistocene climate transition A newer approach uses “clumped isotopes,” a measurement that depends only on temperature, to help pull the ice-volume signal apart from the temperature signal.11Paleoceanography and Paleoclimatology. Revisiting Oxygen‐18 and Clumped Isotopes in Planktic and Benthic Foraminifera Benthic foraminifera living on the seafloor provide records of deep-water temperature and ice-volume changes that are foundational to our understanding of past climate.12Quaternary Science Reviews. A review of benthic foraminiferal oxygen and carbon isotopes
Organic molecules preserved in sediments offer independent temperature estimates. The TEX86 index is based on membrane lipids produced by a group of marine archaea, and it has become one of the most widely used proxies for past sea-surface temperature in climates warmer than today’s. A 15-million-year TEX86 record from the eastern equatorial Atlantic highlighted a complication: the lipids are not produced exclusively in the surface mixed layer, so the resulting temperature signal can reflect a depth-integrated average that is sensitive to subsurface changes, not just the sea surface.13Climate of the Past. A 15-million-year surface- and subsurface-integrated TEX86 temperature record from the eastern equatorial Atlantic Comparing TEX86 with other organic temperature proxies like the alkenone unsaturation index in Australian waters confirmed that TEX86 best matches annual temperature in the upper 200 meters of the water column, while alkenone-based estimates track surface winter temperatures more closely.14Organic Geochemistry. Comparison of U37K′, TEX86H and LDI temperature proxies for reconstruction of south-east Australian ocean temperatures Running multiple proxies in parallel lets researchers triangulate real temperatures rather than relying on any single imperfect recorder.
Pollen, Ancient DNA, and Ecosystem Reconstruction
Not every useful signal in a core comes from the ocean. Lake and wetland cores preserve pollen grains in extraordinary detail, and because different plant species produce distinctively shaped pollen, counting and identifying grains down a core tells you how the vegetation around the site changed over time. A pollen record from Lake Ladoga in northwestern Russia documented 13,900 years of vegetation shifts, from grass and shrub tundra through spruce forests to the birch-dominated landscape marking the onset of Holocene warming, with the Younger Dryas cold snap clearly visible as a return to tundra-steppe conditions around 12,600 years ago.15Boreas. Vegetation and climate changes in northwestern Russia during the Lateglacial and Holocene inferred from the Lake Ladoga pollen record At Lake Ohrid in southeastern Europe, a pollen sequence covering the past 500,000 years showed a clear correspondence between forested and non-forested periods and the glacial-interglacial cycles recognized in the marine oxygen isotope record, demonstrating that major climate rhythms leave consistent signatures on land and at sea.16Biogeosciences. Pollen-based paleoenvironmental and paleoclimatic change at Lake Ohrid (south-eastern Europe) during the past 500 ka
A continent-wide analysis of North American pollen records added a striking modern twist: the rate of abrupt vegetation change across the continent was highest at the end of the last ice age, gradually declined through the Holocene, and then surged to its highest levels of the past 6,500 years during the last two centuries. The recent changes driven by human land use are comparable in magnitude to the ecological upheaval that accompanied the transition from a glacial to an interglacial world.17PubMed Central. North American pollen records provide evidence for macroscale ecological changes in the Anthropocene
Pollen has one major limitation: it only captures organisms that produce pollen. Sedimentary ancient DNA, extracted directly from the mud, can detect organisms that leave no fossil trace at all. By combining two genetic marker genes, one study recovered 91 distinct eukaryotic taxa from marine sediment samples, compared with 59 and 51 from each marker alone, and roughly 17% of the combined genetic material came from eukaryotes, including species that do not fossilize.18Oceanography. The Potential of Sedimentary Ancient DNA to Reconstruct Past Ocean Ecosystems This technique is still young, but it promises to fill in parts of the past that fossils and pollen alone cannot reach.
Recording Earthquakes and Tsunamis
Sediment cores are one of the few tools that can extend earthquake records far beyond the age of written history. When a large submarine earthquake shakes the continental slope, it can trigger underwater landslides that send clouds of sediment cascading into the deep ocean. These settle as graded layers called turbidites, with coarser material at the bottom and finer material on top. If the same turbidite appears in multiple cores spread across a wide area, it likely came from a single large event rather than a localized slump.
This approach has been used extensively along the Cascadia subduction zone off the Pacific Northwest. Cores collected along the continental margin record 18 turbidite events correlated over hundreds of kilometers during the past roughly 10,000 years, with 13 of those occurring since the eruption of Mount Mazama about 7,700 years ago. The average recurrence interval is about 600 years, and the youngest event, roughly 300 years ago, matches the great Cascadia earthquake of 1700 known from coastal evidence and Japanese tsunami records.19Annual Review of Earth and Planetary Sciences. Holocene Earthquake Records from the Cascadia Subduction Zone and Northern San Andreas Fault Based on Precise Dating of Offshore Turbidites More recent work on the same margin has shown how earthquake-triggered landslides on the lower slope produce proximal mass transport deposits that grade offshore into complex, interfingered abyssal turbidites, strengthening the connection between the sediment record and megathrust earthquake history.20PubMed Central. Widespread abyssal turbidites record megathrust earthquake-triggered landslides and coseismic deformation in the Cascadia subduction zone
Similar logic applies in other tectonic settings. In western Baffin Bay, Holocene turbidites in a submarine channel include one dated to about 1933 CE that matches a known magnitude 7.4 earthquake, lending confidence that the older turbidites in the same system also record seismic events.21Canadian Journal of Earth Sciences. Holocene earthquake-triggered submarine landslides and turbidites in western Baffin Bay On coastlines, cores from sheltered bays and marshes can preserve sand layers deposited by tsunamis. Drilling to 10 meters depth in a wetland on Sri Lanka’s east coast identified three probable paleo-tsunami deposits predating written records, the oldest roughly 2,700 years old.22Elsevier. Sedimentological observations and geochemical characteristics of paleo-tsunami deposits along the east coast of Sri Lanka in the Indian Ocean
Human Fingerprints in the Mud
Sediment cores do not only record natural processes. A 900-centimeter core from the Danube Delta’s Sulina distributary captured 600 years of heavy metal history. In the lower half, metals like copper, zinc, lead, chromium, and nickel tracked natural variability controlled mainly by grain size. Above roughly 200 centimeters, corresponding to about 1700 CE onward, concentrations of those metals plus cadmium began to climb, reflecting enrichment from human activity.23Science of The Total Environment. A 600 years sediment record of heavy metal pollution history in the Danube Delta
Microplastics are a newer addition to the sedimentary record, and whether they can serve as a time marker for the Anthropocene has sparked real debate. Cores from the Patos-Mirim lagoon system in southeastern South America showed a clear pattern: a basal zone free of microplastics, then an upper contaminated zone about 70 centimeters thick with an increasing trend matching the intensification of local human activity, which the authors dated to around the early 1970s.24Science of The Total Environment. The use of microplastics as a reliable chronological marker of the Anthropocene onset in Southeastern South America But a study of European lake sediments found microplastic particles throughout the cores, including in layers deposited well before 1950, because the particles migrate downward through unconsolidated sediment. The authors concluded that microplastics cannot be used as a time-synchronous marker horizon for the start of the Anthropocene.25PubMed Central. Downward migrating microplastics in lake sediments are a tricky indicator for the onset of the Anthropocene The disagreement likely comes down to site conditions: in compacted or rapidly accumulating sediments, microplastics may stay put, while in loosely consolidated lake beds they sink.
Varved Sediments and Annual Resolution
Most sediment cores offer resolution on the scale of decades or centuries. But in certain lakes, seasonal cycles produce visible alternating layers called varves, analogous to tree rings, that can be counted year by year. In High Arctic lakes where the growing season is short, each varve consists of a coarser layer deposited during summer meltwater input and a finer layer from winter settling. Because the amount of sediment delivered in summer depends on temperature, varve thickness becomes a proxy for summer warmth.
A core from Lower Murray Lake on northern Ellesmere Island, at 81°N, provided a varve record spanning the last millennium.26The Holocene. A record of climate over the last millennium based on varved lake sediments from the Canadian High Arctic Even winter conditions leave a trace: at another High Arctic site, the thickness of nival (snowmelt-related) sediment units correlated with winter temperature and snowfall over the past four centuries. The coldest decades fell between 1800 and 1880, while the fastest warming rates occurred from 1880 to 1930 and again from 1970 to 2010, the latter at about 0.37°C per decade.27Quaternary Science Reviews. Winter temperature conditions (1670–2010) reconstructed from varved sediments, western Canadian High Arctic These records fill in parts of the planet where weather stations have existed for only a few decades, giving climatologists something close to real-time annual data stretching back centuries.
Deep Time Events Recorded in Rock
At the long end of the time scale, ancient sedimentary cores and their outcrop equivalents preserve evidence of global catastrophes. Black shales, dark layers rich in organic carbon, mark intervals when large areas of the ocean became depleted in oxygen, allowing organic matter to accumulate on the seafloor rather than decompose. These oceanic anoxic events have occurred repeatedly through Earth’s history.28PubMed Central. The origin of Cretaceous black shales: a change in the surface ocean ecosystem and its triggers One of the largest, Oceanic Anoxic Event 2 at the boundary between the Cenomanian and Turonian stages about 94 million years ago, left black shale deposits across wide areas of the world ocean and produced a distinctive positive spike in carbon isotope values that researchers use as a fingerprint for the event.29Paleontological Research. Implication of spatiotemporal distribution of black shales deposited during the Cretaceous Oceanic Anoxic Event-2
Perhaps the most famous single layer in the sedimentary record is the iridium-enriched clay marking the end-Cretaceous mass extinction 66 million years ago. Iridium is rare in Earth’s crust but relatively abundant in asteroids, and its elevated concentration at the Cretaceous-Paleogene boundary was the first line of evidence pointing to a hypervelocity impact. Drill core recovered from the peak ring of the Chicxulub impact structure itself confirmed the presence of a positive iridium anomaly within the crater, a finding verified by four independent laboratories.30PubMed Central. Globally distributed iridium layer preserved within the Chicxulub impact structure
When the Record Lies
For all their power, sediment cores are not perfectly faithful recorders. After sediment is deposited, chemical reactions between the grains and the fluids circulating through them can alter the original signals. This process, called diagenesis, is a persistent headache for anyone trying to read ancient ocean chemistry from carbonates. The geochemistry of elements sensitive to oxygen levels can be overprinted by reactions with pore waters whose composition differs from the overlying seawater, making it difficult to distinguish primary signals from secondary alteration.31Geochimica et Cosmochimica Acta. Modeling the impacts of diagenesis on carbonate paleoredox proxies Magnetic minerals are vulnerable too: hydrogen sulfide produced by microbes in methane-rich sediments can dissolve iron-bearing magnetic grains, releasing reactive iron that gets locked into pyrite. This lowers the magnetic susceptibility of the sediment, potentially erasing or distorting a signal that researchers use to track changes in sediment source and delivery.32Geochemistry, Geophysics, Geosystems. Isolating Detrital and Diagenetic Signals in Magnetic Susceptibility Records From Methane‐Bearing Marine Sediments
Bioturbation, the physical mixing of sediment by burrowing organisms, is another major source of blurring. In all but the most oxygen-depleted settings, worms and other creatures churn the upper few centimeters of the seafloor, smearing sharp signals across a broader depth range. Researchers sometimes seek out sites with minimal bioturbation, such as oxygen minimum zones, or use mathematical models to “un-mix” the signal, but neither solution is perfect.
Who Gets to Drill, and Where
Collecting sediment cores from the deep ocean is not just a scientific challenge; it is a legal and diplomatic one. Under the United Nations Convention on the Law of the Sea, coastal states have sovereignty or sovereign rights over the seabed and subsoil extending out to their exclusive economic zone. Any scientific drilling within that zone requires marine scientific research clearance from the relevant country. Drilling in international waters, by contrast, does not require such clearance.33Scientific Drilling. Preparing for the new age of the Nagoya Protocol in scientific ocean drilling The Nagoya Protocol on access and benefit-sharing adds another layer of complexity, because sediment cores can contain genetic resources, including ancient DNA and novel microbial organisms, that fall under biodiversity agreements. International ocean drilling expeditions now routinely navigate both maritime law and biodiversity regulations before a single core is pulled from the seafloor, a process that can add months of lead time to an expedition.