Ocean research spans an extraordinary range of activities, from mapping the seafloor with sonar to collecting water samples for traces of DNA shed by deep-sea creatures no one has ever seen alive. As of the mid-2020s, only about a quarter of the global seafloor has been directly mapped, observed, or explored, which means the majority of Earth’s solid surface remains essentially unknown territory. That single statistic helps explain why ocean research projects continue to turn up surprises: new species, unexpected pollution pathways, ice-melt processes that upend sea-level forecasts, and ecosystems so fragile that disturbing them could erase biodiversity that took millions of years to develop.
How Much of the Ocean Floor Has Actually Been Mapped
The figure most often cited is that roughly a quarter of the seafloor has been mapped using modern bathymetric methods. A 2025 review of mapping techniques confirmed that only about a quarter of the seabed has been directly observed, largely because traditional detection methods struggle with the sheer scale and depth of the ocean.1Intelligent Marine Technology and Systems. Techniques and methods for seafloor topography mapping: past, present, and future But that global average hides wide variation. The North Atlantic and North Pacific are the best-covered, with over a third of each ocean mapped by the end of 2024. The Arctic, South Atlantic, and Southern Oceans hover around 30 percent coverage. The Indian Ocean lags behind at roughly 17.5 percent.2Frontiers in Marine Science. Status of global seafloor mapping effort and priority areas for future mapping
Coverage also varies by depth. About 40 percent of the seafloor in the middle depth ranges, between 200 and 3,000 meters, has been mapped. Shallow coastal zones and the vast abyssal plains deeper than 3,000 meters each sit at roughly a quarter coverage. One unexpected finding: the hadal zone, the trenches deeper than 6,000 meters, actually has the highest proportional mapping coverage at about 56 percent. That zone makes up only about one percent of the total ocean floor, so it represents a small absolute area, but researchers have deliberately targeted these extreme environments because of their scientific interest.3Frontiers in Marine Science. Status of global seafloor mapping effort and priority areas for future mapping
Closing the remaining gaps is not just a cartographic exercise. Without detailed seafloor maps, researchers cannot model ocean currents accurately, identify habitats worth protecting, or plan safe routes for undersea cables. The Seabed 2030 initiative, a collaboration launched in 2017, has set the ambitious goal of mapping the entire ocean floor by the end of the decade, but progress depends on ship time, funding, and the development of faster survey technologies.
New Species Turning Up Faster Than Scientists Can Name Them
One of the most consistent findings across ocean research projects is that deep-sea biodiversity is vastly underestimated. When researchers trawl the abyssal seafloor, a startling proportion of what comes up in the net has never been formally described. A large survey of the lower bathyal and abyssal fauna off eastern Australia identified over 1,000 distinct taxa from roughly 25,700 specimens. Of the species-level identifications, up to 58 percent appeared to be undescribed, meaning they had no formal scientific name.4Marine Biodiversity Records. The lower bathyal and abyssal seafloor fauna of eastern Australia
That pattern repeats across taxonomic groups and ocean basins. A census of abyssal polychaetes, a group of segmented worms that are among the most common deep-sea animals, assembled over 3,600 records and found 768 species, 358 of which were new to science. Those records represented two centuries of collecting effort.5Deep Sea Research Part II: Topical Studies in Oceanography. A census of abyssal polychaetes Even relatively well-studied groups like bivalves keep yielding novelties. Three new species of the clam genus Axinulus were recently described from the Kuril-Kamchatka and Japan trenches at depths between 3,200 and 9,583 meters, and the researchers noted that most deep-sea members of that family remain unidentified.6PubMed Central. Three new species of Axinulus (Bivalvia: Thyasiridae) from the Japan and Kuril-Kamchatka trenches and abyssal zone of the northern Pacific Ocean
The sheer number of undescribed species makes it difficult to set conservation baselines. You cannot protect what you do not know exists, and the pace of taxonomy, always slow and underfunded, is struggling to keep up with what expeditions bring back.
Environmental DNA as a Deep-Sea Survey Tool
Traditional deep-sea surveys require physical collection of specimens, which is expensive and disturbs the habitat. Environmental DNA, or eDNA, offers a complementary approach: organisms constantly shed genetic material into the water and sediment around them, and analyzing that material can reveal which species are present without ever catching or seeing them. Researchers working off southern Africa used sediment samples and eDNA metabarcoding to recover 444 distinct genetic units across a wide array of species, making it the first such study of deep-sea benthic environments in that region.7Marine Biology. Extending deep-sea benthic biodiversity inventories with environmental DNA metabarcoding
The method has limits, though. Sampling water above the seabed misses much of the bottom-dwelling community. A study comparing sediment and water sampling found that eDNA from water cannot comprehensively survey benthic communities, even when large volumes are collected.8Scientific Reports. Evaluating sediment and water sampling methods for the estimation of deep-sea biodiversity using environmental DNA Sediment grabs capture a richer picture, but they still depend on the sample location and timing.
One creative workaround involves sponges. Deep-sea sponges naturally filter large volumes of water and accumulate DNA from organisms in their environment. Researchers analyzing sponge tissue as a “natural sampler” detected diverse metazoan communities, including both benthic and pelagic organisms, across different latitudes and depths in the North Atlantic. The approach proved effective for evaluating community shifts across large spatial scales.9Communications Biology. North Atlantic deep-sea benthic biodiversity unveiled through sponge natural sampler DNA Essentially, the sponges had already done the water-sampling work. This kind of inventive thinking is characteristic of a field where every data point costs a small fortune to collect.
How the Ocean Stores Carbon
The ocean absorbs a significant fraction of the carbon dioxide that human activity pumps into the atmosphere, and a large share of that removal depends on biological processes collectively called the biological carbon pump. Research quantifying these pathways found that the gravitational settling of organic particles, mostly zooplankton fecal pellets and clumps of dead phytoplankton, accounts for about 70 percent of the total carbon export from the surface to the deep ocean. Vertically migrating animals that feed at the surface and then descend contribute roughly 10 percent, and physical mixing handles the remaining 20 percent.10Global Biogeochemical Cycles. Quantifying the Carbon Export and Sequestration Pathways of the Ocean’s Biological Carbon Pump
These pathways differ in how long they lock carbon away. The gravitational and migrant pumps store carbon for an average of about 140 to 150 years, while the mixing pump keeps it sequestered for only around 50 years.11Global Biogeochemical Cycles. Quantifying the Carbon Export and Sequestration Pathways of the Ocean’s Biological Carbon Pump That distinction matters for climate modeling. If warming waters or changing ecosystems shift the balance from gravitational settling toward mixing-driven export, the ocean’s ability to store carbon on century timescales could weaken right when we need it most.
Microplastics Have Reached the Deepest Seafloor
Research on ocean pollution has moved well beyond tracking floating garbage patches. Deep-sea sediments turn out to be a major resting place for microplastic debris, particularly fibers. Sampling across the Atlantic, Mediterranean, and Indian Oceans found that microplastic concentrations in deep-sea sediments were up to four orders of magnitude higher, per unit volume, than in contaminated surface waters. The researchers described the deep-sea floor as a “large and hitherto unknown repository” for microplastics.12PubMed Central. The deep sea is a major sink for microplastic debris
Where those microplastics end up is not random. Deep-ocean currents driven by differences in water temperature and salinity, known as thermohaline currents, actively sort and concentrate plastic particles. A study of the seafloor in the Tyrrhenian Sea found that these bottom currents create microplastic hotspots with the highest concentrations reported for any seafloor setting: 190 pieces per 50 grams of sediment. The same currents that deliver oxygen and nutrients to deep-sea ecosystems also deliver plastic, meaning that biodiversity hotspots on the seafloor are likely microplastic hotspots as well.13PubMed. Seafloor microplastic hotspots controlled by deep-sea circulation That overlap is a grim coincidence, linking the richest biological communities to the greatest contamination loads.
Mapping the Underside of Antarctic Ice Shelves
One of the most pressing questions in climate science is how fast Antarctic ice shelves are melting from below, and by what mechanisms. Ocean research projects using autonomous underwater vehicles, or AUVs, are providing the first detailed observations of these otherwise inaccessible environments. AUVs were among the earliest tools deployed beneath ice shelves; early missions beneath Fimbul Ice Shelf revealed a spatially complex oceanographic environment, an ice base with widely varying roughness, and water periodically warm enough to accelerate melting.14Geophysical Research Letters. Measurements beneath an Antarctic ice shelf using an autonomous underwater vehicle
More recently, high-resolution sonar mapping of the underside of Dotson Ice Shelf in West Antarctica revealed a whole catalog of melt features that researchers had not anticipated. Convection and occasional intrusions of warm water form terraced structures in quiet areas, while shear-driven turbulence rapidly erodes smooth surfaces and carves teardrop-shaped indentations created by rotating boundary-layer flow. Full-thickness fractures modified by melting were observed throughout the survey area.15PubMed Central. Swirls and scoops: Ice base melt revealed by multibeam imagery of an Antarctic ice shelf The fact that all of these processes were found active under a single ice shelf underscores how much variety exists in basal melting, and how incomplete earlier models were. Other deployments, such as AUV surveys beneath the Sørsdal Ice Shelf in East Antarctica, have confirmed that each shelf has its own unique thermal profile and melt regime, making generalized predictions risky.16Journal of Geophysical Research: Oceans. Cold Ocean Cavity and Weak Basal Melting of the Sørsdal Ice Shelf Revealed by Surveys Using Autonomous Platforms
Mesophotic Reefs Are Not the Safe Haven People Hoped
Coral reefs in the “twilight zone,” at depths of roughly 30 to 150 meters, are known as mesophotic coral ecosystems. For years, a popular hypothesis held that these deeper reefs might serve as climate refuges, surviving warm events that bleach their shallow-water counterparts and eventually reseeding them with larvae. The logic seemed straightforward: deeper water is cooler, so it should buffer corals against heat stress.
That logic turns out to be wrong, or at least far too simple. Research in the Caribbean found that mesophotic reefs actually have lower bleaching thresholds than shallow reefs. Over two thermal stress events, bleaching at depth was driven by a threshold that declined by about 0.26°C for every additional 10 meters of depth. In other words, corals adapted to cooler conditions are stressed by smaller temperature increases than their shallow-water relatives would tolerate.17PubMed. Caribbean mesophotic coral ecosystems are unlikely climate change refugia Any increase above local mean warm-season temperatures can trigger bleaching, regardless of the absolute temperature.
A broader review of mesophotic reef science reached similar conclusions. These reefs show distinct community patterns that segregate shallow-reef species from ecologically unique deeper communities, including endemic species found nowhere else. While some locations show limited refuge potential, the pattern is site-specific at best. Mesophotic reefs are susceptible to many of the same threats degrading shallow reefs, and researchers have argued they need their own conservation strategies rather than being treated as backup copies of shallower ecosystems.18PubMed. Function and stability of mesophotic coral reefs
What Deep-Sea Mining Would Cost Ecosystems
Polymetallic nodules, potato-sized lumps of metal ore scattered across vast stretches of abyssal seafloor, have attracted growing commercial interest. Mining them would mean vacuuming up the top layer of sediment across enormous areas. The ecological stakes are becoming clearer as research accumulates.
The nodules themselves are habitat. A study in the Clarion-Clipperton Zone of the central Pacific found that many abyssal species, including certain corals and other sessile organisms, occur exclusively on these hard surfaces. Removing the nodules would eliminate that habitat, and because nodules grow only a few millimeters per million years, recovery in any meaningful human timeframe is not an option.19Scientific Reports. Threatened by mining, polymetallic nodules are required to preserve abyssal epifauna The concern extends beyond the animals living directly on the nodules. Research modeling food webs in two nodule-rich areas, the Clarion-Clipperton Zone and the Peru Basin, found that removing nodules would trigger cascading effects through the food web. The losses resulted overwhelmingly from non-trophic interactions: sessile organisms attached to the nodules support other fauna that depend on them for shelter and structure, not just for food. Removing the nodules collapses these interaction networks and leads to broad biodiversity loss.20Scientific Reports. Polymetallic nodules are essential for food-web integrity of a prospective deep-seabed mining area in Pacific abyssal plains
Recent work has further emphasized that the heterogeneity of the habitat, the mix of nodule patches, bare sediment, and transitional zones, is itself what sustains community diversity. Understanding how that patchwork functions is considered critical for developing environmental management plans if mining proceeds.21Frontiers in Marine Science. The influence of habitat heterogeneity and disturbance on benthic community structure in deep-sea polymetallic nodule environments and management implications for seabed mining
Ocean Heat and Hurricane Intensity
The relationship between upper-ocean heat and tropical cyclone strength is an active area of research with direct implications for storm forecasting. Warmer surface waters feed energy to developing hurricanes, but the standard measure of ocean heat content does not capture the full picture. Research has shown that incorporating the strength of the thermocline, the sharp temperature boundary sitting below the mixed layer, improves correlations with hurricane intensity changes. A modified heat-content variable that accounts for this stratification tracked hurricane behavior better than simple anomalies alone, as demonstrated in the case of Hurricane Juliette in 2001.22Monthly Weather Review. Oceanic Heat Content Variability in the Eastern Pacific Ocean for Hurricane Intensity Forecasting
A study examining the northwest Pacific found that upper-ocean warming alone could not fully explain trends in extremely intense tropical cyclones. A newer formulation of heat content appeared to have a closer relationship with cyclone activity than the conventional measure, suggesting that the vertical structure of ocean warmth matters as much as its magnitude.23Hurricane Risk. Is the Increasing Trend in the Environmental Upper-Ocean Heat Content in the Northwest Pacific Strengthening Tropical Cyclone Intensity? For forecasters, the takeaway is that sea-surface temperature maps are not sufficient. Predicting how much a storm will strengthen over a patch of ocean requires knowing how that warmth is distributed with depth.
Robots Built for Full Ocean Depth
Exploring the deepest parts of the ocean requires vehicles that can withstand crushing pressure. A growing area of engineering research focuses on making underwater gliders and autonomous vehicles more energy-efficient at extreme depths. One recent development is a buoyancy engine designed for full-ocean-depth operation, where pressures reach about 1,134 bar. Engineers developed a dual-stage pump system that reduced internal leakage by 58 percent and improved overall efficiency by 25 percent compared to conventional designs. An adaptive pumping strategy cut energy consumption per dive profile by over a third.24Ocean Engineering. A Novel ultra-high-pressure buoyancy engine for enhanced energy efficiency of full-ocean-depth underwater vehicles The system was integrated into an operational deep-sea glider, extending its capacity for long-duration observation missions in the abyss.
These engineering advances matter because the bottleneck for deep-ocean science is often just getting instruments to the right place for long enough. Traditional crewed submersibles are expensive and brief. Autonomous gliders and AUVs that can stay deployed for weeks or months, profiling the water column or surveying the seabed, dramatically expand what a single research expedition can accomplish.
Hydrothermal Vents and Bioprospecting
Hydrothermal vents, where superheated water laced with dissolved minerals gushes from cracks in the seafloor, have been a feature of Earth’s oceans since the planet’s earliest days. The ecosystems around them run on chemosynthesis rather than sunlight: microbes harness the chemical energy in vent fluids to fix carbon, and entire food chains build on that microbial productivity. The animal communities at vents tend to be low in species diversity but high in abundance, dominated by specialists found nowhere else.
Beyond their ecological interest, vent microbes are attracting attention for biotechnology. Deep-sea microorganisms produce enzymes that function under extreme pressures, temperatures, and chemical conditions, properties that make them useful for industrial processes, pharmaceutical discovery, and environmental remediation. A review of deep-sea microbial genetic resources highlighted the vast untapped diversity of these organisms and their potential for bioprospecting, while also flagging climate change and deep-sea mining as threats to the genetic resources themselves.25PubMed Central. Deep-sea microbial genetic resources: new frontiers for bioprospecting The tension between commercial interest in vent organisms and the need to protect their habitats mirrors the broader debate around deep-sea mining.
Listening to the Ocean
Not all ocean research involves looking at or collecting things. Passive acoustic monitoring, or PAM, uses underwater microphones to record and analyze sound. In the Arctic, where shrinking sea ice is opening new shipping lanes and changing the underwater sound environment, PAM-based studies have focused on tracking long-term changes in underwater noise, characterizing the properties of the seabed using sounds from passing ships, and developing automated methods to classify marine mammal vocalizations.26The Journal of the Acoustical Society of America. Korea’s passive acoustic monitoring research on underwater noise and marine mammals in a changing Arctic Ocean International organizations like the Arctic Council have emphasized the importance of this kind of monitoring as human activity in Arctic waters increases.
Acoustics also play a role in sediment core research for reconstructing past climates. Drilling into the seafloor to retrieve cores of layered sediment allows scientists to read the climate record preserved in those layers. A recent study of a core from the South China Sea identified 129 event layers, sudden deposits from submarine landslides or volcanic events, interspersed in 800,000 years of sediment. These instantaneous deposits accounted for about 16 percent of the total core thickness and, when not properly accounted for, significantly distorted the age model and paleoclimate reconstruction. Removing them produced a cleaner, more reliable climate record.27Geochemistry, Geophysics, Geosystems. Event Deposits Recorded Over the Past 800 kyr in IODP Hole U1433A (South China Sea): Implications of a Revised Methodology for Chronology and Paleoclimate Reconstruction
Traditional Knowledge and the Future of Ocean Science
A recurring theme across modern ocean research is that scientific data alone is not enough. Indigenous and traditional communities in coastal regions hold generations of ecological knowledge that formal science has often overlooked. In Fiji, for example, the indigenous iTaukei people have long managed marine resources using detailed understanding of local ecosystems. The Women in Fisheries Network there has demonstrated how empowering women in fisheries decision-making can complement and strengthen scientific knowledge for sustainable management.28Marine Policy. Weaving science and traditional knowledge: Toward sustainable solutions for ocean management
Integrating these perspectives is a stated priority for the UN Decade of Ocean Science for Sustainable Development, which runs through 2030. The challenge is practical: traditional knowledge systems are place-based and qualitative, while the scientific research community runs on data and publications. Bridging the two requires co-production, where local knowledge holders and researchers design questions and interpret results together, rather than one side merely consulting the other. Whether ocean science institutions can actually make that shift will shape how effectively the next wave of ocean research serves the communities most directly affected by what it discovers.