Oceanography

The big picture

On 23 March 1875, in the western Pacific, the crew of HMS Challenger spent most of a day lowering a hemp line with a weight on the end. They were doing what they had already done hundreds of times: measuring how far it was to the bottom. This time the line ran out to 4,475 fathoms, about 8,184 meters, and they wrote the number in the log. They had just found the deepest place any human had measured, and they had done it with rope, lead, and patience. The spot is now called the Challenger Deep in their honor, and modern sonar puts the true depth near 10,935 meters. They were 2.7 kilometers short and still made one of the great measurements in the history of science.

That day is a good picture of what oceanography is. The ocean does not hand over its data. Every number you will use in this course, every temperature, every current speed, every carbon measurement, was pulled out of a hostile, opaque, corrosive, moving fluid by somebody who had to build the instrument first. When you know how a measurement was made, you know how far to trust it, and you know what the next generation of instruments will probably change.

Oceanography is the scientific study of the ocean: its water, its floor, its chemistry, its motion, and its life. It is not one discipline but four woven together, and this course is organized around the weave. What makes ocean science distinctive is not any single method. It is that the object of study is enormous, three-dimensional, in constant motion, mostly dark, and almost entirely out of reach.

The four branches, and why they refuse to stay separate

Textbooks split oceanography into four traditional branches, and it helps to name them, as long as you remember that no real question sits inside only one.

• Geological oceanography studies the shape and history of the ocean floor: bathymetry, sediments, plate tectonics, and the record locked in seafloor cores.
• Chemical oceanography studies what is dissolved in seawater: salts, gases, nutrients, carbon, and the reactions that move them around.
• Physical oceanography studies motion and structure: temperature, density, currents, waves, tides, and the exchange of heat with the atmosphere.
• Biological oceanography studies life in the sea and how it depends on and alters everything above.

Now take one ordinary question and watch the branches collapse into each other. Why is the water off the coast of Peru so extraordinarily rich in fish? The wind blows along the coast, and Earth's rotation turns the surface water offshore, which is physics. Cold water rises from below to replace it, carrying nitrate and phosphate, which is chemistry. Those nutrients feed a bloom of diatoms that feeds anchoveta, which is biology. And the whole system sits above a subduction zone that gives Peru a narrow shelf and a deep trench just offshore, which is geology. One question, four branches. That is normal.

Key idea: Oceanography is defined less by a single method than by its object, and every serious ocean question crosses the geological, chemical, physical, and biological branches at once.

Before Challenger: the ocean as a surface

For most of recorded history the ocean was studied only where it touched people: as a surface to cross, a shore to defend, and a place to fish. Polynesian navigators read swell patterns, cloud forms, and bird behavior with an accuracy Europeans did not match for centuries, and they did it without writing anything down. European practical knowledge accumulated as sailing directions and current charts. Benjamin Franklin, working with the Nantucket whaling captain Timothy Folger, published a chart of the Gulf Stream in the 1770s because mail packets were taking two weeks longer than merchant ships that knew to stay out of it.

The first person to make ocean data systematic was the United States naval officer Matthew Fontaine Maury. In the 1840s and 1850s he collected thousands of old ships' logs, extracted wind and current observations, and published standardized charts that cut sailing times dramatically. In 1855 he published The Physical Geography of the Sea, often called the first oceanography textbook. He also produced, in 1854, the first bathymetric chart of the North Atlantic, assembled from deep-sea soundings taken for a very practical reason: the telegraph industry wanted to lay a cable and needed to know what it would be lying on.

That is worth pausing on. Deep-sea science got its first sustained funding because a business needed the answer. This keeps happening. Sonar was developed for submarines, satellite altimetry was refined for navies, and much modern seafloor mapping is paid for by offshore energy and telecommunications. Ocean science has almost always ridden on somebody else's budget.

Key idea: Before the 1870s the ocean was charted as a surface for practical navigation and cable laying, and the deep interior was essentially unexplored.

The Challenger expedition, 1872 to 1876

The voyage that turned this into a science was the circumnavigation of HMS Challenger, a British warship stripped of most of her guns and rebuilt as a floating laboratory. Under the scientific direction of Charles Wyville Thomson, she sailed roughly 68,890 nautical miles, about 127,600 kilometers, over three and a half years, and stopped at 362 designated stations.

At each station the crew ran a full routine: sound the depth, sample the bottom, measure temperature at several depths with protected thermometers, collect water for chemical analysis, and drag nets and dredges for life. The routine mattered more than any single result. Because they did the same things in the same order everywhere, the observations could be compared, and comparison is what turns a collection of curiosities into a data set.

The findings reshaped what people believed. Challenger collected roughly 4,700 species new to science, which demolished the then-popular idea that the deep sea below a few hundred fathoms was lifeless. The bottom samples showed that different regions of the seafloor are blanketed in systematically different sediments. The temperature series showed a deep ocean that is uniformly cold everywhere, even beneath the tropics, which is a direct clue that deep water must come from the poles. Careful chemical analysis of the water samples, done later by William Dittmar, established that the major dissolved salts occur in essentially fixed proportions to one another everywhere in the open ocean, a result you will use in Lesson 5. The expedition's report ran to 50 volumes and roughly 29,500 pages, published between 1880 and 1895.

Key idea: Challenger invented the standardized ocean station, and its results established that the deep sea is inhabited, cold everywhere, systematically sedimented, and chemically consistent.

The instrument revolutions

Everything after Challenger is a story of instruments. Four jumps matter most, and each one changed what questions could even be asked.

Echo sounding, from the 1920s. Instead of lowering a line for hours, you emit a sound pulse and time its return. The German research vessel Meteor, working the South Atlantic from 1925 to 1927, ran continuous echo-sounding profiles and revealed that the Mid-Atlantic Ridge was not an isolated bump but a continuous mountain range. The arithmetic is easy and worth doing once. Sound travels through seawater at roughly 1,500 meters per second. If a pulse returns 5.0 seconds after you send it, the round trip is 5.0 times 1,500, which is 7,500 meters, so the depth is half of that, 3,750 meters. Compare that to a whole day with a hemp line.

Mapping and interpretation, from the 1950s. At Columbia University, Marie Tharp spent years converting Bruce Heezen's raw echo-sounding profiles into physiographic maps of the seafloor. She noticed a continuous V-shaped rift valley running down the crest of the Mid-Atlantic Ridge and connected it to the belt of earthquake epicenters. Heezen initially dismissed the idea as continental drift talk. The map she and Heezen published in 1957, and the world ocean floor panorama that followed, became one of the most persuasive pieces of evidence for seafloor spreading.

Going down, from the 1960s. The deep submergence vehicle Alvin, commissioned in 1964, let scientists observe the deep seafloor with their own eyes. In 1977 an Alvin dive on the Galapagos Rift found hot vents surrounded by dense animal communities living without sunlight, which rewrote biology. In parallel, the Deep Sea Drilling Project began coring the seafloor in 1968 with the ship Glomar Challenger, deliberately named after the 1870s expedition, and its successors continue as the international scientific ocean drilling programs.

Going up and going autonomous, from the 1990s. Satellite altimeters measure the height of the sea surface to a few centimeters from orbit. Because the ocean's surface bulges over dense currents and slopes across pressure gradients, altimetry gives you global maps of currents, eddies, and sea level. TOPEX/Poseidon launched in 1992 and started a continuous record maintained by the Jason series, Sentinel-6 Michael Freilich, and now SWOT. Meanwhile the Argo program, beginning around 1999, deployed a global array of drifting floats. Each float sinks to a parking depth near 1,000 meters, drifts about ten days, descends to roughly 2,000 meters, then rises while recording temperature and salinity, surfaces, transmits by satellite, and repeats. With roughly 3,900 floats active at any time, Argo delivers over 100,000 profiles a year and gave physical oceanography its first genuinely global, year-round view of the upper ocean.

Key idea: Each instrument revolution, echo sounder, deep submersible, satellite altimeter, and autonomous float, changed the ocean from a place you visit into a system you can monitor continuously.

What we still do not know

It is tempting to think the mapping is finished. It is not. As of the mid-2020s, only about a quarter of the global seafloor had been directly measured at modern resolution by ships or autonomous vehicles. The familiar colorful maps of the whole ocean floor are mostly not direct measurements at all: they are inferred from satellite altimetry, because a large seamount has enough gravitational pull to raise the sea surface above it by a few meters, and that bump can be read from orbit. Gravity-derived bathymetry has a horizontal resolution of a few kilometers, good enough to find seamounts, far too coarse to plan a cable route or resolve a canyon. The international Seabed 2030 project aims to close the gap.

Put the numbers side by side and the imbalance is stark. We have complete high-resolution topographic maps of the Moon, Mars, and Venus, and only partial direct coverage of our own seafloor. The reason is physics, not neglect: radar and laser pulses that map a planet from orbit cannot penetrate seawater, so the only way to measure the bottom precisely is to put an instrument in the water and drive it back and forth, one swath at a time, over 361 million square kilometers.

Key idea: Most published seafloor maps are inferred from satellite gravity rather than directly sounded, so the ocean floor remains the least accurately mapped solid surface in the inner solar system.

A worked comparison: how much has observing changed?

Suppose you want one temperature profile of the upper 2,000 meters of ocean. In 1875 that meant stopping the ship, lowering reversing thermometers on a wire, waiting for them to equilibrate, hauling them back, and reading them: call it several hours for a handful of depths, and you get maybe two or three profiles a day if the weather cooperates. A four-year expedition managed 362 stations, so about 90 stations per year.

An Argo float produces one full profile every ten days without a human present. That is 36.5 profiles per year per float. With 3,900 floats, the array yields roughly 142,000 profiles per year. Divide: 142,000 divided by 90 is about 1,580. The global float array collects in one year what the Challenger expedition, at its pace, would have needed roughly sixteen centuries to gather. And the floats do it in winter, in storms, in the Southern Ocean, where ships almost never go.

That single ratio explains why the modern textbook can state global averages with confidence that would have been reckless in 1950. It should also make you cautious about long-term trends before about 1970, when coverage was thin and biased toward shipping lanes in the Northern Hemisphere summer.

Common misconceptions

• "The seafloor has all been mapped." No. Only about a quarter is directly measured at modern resolution. The rest is a gravity-based estimate from satellites.
• "Oceanography is marine biology." Marine biology is one branch. Most working oceanographers spend their days on physics, chemistry, geology, or data analysis.
• "Challenger discovered the Mariana Trench's true depth." They found the deepest spot then known and were about 2.7 kilometers short of the modern value. Great measurement, not a final one.
• "Satellites can see the bottom." Not directly. Light and radar do not penetrate deep water. Satellites infer depth from the tiny sea-surface bumps that seafloor mass creates.
• "The deep sea is lifeless." Challenger disproved this in the 1870s, and the 1977 vent discovery showed entire ecosystems that never use sunlight at all.

Recap

• Oceanography has four traditional branches, geological, chemical, physical, and biological, and real questions cross all four.
• Maury systematized wind, current, and depth data in the mid-1800s, largely for navigation and telegraph cables.
• HMS Challenger, 1872 to 1876, invented the standardized ocean station across 362 sites and 50 volumes of results.
• Challenger established a living, uniformly cold deep sea and constant proportions among the major salts.
• Echo sounding, deep submersibles, satellite altimetry, and Argo floats each opened a new class of question.
• Argo alone yields roughly 142,000 profiles per year, more than a thousand times the Challenger rate.
• About a quarter of the seafloor is directly mapped; the rest is inferred from satellite gravity data.

Sources

1. National Ocean Service. (n.d.). How much of the ocean have we explored? NOAA. oceanservice.noaa.gov
2. Encyclopaedia Britannica. (n.d.). Oceanography. britannica.com
3. Encyclopaedia Britannica. (n.d.). Challenger Expedition. britannica.com
4. Argo Program. (n.d.). How Argo floats work. Scripps Institution of Oceanography, UC San Diego. argo.ucsd.edu
5. Nippon Foundation and GEBCO. (n.d.). Seabed 2030 project. seabed2030.org
6. Woods Hole Oceanographic Institution. (n.d.). Ocean learning hub. whoi.edu

The big picture

Imagine draining the Atlantic. You walk east from the New Jersey shore across a gently sloping plain, so gently sloping that after 100 kilometers you have descended only about 100 meters. Then the ground tilts under you, and over the next 50 kilometers you drop three and a half kilometers. You come out onto an apron of debris, then onto the flattest surface on the planet, a plain so level that it changes elevation by less than a meter per kilometer for a thousand kilometers. Halfway across you begin climbing a mountain range 2,500 meters high that runs from the Arctic to the Southern Ocean, and at its crest you find a rift valley splitting it down the middle.

That walk is the anatomy of an ocean basin, and every feature on it was made by a process you can name. This lesson gives you the map. The next lesson gives you the engine.

Start with scale, because the ocean is bigger than the numbers usually convey. The ocean covers about 361 million square kilometers, roughly 71 percent of Earth's surface, and holds about 1.335 billion cubic kilometers of water, roughly 97 percent of all the water on the planet. Its average depth is about 3,682 meters. Compare that with the average elevation of land, about 840 meters. The ocean is more than four times as deep, on average, as the continents are tall.

The hypsographic curve: Earth has two surfaces, not one

Plot the fraction of Earth's solid surface that lies at each elevation and you do not get a smooth bell curve. You get two humps. One cluster sits near sea level, from about 1,000 meters above it down to a few hundred meters below. The other sits between about 4,000 and 6,000 meters below sea level. Between them, in the range from roughly 200 to 4,000 meters deep, the curve drops steeply, meaning relatively little of Earth's surface has those elevations. This plot is the hypsographic curve, and its two-humped shape is one of the most important facts in Earth science.

The reason is that Earth has two kinds of crust with two different densities, and both float on the mantle like blocks of wood of different thickness. Continental crust is granitic, thick, about 35 kilometers on average, and relatively light. Oceanic crust is basaltic, thin, about 7 kilometers, and denser. Each floats to its own level, a principle called isostasy. The steep stretch between the humps is the continental slope, the narrow transition where one kind of crust gives way to the other. The ocean is not simply a low spot that filled with water. It is a fundamentally different kind of surface.

Key idea: Earth's surface has two dominant elevations because it has two kinds of crust with different densities and thicknesses, and each floats isostatically to its own level.

The continental margin

The continental margin is the submerged edge of a continent, and it has three parts.

The continental shelf is the shallow, nearly flat platform extending from the shoreline out to the shelf break. Its average width is about 65 to 80 kilometers, though it ranges from almost nothing off western South America to over 1,000 kilometers in the Arctic. The shelf break, where the gradient sharply steepens, sits at an average depth near 135 meters. The shelf's average gradient is about 0.1 degrees, which is a slope of roughly 1 in 1,000: descend one meter for every kilometer you travel. Geologically the shelf is continental crust, not ocean floor, which is why nations claim it. During the last glacial maximum, about 20,000 years ago, sea level was roughly 120 to 130 meters lower, so most of the world's shelves were dry land. People walked and hunted there, and their tools and mammoth bones come up in trawl nets today.

The continental slope runs from the shelf break down toward the deep ocean, with a much steeper average gradient of about 3 to 6 degrees. It is the true edge of the continent. Cutting across it are submarine canyons, some rivaling the Grand Canyon in scale, carved and maintained by turbidity currents: dense, sediment-laden avalanches of muddy water that race downslope, sometimes at tens of kilometers per hour. We know their speed because in 1929 an earthquake on the Grand Banks triggered one that snapped transatlantic telegraph cables in sequence, and the times at which each cable failed gave engineers a set of stopwatch readings on a flow nobody had seen.

The continental rise is the gently sloping apron of accumulated sediment at the base of the slope, built from the deposits those turbidity currents dump when they lose speed on the flat.

Key idea: A margin has a shallow shelf at roughly 0.1 degrees, a steep slope at 3 to 6 degrees marking the true continental edge, and a sedimentary rise built by turbidity currents.

Passive versus active margins

Not all margins have all three parts, and the difference tells you where you are in the plate tectonic story.

Feature	Passive margin	Active margin
Tectonic setting	Within a plate, far from a boundary	At a plate boundary, usually a subduction zone
Example	Eastern North America, west Africa	Western South America, Japan
Shelf	Wide, often 100 km or more	Narrow, often under 20 km
Rise present?	Yes, thick sediment apron	Usually absent; a trench takes its place
Earthquakes and volcanoes	Rare	Common
Sediment thickness	Very thick, kilometers	Thin, or scraped into an accretionary wedge

This single table explains a stubborn asymmetry. The Atlantic is rimmed by passive margins, so it has broad shelves, thick sediment, and only a handful of trenches. The Pacific is rimmed by active margins, the Ring of Fire, so it has narrow shelves, deep trenches, and most of the world's subduction earthquakes. When someone tells you a coastline has a 200-kilometer shelf, you already know it is probably not sitting over a subduction zone.

Key idea: Passive margins are wide, thickly sedimented, and quiet; active margins are narrow, trench-bounded, and seismically busy.

The deep ocean floor

Beyond the rise lie the abyssal plains, at depths of roughly 4,000 to 6,000 meters. These are the flattest surfaces on Earth, with gradients often less than 1 meter of relief per kilometer. They are flat because fine sediment, raining down over tens of millions of years, has buried an originally rough basaltic surface the way snow smooths a rocky field. Where the sediment blanket is thinner, the buried topography pokes through as abyssal hills, low bumps a few hundred meters high that are, by area, the most common landform on the planet.

Rising from the plains are seamounts, submarine volcanoes taller than 1,000 meters. Estimates run well over 100,000 of them, and only a small fraction have ever been surveyed in detail. Some have flat tops. These are guyots, seamounts that once reached the surface, were planed flat by waves, and then subsided as the aging plate beneath them cooled and sank. In the tropics, subsiding volcanic islands with fringing coral reefs can leave behind a ring of reef around a lagoon: an atoll. Charles Darwin proposed this subsidence explanation in 1842 from surface observation alone. It was confirmed in 1952, when drilling on Enewetak Atoll passed through more than a kilometer of reef carbonate and struck volcanic basalt underneath, exactly as Darwin's model required.

Key idea: Abyssal plains are sediment-smoothed basalt, abyssal hills are the buried roughness showing through, and seamounts, guyots, and atolls record volcanoes riding a cooling, subsiding plate.

Ridges and trenches: the extremes

The mid-ocean ridge system is the largest single geological feature on Earth: a continuous volcanic mountain chain roughly 65,000 kilometers long, winding through every ocean basin. Its crest typically lies about 2,500 meters below sea level, standing 2 to 3 kilometers above the neighboring abyssal plains. Slow-spreading ridges such as the Mid-Atlantic Ridge are rugged and carry a deep central rift valley, sometimes 1 to 3 kilometers deep. Fast-spreading ridges such as the East Pacific Rise are smoother and broader, with only a shallow axial trough. Lesson 3 explains why.

Ocean trenches are the deepest places on Earth, narrow arcs where one plate bends and dives beneath another. Most exceed 7,000 meters. The Mariana Trench holds the Challenger Deep at about 10,935 meters. Drop Mount Everest, 8,849 meters, into it and its summit would still lie more than 2 kilometers underwater. Depths below 6,000 meters define the hadal zone, which sounds vast but occupies only a fraction of a percent of the ocean floor.

Key idea: The mid-ocean ridge is a 65,000-kilometer volcanic chain built by plates separating, and trenches are narrow, hadal-depth scars where plates converge.

A warning about every seafloor profile you will ever see

Almost every cross-section of an ocean basin is drawn with enormous vertical exaggeration, and it will mislead you unless you correct for it. Work the numbers. The Atlantic is roughly 5,000 kilometers wide at mid-latitudes and about 4 kilometers deep. The true aspect ratio is 5,000 to 4, or 1,250 to 1. Drawn honestly on this page, the entire Atlantic basin would be a hairline scratch. So textbooks stretch the vertical scale, often by a factor of 20 to 100.

Here is what that does. The continental slope has a real gradient near 4 degrees, which is gentler than a wheelchair ramp is steep. At 20 times vertical exaggeration, the tangent of the drawn angle is 20 times the true tangent: tan(4 degrees) is about 0.070, times 20 gives 1.40, and the arctangent of 1.40 is about 54 degrees. The gentlest of ramps is drawn as a cliff. Whenever you read a profile, find the vertical exaggeration in the caption, and mentally flatten everything.

Key idea: Seafloor profiles use vertical exaggeration of 20 to 100 times, so the real ocean floor is vastly gentler than any diagram of it suggests.

Common misconceptions

• "The seafloor is a rugged, canyon-riddled landscape everywhere." Most of it is the flattest terrain on Earth. The drama is concentrated at ridges, trenches, and margins.
• "The continental shelf is part of the ocean floor." Geologically it is continental crust that happens to be flooded, which is exactly why coastal states claim rights to it.
• "The Mid-Atlantic Ridge is a chain of separate underwater volcanoes." It is one continuous system, the longest mountain range on the planet, threading through every ocean.
• "Trenches make up a large share of the deep sea." Hadal depths below 6,000 meters cover well under one percent of the ocean floor.
• "Textbook profiles show the real steepness." They do not. A 4-degree slope drawn at 20 times exaggeration appears to be over 50 degrees.

Recap

• The ocean covers about 71 percent of Earth, averages 3,682 meters deep, and holds 97 percent of the planet's water.
• The hypsographic curve is two-humped because continental and oceanic crust differ in density and thickness and float isostatically.
• A margin runs shelf (about 0.1 degrees, break near 135 meters), slope (3 to 6 degrees), then rise.
• Submarine canyons are maintained by turbidity currents, timed for the first time by the 1929 Grand Banks cable breaks.
• Passive margins are wide and quiet; active margins are narrow, trench-bounded, and seismic.
• Abyssal plains are sediment-smoothed; seamounts, guyots, and atolls track a subsiding, cooling plate.
• The mid-ocean ridge runs 65,000 kilometers; trenches reach 10,935 meters at the Challenger Deep.

Sources

1. National Ocean Service. (n.d.). How deep is the ocean? NOAA. oceanservice.noaa.gov
2. National Ocean Service. (n.d.). What is a submarine canyon? NOAA. oceanservice.noaa.gov
3. Encyclopaedia Britannica. (n.d.). Continental shelf. britannica.com
4. Encyclopaedia Britannica. (n.d.). Oceanic trench. britannica.com
5. U.S. Geological Survey. (n.d.). Coastal and marine hazards and resources program. usgs.gov
6. Wikipedia contributors. (n.d.). Hypsometric curve. Wikipedia. en.wikipedia.org

The big picture

In the early 1960s, ships towing magnetometers across the northeast Pacific brought back a pattern nobody expected: long parallel stripes of alternately stronger and weaker magnetism in the seafloor rock, running for hundreds of kilometers and mirrored symmetrically on either side of a ridge. Drawn in black and white, the map looked like a zebra skin. In 1963 Fred Vine and Drummond Matthews in Cambridge, and independently Lawrence Morley in Canada, saw what it meant. Earth's magnetic field reverses polarity from time to time. If new crust is continuously being made at a ridge and carried outward in both directions, then each strip of lava freezes in the field direction of its own moment of birth, and the ridge writes the same magnetic barcode onto both of its outbound conveyor belts at once.

That is a spectacular piece of reasoning, and it settled a fifty-year argument. The seafloor is not an ancient permanent floor. It is a young, moving surface, manufactured at one set of boundaries and destroyed at another. This lesson explains the machine that builds and unbuilds ocean basins, because almost everything in the previous lesson, ridges, trenches, margin type, even shelf width, is downstream of it.

Alfred Wegener had proposed continental drift in 1912 on the basis of matching coastlines, matching fossils across the Atlantic, and matching rock sequences. He was largely right about the pattern and had no plausible mechanism, so geologists rejected him. The ocean floor supplied the mechanism.

Seafloor spreading

In 1962 Harry Hess of Princeton published a paper he modestly called "an essay in geopoetry." His proposal, now called seafloor spreading, was that mantle material rises beneath mid-ocean ridges, cools into new basaltic oceanic crust, and moves outward on both sides as still more rises behind it. Where that crust eventually meets a continent or another plate, it bends down into a trench and returns to the mantle. The ocean floor is a conveyor belt with a source and a sink.

This reframes plate tectonics correctly. The lithosphere, the rigid outer shell of crust plus uppermost mantle, is broken into about a dozen major plates that move over the weaker, ductile asthenosphere below. Plates meet at three kinds of boundary: divergent boundaries where they separate and new crust is made, convergent boundaries where one dives beneath the other, and transform boundaries where they slide past each other.

Key idea: Ocean floor is manufactured at divergent ridges and consumed at convergent trenches, so the seafloor is a conveyor belt rather than a permanent basin bottom.

Four independent lines of evidence

What makes spreading convincing is not one measurement but several that were made for different reasons and agree.

• Magnetic stripes. The symmetric zebra pattern about the ridge axis, matched to the independently dated timescale of geomagnetic reversals, gives an age for every stripe.
• Drilled age. Beginning in 1968, the Glomar Challenger cored basement rock at increasing distances from ridges. The age of the oldest sediment sitting directly on basalt increased steadily with distance, exactly as predicted, and matched the magnetic ages.
• Sediment thickness. Sediment accumulates at a roughly steady rate, so older crust has had longer to collect it. Thickness is essentially zero at a ridge crest and grows outward. If the seafloor were ancient everywhere, sediment would be kilometers thick everywhere.
• Heat flow and depth. Heat escaping from the seafloor is highest at ridge crests and falls off with distance. Depth increases with distance too, because cooling rock contracts and gets denser, so the plate rides lower.

Add the earthquake maps. Shallow earthquakes trace ridges and transforms; deep earthquakes, down to about 700 kilometers, occur only in inclined slabs beneath trenches. A slab of cold, brittle rock descending into the mantle is the only sensible reason for earthquakes at that depth.

Key idea: Magnetic stripes, drilled basement ages, sediment thickness, heat flow, depth, and earthquake depth distribution were measured independently and all converge on seafloor spreading.

Working the numbers: spreading rates and ages

Spreading rate is where the theory becomes arithmetic. Rates are quoted either as half rates, the speed of one flank away from the axis, or full rates, the rate at which the two plates separate. Always check which one you have.

Example 1. A magnetic stripe dated at 4.0 million years old lies 50 kilometers from the ridge axis. The half rate is 50 kilometers divided by 4.0 million years. Convert: 50 km is 5,000,000 cm, so 5,000,000 divided by 4,000,000 is 1.25 centimeters per year. The full spreading rate is twice that, 2.5 centimeters per year, which is close to the modern Mid-Atlantic Ridge.

Example 2. How long did it take to open the Atlantic? Take a width of about 5,000 kilometers at mid-latitudes and a full rate of 2.5 centimeters per year. In centimeters, 5,000 km is 5 times 10 to the 8th. Divide by 2.5 and you get 2 times 10 to the 8th years, or 200 million years. The geological record puts the opening of the central Atlantic at roughly 180 million years ago. A one-line calculation lands within about ten percent of decades of careful stratigraphy.

Example 3. The East Pacific Rise spreads at a full rate near 15 centimeters per year, six times faster than the Atlantic. Fast-spreading ridges are broad and smooth with only a shallow axial trough, because the crust stays hot and pliable and magma supply is generous. Slow-spreading ridges are rugged with a deep central rift valley, because the lithosphere cools and becomes brittle close to the axis and stretches by faulting. The morphology of a ridge tells you its speed.

Key idea: Spreading rate equals distance divided by age, and it explains ridge shape: fast ridges are smooth and broad, slow ridges are rugged with deep rift valleys.

Why the seafloor deepens as it ages

Newly made lithosphere is hot and therefore relatively low in density, so it floats high. As it moves away from the axis it loses heat to the ocean, contracts, and becomes denser, so it sinks isostatically. To a good approximation, for crust younger than about 80 million years, the depth follows a square-root-of-age law:

depth in meters is about 2,500 plus 350 times the square root of the age in millions of years.

Check it. At age zero the formula gives 2,500 meters, the typical depth of a ridge crest. At 25 million years the square root is 5, so the depth is 2,500 plus 1,750, which is 4,250 meters. At 100 million years the square root is 10, so the depth is 2,500 plus 3,500, which is 6,000 meters, the deep abyssal plain. The abyssal plains of the previous lesson are not a separate kind of place. They are simply old ridge crest, cooled and sunk.

This also explains guyots. A volcano built at a young, high ridge flank reaches the surface. Waves plane its top flat. Then, as the plate ages and subsides according to that same square-root law, the flat summit is carried down into permanent deep water.

Key idea: Seafloor depth grows as the square root of crustal age because cooling lithosphere contracts and sinks, which is why old basins are deep and ridge crests are shallow.

Convergence: where ocean floor dies

Because Earth is not expanding, everything made at ridges must be destroyed somewhere. At a subduction zone the older, colder, denser oceanic plate bends and descends beneath its neighbor at an angle typically between 30 and 60 degrees. The bend itself creates the trench. Descending sediment and hydrated minerals release water into the overlying mantle wedge, which lowers the melting point of that rock and generates magma, which rises to build a chain of volcanoes: a volcanic island arc if the overriding plate is oceanic, as in Japan or the Aleutians, or a continental arc such as the Andes if it is continental. Scraped-off sediment piles into an accretionary wedge on the landward wall of the trench.

One consequence deserves emphasis. Because ocean crust is always being recycled, it never gets very old. The oldest oceanic crust anywhere is roughly 180 to 200 million years, in the western Pacific and off northwest Africa. Continental crust, which is too buoyant to subduct, includes rocks over 4 billion years old. The ocean floor is younger than most dinosaurs; the continents remember almost the whole history of the planet.

Transform boundaries complete the set. Ridges are offset in steps by transform faults, and the inactive extensions of those offsets continue across the basin as fracture zones, long scars visible on any bathymetric map.

Key idea: Subduction destroys ocean floor and builds trenches, arcs, and accretionary wedges, which is why no oceanic crust is older than about 200 million years.

Hotspots, and a second way to measure plate motion

Not all volcanism sits at plate boundaries. A hotspot is a persistent source of magma, apparently rooted deep in the mantle and roughly fixed while a plate slides over it. The Hawaiian chain is the classic case: an active volcano at the southeast end and a line of progressively older, more eroded, more deeply subsided islands and seamounts stretching northwest.

You can measure plate speed from it. Kauai's shield volcanism ended around 5.1 million years ago, and Kauai lies roughly 520 kilometers northwest of the active volcanism on the Big Island. Divide: 520 kilometers is 52,000,000 centimeters; divided by 5,100,000 years that is about 10.2 centimeters per year. Independent GPS measurements of Pacific plate motion give about 9 to 10 centimeters per year. Two entirely different methods, one geological and one geodetic, agree.

Follow the chain far enough northwest and it bends sharply north into the Emperor Seamounts, at a location dated near 47 million years. Something changed then, either the plate's direction or the hotspot's position, and the argument over which is still live.

Key idea: Hotspot island chains record plate motion directly, and the Hawaiian chain gives about 10 centimeters per year, matching GPS measurements of the Pacific plate.

The Wilson cycle: an ocean basin has a life story

J. Tuzo Wilson recognized that ocean basins open and close in a repeating cycle. Every stage exists on Earth right now, so you can see the whole life history at once.

Stage	What is happening	Modern example
Embryonic	A continent begins to rift and stretch	East African Rift
Young	A narrow sea floods the rift; new ocean crust begins	Red Sea, Gulf of California
Mature	A wide basin with an active ridge and passive margins	Atlantic Ocean
Declining	Subduction zones ring the basin and it begins closing	Pacific Ocean
Terminal	Continents converge; only a remnant sea remains	Mediterranean Sea
Suturing	Basin gone; a mountain belt marks the seam	Himalaya, Ural Mountains

The Atlantic is widening by a few centimeters a year while the Pacific, ringed by trenches, is shrinking. Marine fossils high in the Himalaya are the closing argument: that rock was once the floor of the Tethys Ocean, which no longer exists.

Key idea: Ocean basins open and close on a Wilson cycle of a few hundred million years, and every stage of that cycle is visible somewhere on Earth today.

Common misconceptions

• "Continents plow through the ocean floor." No. Continents ride on plates that include ocean floor. The whole plate moves as a unit.
• "The seafloor is the oldest rock on Earth." The opposite. It is among the youngest, nowhere older than about 200 million years, because subduction recycles it.
• "Magnetic stripes are stripes of magnetic rock and non-magnetic rock." They are all basalt. The stripes record alternating polarity of the field frozen in at the moment each strip cooled.
• "Earth must be expanding if new crust is made at ridges." No. Subduction destroys ocean floor at about the rate ridges create it, so total area is roughly conserved.
• "Ridges are deep because they are in the middle of the ocean." Ridge crests are the shallowest deep-ocean terrain, near 2,500 meters, because the young crust beneath them is still hot and buoyant.

Recap

• Seafloor spreading makes crust at ridges and destroys it at trenches; the floor is a conveyor, not a permanent basin.
• Magnetic stripes, drilled ages, sediment thickness, heat flow, and earthquake depths independently confirm it.
• Spreading rate equals distance over age; the Atlantic at 2.5 cm/yr and 5,000 km wide implies about 200 million years.
• Fast ridges are smooth and broad; slow ridges are rugged with a deep axial rift valley.
• Depth is about 2,500 plus 350 times the square root of age in millions of years, so old crust lies deep.
• Subduction builds trenches, arcs, and accretionary wedges and keeps ocean crust younger than about 200 million years.
• Hotspot chains such as Hawaii measure plate motion at about 10 cm/yr, matching GPS.
• The Wilson cycle runs from rift to young sea to mature basin to closing basin to suture, all visible today.

Sources

1. U.S. Geological Survey. (n.d.). Understanding plate motions. This Dynamic Earth. pubs.usgs.gov
2. Encyclopaedia Britannica. (n.d.). Plate tectonics. britannica.com
3. Encyclopaedia Britannica. (n.d.). Seafloor spreading. britannica.com
4. National Ocean Service. (n.d.). What is the mid-ocean ridge? NOAA. oceanservice.noaa.gov
5. U.S. Geological Survey. (n.d.). Hawaiian volcanoes and the hotspot. Hawaiian Volcano Observatory. usgs.gov
6. Wikipedia contributors. (n.d.). Wilson cycle. Wikipedia. en.wikipedia.org

The big picture

A drilling ship holds position over 4,000 meters of water and pulls up a plastic tube of grey-white mud. Split it lengthwise and it looks like nothing: soft, faintly banded, unremarkable. Put a pinch of it under a microscope and it turns out to be almost entirely shells, hundreds of thousands of them per gram, each one the calcite skeleton of a single-celled organism that lived in the sunlit surface water and sank when it died. Measure the oxygen isotopes in those shells down the length of the core and the banding resolves into ice ages: a saw-toothed record of glacial and interglacial cycles marching back a million years, in order, undisturbed.

That is why marine sediments matter out of all proportion to how boring they look. The deep seafloor is the quietest depositional environment on the planet. Material rains down slowly, nothing burrows very deeply, and nothing washes it away. The result is Earth's most continuous archive of its own past, and essentially everything we know about ice ages, past ocean chemistry, and abrupt climate change before the ice-core record comes out of it.

Sediments also matter for three practical reasons. They are habitat, since most seafloor animals live in or on them. They are resource, since offshore oil, gas, sand, and proposed deep-sea mining targets all sit in or under sediment. And they are hazard, since unstable sediment on a slope can fail catastrophically and generate a tsunami.

Four origins

Oceanographers classify sediment first by where the particles came from.

Lithogenous (also called terrigenous) sediment is broken rock: quartz, clay minerals, feldspar. Rivers deliver roughly 20 billion tonnes of it to the ocean every year, most of it dropped near the coast. Wind carries fine dust enormous distances; NASA satellite estimates put roughly 180 million tonnes of dust leaving North Africa each year, with tens of millions of tonnes settling across the Atlantic. Glaciers carry coarse debris out to sea in icebergs and drop it when the ice melts, a process called ice rafting that leaves distinctive layers of sand and pebbles in otherwise fine polar mud. In the deep open ocean, far from all of these, the lithogenous fraction becomes an extremely fine abyssal clay, often reddish-brown from oxidized iron.

Biogenous sediment is the hard parts of organisms. Two minerals dominate. Calcium carbonate comes from foraminifera, which are single-celled animals, and coccolithophores, which are single-celled algae that build plates of calcite. Silica, specifically opal, comes from diatoms, which are algae, and radiolarians, which are single-celled animals. When biogenous material makes up more than 30 percent of a deep-sea sediment by weight, the sediment is called an ooze: calcareous ooze or siliceous ooze depending on the mineral.

Hydrogenous sediment precipitates directly out of seawater. The famous example is the manganese nodule, a lumpy black concretion of manganese and iron oxides that grows in concentric layers around a nucleus, at a rate of roughly 1 to 10 millimeters per million years. That is the slowest geological process anyone routinely measures. Evaporites such as gypsum and halite, and phosphorite deposits under upwelling zones, are also hydrogenous.

Cosmogenous sediment is extraterrestrial: micrometeorites and tektites. It is real, and it is a rounding error by mass everywhere except in unusually starved settings.

Key idea: Marine sediment is lithogenous (broken rock), biogenous (shells and skeletons), hydrogenous (precipitated from seawater), or cosmogenous (from space), and the first two dominate almost everywhere.

Where each type ends up

The global distribution is not random, and one rule explains most of it: whichever source is supplying material fastest wins.

Near continents, rivers deliver so much rock debris that lithogenous sediment swamps everything else. These neritic deposits, on shelves and slopes, are coarse, fast-accumulating, and often thick. Move out into the deep basins and river supply collapses. Now the slow rain of plankton shells competes with an even slower rain of wind-blown clay, so pelagic deposits are dominated by oozes wherever surface productivity is decent and by abyssal clay where it is not.

Within the oozes, the pattern follows the plankton. Calcareous ooze covers the largest single share of the deep seafloor, roughly half, concentrated in the warm, shallower parts of the Atlantic and Indian Oceans and along ridge flanks. Siliceous ooze forms two great belts: a diatom belt circling the Southern Ocean, where cold nutrient-rich water supports enormous diatom blooms, and an equatorial band of radiolarian ooze under the divergence zone. Abyssal clay dominates the deep central North Pacific, which is far from rivers, low in productivity, and, crucially, too deep for carbonate to survive.

Key idea: Coarse lithogenous sediment dominates near continents, oozes dominate the deep sea beneath productive water, and abyssal clay takes over where productivity is low and the water is too deep for carbonate.

The carbonate compensation depth

That last clause hides one of the most important ideas in marine geology. Calcium carbonate is more soluble in cold water and under high pressure, and it dissolves faster where dissolved carbon dioxide is high, because CO2 makes seawater more corrosive to carbonate. All three conditions intensify with depth.

So a foraminifer shell sinking through the water column enters increasingly hostile chemistry. At some depth, called the lysocline, shells begin to show obvious dissolution: pitted, thinned, fragmented. Deeper still is the carbonate compensation depth, or CCD, defined as the depth at which carbonate dissolves exactly as fast as it arrives. Below the CCD, no calcareous sediment accumulates at all. The seafloor there is clay, not because nothing calcareous fell on it, but because everything calcareous that fell on it dissolved.

The CCD sits near 4,500 meters in the Atlantic and around 3,500 meters in the Pacific, and that difference is a beautiful piece of reasoning. Deep water forms in the North Atlantic and travels for a thousand years or more before reaching the deep North Pacific. All along the way, sinking organic matter is respired by bacteria, releasing CO2 into the water. Deep Pacific water is therefore the oldest, most CO2-charged, most corrosive deep water on Earth, so its CCD is shallower. The sediment map of the Pacific is, in part, a map of how long its deep water has been away from the surface.

Key idea: Below the carbonate compensation depth, calcite dissolves as fast as it arrives, so carbonate sediment vanishes; the CCD is shallower in the Pacific because its deep water is older and richer in respired CO2.

Accumulation rates: turning thickness into time

Sediment is a clock if you know its rate. Typical values are worth memorizing to an order of magnitude.

Setting	Typical rate	Time to build 1 meter
Continental margin, near a river	10 to 100 cm per 1,000 years	1,000 to 10,000 years
Calcareous ooze, open ocean	1 to 6 cm per 1,000 years	about 17,000 to 100,000 years
Abyssal clay, deep Pacific	about 0.1 cm per 1,000 years	about 1,000,000 years
Manganese nodule growth	1 to 10 mm per million years	100 million to 1 billion years

Work two examples. First: you core 3 meters of calcareous ooze accumulating at 2 centimeters per 1,000 years. Convert 3 meters to 300 centimeters, divide by 2 centimeters per 1,000 years, and you get 150 units of 1,000 years, which is 150,000 years. That single short core spans the whole last glacial cycle.

Second: you core 3 meters of abyssal clay at 0.1 centimeters per 1,000 years. Now 300 divided by 0.1 gives 3,000 units of 1,000 years, which is 3 million years. Same length of tube, twenty times the span, and vastly coarser time resolution. This is exactly the trade-off paleoceanographers manage: fast-accumulating sites give fine detail over short intervals, slow sites give long records with each centimeter smearing together thousands of years.

One more useful consequence. Average deep-sea sediment thickness is only a few hundred meters, and it is thinnest on young crust near ridges. A 200-meter blanket at 0.5 centimeters per 1,000 years takes 40 million years to build, which is entirely consistent with seafloor that is nowhere older than about 200 million years. Sediment thickness and plate tectonics agree.

Key idea: Time equals thickness divided by accumulation rate, and the enormous range of rates sets the trade-off between record length and time resolution.

Reading a core: what sediment actually tells you

Several independent signals live in the same mud.

• Oxygen isotopes. Seawater contains oxygen-16 and the heavier oxygen-18. Evaporation preferentially removes the lighter isotope, and during glacial periods that light water is locked away in continental ice sheets, leaving the ocean enriched in oxygen-18. Foraminifera building shells from that water record the enrichment. The ratio in fossil shells therefore tracks global ice volume, and it produced the first continuous ice-age record.
• Orbital cycles. Spectral analysis of those isotope records reveals periodicities near 100,000, 41,000, and 23,000 years, matching Earth's orbital eccentricity, axial tilt, and precession. A landmark 1976 study of deep-sea cores made the Milankovitch theory of ice ages testable and then supported it.
• Assemblage. Which species are present indicates surface temperature, because warm-water and cold-water plankton have known preferences.
• Event layers. Volcanic ash beds date and correlate cores across basins. A worldwide clay layer enriched in iridium marks the asteroid impact 66 million years ago that ended the Cretaceous, a discovery made in 1980 partly from marine sections.

Key idea: A single core carries ice volume in its isotopes, orbital cycles in its rhythms, sea-surface temperature in its species, and datable events in its ash and impact layers.

Common misconceptions

• "Sediment is just dirt washed off the land." Roughly half the deep seafloor is covered by shells of plankton, not rock debris.
• "Deep-sea sediment is kilometers thick everywhere." On average it is a few hundred meters, and essentially zero at ridge crests, because the crust is young.
• "Carbonate is absent from the deep Pacific because plankton do not live there." They do. Their shells dissolve on the way down or on the seafloor, below a CCD near 3,500 meters.
• "A thicker core always means a longer record." No. A 3-meter core spans 150,000 years of ooze or 3 million years of abyssal clay, depending on rate.
• "Manganese nodules can be farmed sustainably because they regrow." They grow at millimeters per million years, which is not a human timescale in any sense.

Recap

• Sediments are lithogenous, biogenous, hydrogenous, or cosmogenous, and the first two dominate by mass.
• Ooze means more than 30 percent biogenous material, either calcareous or siliceous.
• Coarse lithogenous material dominates margins; oozes and abyssal clay dominate the deep sea.
• Below the CCD, near 4,500 m in the Atlantic and 3,500 m in the Pacific, carbonate dissolves as fast as it falls.
• The Pacific CCD is shallower because its deep water is older and carries more respired CO2.
• Time equals thickness divided by rate: 3 m of ooze at 2 cm per 1,000 years is 150,000 years.
• Cores record ice volume through oxygen isotopes, orbital cycles, past temperatures, and datable ash and impact layers.

Sources

1. Encyclopaedia Britannica. (n.d.). Marine sediment. britannica.com
2. National Ocean Service. (n.d.). What is marine snow? NOAA. oceanservice.noaa.gov
3. NASA Earth Observatory. (2015). Saharan dust feeds Amazon's plants. earthobservatory.nasa.gov
4. National Centers for Environmental Information. (n.d.). Marine geology and geophysics: seafloor sediment data. NOAA. ncei.noaa.gov
5. Wikipedia contributors. (n.d.). Carbonate compensation depth. Wikipedia. en.wikipedia.org
6. Woods Hole Oceanographic Institution. (n.d.). Ocean learning hub: seafloor and sediments. whoi.edu

The big picture

Two questions have been asked about seawater for as long as people have tasted it. Why is the ocean salty? And why is it not getting saltier? The first is easy and the second is the interesting one, because rivers have been carrying dissolved rock into the sea for billions of years and the ocean is not a evaporating pan. Something must be taking the salt back out at almost exactly the rate it comes in. Working out what, and how fast, is the core of chemical oceanography.

Here is the frame for this lesson. Seawater is a dilute solution in which almost everything is present and almost nothing is abundant. About 96.5 percent of it is water. The dissolved 3.5 percent is dominated by six ions. On top of that sit dissolved gases, a carbon system that buffers the whole ocean's acidity, and nutrients present in vanishingly small amounts that nonetheless decide where life can grow. Each of those four topics has a governing idea, and you can carry all four away from this lesson.

Salinity and the six ions that matter

Salinity is the total mass of dissolved inorganic solids in a given mass of seawater. Average open-ocean salinity is about 35, meaning 35 grams of dissolved salts per kilogram of seawater, which is 3.5 percent by mass. Take a kilogram of seawater and you have roughly 965 grams of water and 35 grams of salts. Modern practice reports salinity as a dimensionless number on the practical salinity scale, derived from electrical conductivity, but you can safely read "35" as "35 grams per kilogram."

Six ions supply over 99 percent of that mass.

Ion	Grams per kg of seawater at S = 35	Percent of dissolved solids
Chloride	19.35	55.0
Sodium	10.78	30.6
Sulfate	2.71	7.7
Magnesium	1.28	3.7
Calcium	0.41	1.2
Potassium	0.40	1.1

Notice that ordinary table salt, sodium chloride, accounts for about 86 percent of the total. That is why seawater tastes mostly of salt, with a bitter edge from magnesium and sulfate.

Now the result that makes chemical oceanography tractable. The principle of constant proportions, established by Georg Forchhammer and confirmed rigorously by William Dittmar on the Challenger samples, states that although total salinity varies from place to place, the ratios of the major ions to one another are essentially the same throughout the open ocean. Water from the Arctic and water from the tropics differ in how much salt they carry, not in what kind.

This is enormously practical. If the ratios are fixed, you can measure any one property that scales with total dissolved ions and get all of them. Historically chemists measured chlorinity by titration and converted. Since the 1970s instruments measure electrical conductivity, which responds to total ion content, and every CTD instrument lowered from a ship or carried by an Argo float measures Conductivity, Temperature, and Depth for exactly this reason. Constant proportions is what makes salinity measurable by a robot.

Key idea: Salinity averages 35 grams per kilogram, six ions supply over 99 percent of it, and their ratios are constant in the open ocean, which is why conductivity alone can measure salinity.

Sources, sinks, and residence time

Salt enters the ocean from chemical weathering of rock carried by rivers, from volcanic outgassing, and from hydrothermal circulation through hot young crust at ridges. It leaves through several doors: sea spray blown onto land, burial in sediments and pore water, precipitation as evaporite minerals in restricted basins, uptake by organisms, adsorption onto sinking clay particles, and chemical exchange with basalt in hydrothermal systems. Over long times the doors match the taps, which is why the composition is steady.

The quantitative tool is residence time: the average time a dissolved atom spends in the ocean before being removed. For a system in rough balance,

residence time equals the amount in the ocean divided by the rate of supply or removal.

Work the cleanest case, water itself. The ocean holds about 1.335 times 10 to the 18th cubic meters. Evaporation removes roughly 4.25 times 10 to the 14th cubic meters per year. Divide: 1.335 times 10 to the 18th divided by 4.25 times 10 to the 14th is about 3,100 years. A water molecule spends roughly three millennia in the sea between evaporations.

Now compare elements. The numbers span nine orders of magnitude.

Constituent	Approximate residence time	Interpretation
Chloride	about 100 million years	Almost nothing removes it; it just accumulates
Sodium	about 75 million years	Conservative, mixed uniformly everywhere
Calcium	about 1 million years	Removed by shell-building organisms
Silicon	about 20,000 years	Taken up rapidly by diatoms
Water	about 3,100 years	Cycled by evaporation
Iron	about 100 to 200 years	Scavenged onto particles almost immediately

Read the table as a rule. If residence time is much longer than the roughly 1,000-year time it takes to mix the whole ocean, the element gets stirred uniform and its concentration is the same everywhere: it is conservative. If residence time is much shorter than the mixing time, the element is consumed before it can spread, so it shows sharp gradients: it is non-conservative. That one comparison predicts, without any further chemistry, that sodium will be boringly uniform and iron will be patchy and scarce, which is exactly what the data show and exactly why iron limits life across huge stretches of ocean.

Key idea: Residence time equals inventory divided by flux; constituents with residence times longer than the roughly 1,000-year ocean mixing time are uniform, and shorter ones are patchy and often limiting.

Why salinity varies, even with constant proportions

If the ratios are fixed, what changes the total? Almost entirely the balance between evaporation and precipitation at the surface, plus river input and ice.

Evaporation removes water and leaves salt behind, raising salinity. Precipitation and river runoff add water and dilute it. So the map of surface salinity is close to a map of evaporation minus precipitation. Salinity peaks near 36.5 to 37 in the subtropical gyres around 25 degrees latitude, where descending dry air makes deserts and the ocean surface evaporates hard. It dips near the equator, where the intertropical convergence zone dumps rain, and again at high latitudes where evaporation is weak and rivers and melting ice deliver fresh water. The Baltic Sea, nearly enclosed and fed by rivers, runs near 7. The Red Sea, hot, enclosed, and essentially riverless, exceeds 40.

Ice deserves a note because it acts in both directions. When seawater freezes, the ice crystal lattice largely excludes salt, so the ice is nearly fresh and the salt is expelled into the water below. This brine rejection makes very cold, very salty, very dense water that sinks, and it is the primary engine of the deep circulation you will meet in Lesson 8. When that ice melts, it freshens the surface again.

Key idea: Surface salinity tracks evaporation minus precipitation, peaking in the subtropics and dipping at the equator and poles, while freezing sea ice rejects brine and makes the densest water in the ocean.

Dissolved gases

Gases dissolve at the sea surface and are carried down by mixing and sinking water. Two rules govern how much dissolves. Solubility increases as temperature falls, so cold water holds more gas, and solubility decreases slightly as salinity rises. Warm tropical surface water might hold about 6 milligrams of oxygen per liter at saturation, while near-freezing polar water holds close to 11.

Compare the gas mixture in air with the mixture dissolved in surface seawater and something jumps out. Air is roughly 78 percent nitrogen, 21 percent oxygen, and 0.04 percent carbon dioxide. The gas dissolved in surface seawater is roughly 48 percent nitrogen, 36 percent oxygen, and 15 percent carbon dioxide by volume. Carbon dioxide is enriched by a factor of hundreds. The reason is that CO2 does not merely dissolve, it reacts. That reaction is the next section, and it is why the ocean holds roughly fifty times more carbon than the atmosphere does.

Oxygen has a distinctive depth profile worth knowing. It is high at the surface, where the atmosphere and photosynthesis supply it. It falls sharply with depth as sinking organic matter is respired by bacteria, reaching a broad oxygen minimum zone typically between about 200 and 1,000 meters. Below that it often rises again, because the deep water in that basin sank recently from a cold, oxygen-rich polar surface and has not yet had all its oxygen consumed. Reading an oxygen profile therefore tells you about both biology and circulation at once.

Key idea: Cold water holds more dissolved gas, and CO2 is enormously enriched in seawater relative to air because it reacts chemically rather than simply dissolving.

The carbonate system and why the ocean has a stable pH

When carbon dioxide enters seawater it goes through a chain of reactions. Written in plain text:

CO2 plus H2O gives H2CO3, carbonic acid; H2CO3 gives H+ plus HCO3-, bicarbonate; and HCO3- gives H+ plus CO3 2-, carbonate.

The sum of all these carbon-bearing species is called dissolved inorganic carbon, or DIC, and in typical surface seawater it totals about 2,000 micromoles per kilogram. Crucially, the three forms are not evenly split. At the surface ocean pH of about 8.1, roughly 88 percent of DIC is bicarbonate, roughly 11 percent is carbonate, and less than 1 percent is dissolved CO2 gas.

That skewed distribution is the buffer. Add acid, meaning extra H+ ions, and carbonate ions mop them up by converting to bicarbonate. Remove acid and bicarbonate gives H+ back. Because there is a large reservoir of bicarbonate and carbonate standing by, the ocean's pH resists change far better than pure water would. Surface ocean pH sits in a narrow band around 8.1, slightly alkaline, and has done so for a very long time.

A second quantity, total alkalinity, measures the ocean's overall capacity to neutralize acid and runs near 2,300 micromoles per kilogram. The pair of DIC and alkalinity, together with temperature and salinity, determines every other property of the carbonate system, including pH and the saturation state for shell-building. Lesson 15 takes that machinery apart in detail when we work through ocean acidification. For now hold the shape of it: the buffer is real, it is powerful, and it is finite.

Key idea: Dissolved CO2 reacts to form bicarbonate and carbonate, and that reservoir buffers seawater near pH 8.1, which is why the ocean can absorb enormous amounts of carbon without becoming acidic.

Nutrients: the scarce ingredients that run the ocean

Life needs more than carbon. The macronutrients are nitrogen, mostly as nitrate, and phosphorus, as phosphate, plus silicon as silicate for diatoms and radiolarians. These are present at micromolar levels, millions of times more dilute than sodium, and they are the constraint on almost all marine production.

Their vertical profile is the mirror image of oxygen's. Nutrients are stripped almost to zero in the sunlit surface layer, where phytoplankton consume them, and they are high at depth, where sinking organic matter decays and releases them. Nitrate in the middle of a subtropical gyre may be under 0.1 micromoles per kilogram at the surface and 35 or more at 1,000 meters. This is the fundamental tension of the ocean: light is at the top and nutrients are at the bottom, and life happens only where a process brings them together. That is why Module 3's circulation lessons are also biology lessons.

The proportions are remarkably consistent. Alfred Redfield noticed in the 1930s that marine plankton, and the deep water made from their decay, contain carbon, nitrogen, and phosphorus in an atomic ratio close to 106 to 16 to 1. This Redfield ratio lets you predict which nutrient runs out first.

Worked example. A parcel of water contains 4.0 micromoles of nitrate and 0.80 micromoles of phosphate per kilogram. Which limits growth? Using all the phosphate would require 16 times 0.80, which is 12.8 micromoles of nitrate. You have only 4.0. So nitrate is exhausted first: nitrogen is the limiting nutrient. When it runs out, the phosphate consumed is 4.0 divided by 16, which is 0.25 micromoles, leaving 0.55 micromoles of phosphate unused. This is why measuring leftover phosphate in a bloom tells you nitrogen was the constraint.

Finally, some regions have plenty of nitrate and phosphate and still grow little phytoplankton. These high-nutrient, low-chlorophyll regions include the Southern Ocean, the equatorial Pacific, and the subarctic North Pacific. The missing ingredient is iron, a micronutrient with a residence time of only a century or two because it is scavenged onto particles almost as fast as it arrives. Iron reaches the open ocean mainly as wind-blown dust, and the dust-starved Southern Ocean is the largest iron-limited region on Earth.

Key idea: Nutrients are depleted at the surface and rich at depth, plankton consume them near the Redfield ratio of 106 carbon to 16 nitrogen to 1 phosphorus, and in a third of the ocean the true limit is trace iron.

Common misconceptions

• "The ocean is getting saltier." Salt is removed about as fast as it is delivered, so total salinity has been broadly stable for hundreds of millions of years.
• "Constant proportions means salinity is the same everywhere." No. Total salinity varies from about 7 in the Baltic to over 40 in the Red Sea. The ratios among ions are what stay constant.
• "Sea ice is salty." Freezing largely excludes salt, so sea ice is nearly fresh, and the rejected brine sinks as some of the densest water in the ocean.
• "The ocean holds a lot of CO2 because CO2 dissolves easily." It holds a lot because CO2 reacts to form bicarbonate and carbonate, storing about fifty times as much carbon as the atmosphere.
• "Adding fertilizer nitrogen would grow plankton anywhere." In high-nutrient, low-chlorophyll regions the limiting factor is iron, not nitrogen, so nitrate additions do very little.

Recap

• Average salinity is 35 grams per kilogram; six ions supply more than 99 percent of it, with sodium chloride about 86 percent.
• Constant proportions means the ion ratios are fixed, which is why conductivity measures salinity.
• Residence time equals inventory divided by flux; water is about 3,100 years, sodium 75 million years, iron only a century or two.
• Long residence times relative to the roughly 1,000-year mixing time give uniform, conservative distributions.
• Surface salinity follows evaporation minus precipitation; brine rejection during freezing makes the densest seawater.
• Cold water holds more gas; the oxygen minimum zone at 200 to 1,000 meters records respiration of sinking organic matter.
• DIC is about 88 percent bicarbonate, 11 percent carbonate, under 1 percent dissolved CO2, buffering pH near 8.1.
• Redfield's 106:16:1 ratio identifies the limiting nutrient; iron limits the high-nutrient, low-chlorophyll regions.

Sources

1. National Ocean Service. (n.d.). Why is the ocean salty? NOAA. oceanservice.noaa.gov
2. Encyclopaedia Britannica. (n.d.). Seawater. britannica.com
3. Encyclopaedia Britannica. (n.d.). Salinity. britannica.com
4. Pacific Marine Environmental Laboratory. (n.d.). Ocean carbon and the carbonate system. NOAA. pmel.noaa.gov
5. National Centers for Environmental Information. (n.d.). World Ocean Atlas: temperature, salinity, oxygen, and nutrients. NOAA. ncei.noaa.gov
6. Wikipedia contributors. (n.d.). Redfield ratio. Wikipedia. en.wikipedia.org

The big picture

If you have ever swum in a lake in late summer and dropped your feet into water that felt shockingly cold a meter below the warm surface, you have felt a thermocline. The ocean has the same structure, but scaled up: a warm, well-stirred lid a hundred meters or so thick, then a zone where temperature falls fast, then a vast cold interior. Roughly three quarters of the ocean by volume sits between about 0 and 5 degrees Celsius, in permanent darkness, regardless of what the surface above it is doing.

That layering is not a detail. It is the single most consequential fact in physical oceanography, because layers separated by a density step do not mix easily. Heat added at the top stays near the top. Nutrients regenerated at depth stay at depth. Oxygen breathed in at the surface reaches the interior only where the layering is weak. Almost every question in this course, from why the tropics are biological deserts to how fast the ocean absorbs the heat we add to the atmosphere, comes back to how strongly the water column is stratified and where that stratification breaks down.

So this lesson builds the vertical structure carefully, starting from what makes one parcel of seawater denser than another.

What sets the density of seawater

Seawater density depends on three things: temperature, salinity, and pressure. Cold water is denser than warm water. Salty water is denser than fresh water. Deep water is very slightly denser than shallow water because it is compressed. Typical open-ocean density is about 1,027 kilograms per cubic meter, about 2.7 percent denser than pure fresh water.

Because the interesting variations are tiny, oceanographers usually report sigma-t, which is simply the density in kilograms per cubic meter minus 1,000. A sigma-t of 27.0 means 1,027.0. The whole open ocean spans roughly sigma-t 22 to 28. You are looking at differences of a few parts per thousand, and those few parts per thousand organize the entire circulation.

The changes are close to linear over ordinary ranges, which lets you calculate. A useful pair of coefficients: the thermal expansion coefficient is about 2 times 10 to the minus 4 per degree Celsius for warm surface water, and the haline contraction coefficient is about 7.6 times 10 to the minus 4 per unit of salinity.

Worked comparison. Cool a warm surface parcel by 1 degree. The density increase is 1,027 times 2 times 10 to the minus 4, which is about 0.21 kilograms per cubic meter. Now instead raise its salinity by 0.1. The density increase is 1,027 times 7.6 times 10 to the minus 4 times 0.1, which is about 0.078. So in warm water, cooling by 1 degree does roughly the same job as raising salinity by 0.26. Temperature is doing most of the work.

Now redo it in polar water. The thermal expansion coefficient of seawater shrinks dramatically as temperature approaches freezing, to roughly 0.5 times 10 to the minus 4 near 0 degrees. Cooling by 1 degree now buys only about 0.05 kilograms per cubic meter, which is equivalent to a salinity change of just 0.07. Temperature has almost stopped mattering; salinity now dominates.

Key idea: Density is set by temperature, salinity, and pressure, and temperature dominates in warm water while salinity dominates in polar water because thermal expansion nearly vanishes near freezing.

Two peculiarities of salt water worth knowing

Fresh water is famously strange: it reaches maximum density at 4 degrees Celsius and then becomes less dense as it cools further, which is why lakes freeze from the top and fish survive the winter. Seawater does not behave that way. Above a salinity of about 24.7, the temperature of maximum density falls below the freezing point, so seawater simply keeps getting denser all the way down to the moment it freezes. There is no 4-degree turnaround in the open ocean. Cold surface water sinks, full stop, and that is one reason deep convection is possible at high latitudes.

Second, dissolved salt depresses the freezing point. Seawater at salinity 35 freezes at about minus 1.9 degrees Celsius rather than 0. Polar surface water sitting at minus 1.8 degrees is liquid, not supercooled, and it is among the densest water on the planet.

Key idea: Seawater has no fresh-water density maximum above freezing and freezes near minus 1.9 degrees, so cold polar surface water can sink without limit.

The three-layer ocean

Put a temperature profile from a mid-latitude or tropical site on a page and you will see three zones.

The mixed layer is the top 25 to 200 meters. Wind, waves, and nighttime cooling stir it so thoroughly that temperature and salinity are nearly uniform from top to bottom. This is the layer that talks to the atmosphere. It holds essentially all the light, most of the life, and it deepens in winter and shallows in summer.

The pycnocline is the zone below it, typically from about 200 to 1,000 meters, where density rises rapidly with depth. When that density change is driven mostly by temperature it is called a thermocline; when driven by salinity, a halocline. In most of the ocean the thermocline and pycnocline are the same feature. Below about 1,000 meters lies the deep zone, cold, dark, nearly uniform, and containing roughly 80 percent of the ocean's volume.

The layer names matter less than what the pycnocline does. A strong density gradient is a mechanical barrier. Pushing a parcel of water across it requires work against buoyancy, so vertical exchange is suppressed. The pycnocline is the lid on the ocean's interior.

Two refinements. First, many mid-latitude regions develop a seasonal thermocline in the upper tens of meters each summer, riding above the permanent one, which then gets erased by autumn storms and winter cooling. Second, and more important, the permanent thermocline is a low and mid-latitude feature. Go poleward of about 60 degrees and the surface water is already close to the temperature of the deep water, so the profile becomes nearly uniform from top to bottom. There is no lid. That is precisely why deep water forms at high latitudes and nowhere else.

Key idea: The ocean is a stirred mixed layer over a pycnocline over a cold deep zone, and the pycnocline suppresses vertical exchange everywhere except at high latitudes where it disappears.

Latitude changes everything

Region	Surface temperature	Stratification	Consequence
Tropics	26 to 30 C	Very strong permanent thermocline	Nutrient-starved surface; clear blue, low productivity
Mid-latitudes	Seasonal, 5 to 25 C	Seasonal thermocline over a permanent one	Spring and autumn blooms as mixing changes
High latitudes	Minus 1.9 to 5 C	Weak or absent	Deep mixing, deep water formation, high productivity in summer

Read that table backwards and it explains ocean color from a plane. The famously clear, deep blue water of a tropical gyre looks beautiful because there is nothing living in it, and there is nothing living in it because a permanent thermocline keeps nutrients out of the light. The murky green water of a high-latitude summer is a bloom.

Key idea: Strong permanent stratification makes the tropical open ocean a biological desert, while weak stratification at high latitudes allows the nutrient resupply that supports blooms.

Temperature-salinity diagrams and water masses

Away from the surface, temperature and salinity are conservative properties: nothing in the interior heats a parcel or adds salt to it, so they change only by mixing with other water. That makes them a fingerprint. A parcel of water carries the temperature and salinity it acquired at the surface where it sank, sometimes for a thousand years.

Oceanographers exploit this with a T-S diagram, plotting temperature against salinity for every sample in a cast. A distinct water mass appears as a tight cluster; mixing between two water masses appears as a straight line joining their clusters. Four well-known examples:

• Antarctic Bottom Water, formed in the Weddell and Ross Seas: about minus 0.5 degrees, salinity 34.7. The densest water in the world ocean, hugging the deepest floor.
• North Atlantic Deep Water, formed in the Labrador and Nordic Seas: about 2 to 4 degrees, salinity 34.9. Fills much of the Atlantic between roughly 1,500 and 4,000 meters.
• Antarctic Intermediate Water: about 3 to 5 degrees, salinity near 34.2, recognizable as a low-salinity tongue at around 800 to 1,000 meters far into the Northern Hemisphere.
• Mediterranean Water: warm and extremely salty, near 11 degrees and salinity 36.5, spilling over the sill at Gibraltar and spreading across the Atlantic at about 1,000 meters.

This is genuinely detective work. Find a salty warm anomaly at 1,000 meters off the coast of Ireland and you have found water that left the Mediterranean years earlier. The T-S diagram is how physical oceanographers trace where interior water came from without ever tagging it.

Key idea: Temperature and salinity are conserved below the surface, so a T-S diagram identifies water masses by their formation-region fingerprint and reveals how they mix.

Why stratification is a climate variable

Stratification is not a fixed feature of the ocean. It changes, and it is changing now.

Warming the surface makes the top layer lighter. Melting ice and increased high-latitude rainfall freshen the surface, which also makes it lighter. Both effects increase the density contrast between the surface and the interior, which strengthens stratification. Observational syntheses find that upper-ocean stratification has increased measurably over recent decades.

Three consequences follow directly from what you now know. Stronger stratification means less nutrient supply from below, which tends to reduce primary production in already stratified regions. It means less oxygen delivered to the interior, contributing to the expansion of oxygen minimum zones. And it means heat and carbon taken up at the surface penetrate downward more slowly, which concentrates warming in the thin surface layer where marine heatwaves and coral bleaching happen. A single physical property, the strength of a density gradient, ties biology, chemistry, and climate together.

Key idea: Surface warming and freshening are strengthening stratification, which reduces nutrient and oxygen exchange with the interior and concentrates added heat near the surface.

Common misconceptions

• "The ocean is uniformly cold or uniformly warm." Neither. It is layered, and roughly three quarters of its volume sits between 0 and 5 degrees regardless of surface conditions.
• "Seawater is densest at 4 degrees, like fresh water." No. Above salinity 24.7 the density maximum lies below the freezing point, so seawater keeps densifying until it freezes.
• "Salinity is the main control on density everywhere." Only in polar water. In warm water, temperature dominates because thermal expansion is much larger there.
• "The pycnocline is a surface you could touch." It is a zone of rapid change, often hundreds of meters thick, not a boundary.
• "Clear blue tropical water is the healthiest water." Clarity means few particles, which usually means few plankton. Deep blue open ocean is nutrient-starved.

Recap

• Seawater density depends on temperature, salinity, and pressure, and averages about 1,027 kilograms per cubic meter.
• Sigma-t is density minus 1,000; the open ocean spans roughly 22 to 28.
• In warm water 1 degree of cooling equals about 0.26 salinity units; in polar water the same cooling equals only about 0.07, so salinity dominates there.
• Seawater has no 4-degree density maximum and freezes near minus 1.9 degrees at salinity 35.
• The ocean has a mixed layer, a pycnocline, and a cold deep zone holding about 80 percent of the volume.
• The permanent thermocline is absent at high latitudes, which is why deep water forms only there.
• T-S diagrams identify water masses such as AABW, NADW, AAIW, and Mediterranean Water by their conservative fingerprints.
• Warming and freshening are strengthening stratification, reducing nutrient and oxygen exchange with the interior.

Sources

1. National Ocean Service. (n.d.). Why is the ocean different colors in different places? NOAA. oceanservice.noaa.gov
2. National Centers for Environmental Information. (n.d.). World Ocean Atlas temperature and salinity climatologies. NOAA. ncei.noaa.gov
3. Encyclopaedia Britannica. (n.d.). Thermocline. britannica.com
4. Fox-Kemper, B., Hewitt, H. T., Xiao, C., et al. (2021). Chapter 9: Ocean, cryosphere and sea level change. In Climate change 2021: The physical science basis (IPCC AR6 WGI). ipcc.ch
5. Wikipedia contributors. (n.d.). Water mass. Wikipedia. en.wikipedia.org

The big picture

In 1893 Fridtjof Nansen deliberately froze his ship Fram into the Arctic ice to drift with it. He noticed something odd and wrote it down: the ice did not travel downwind. It consistently drifted 20 to 40 degrees to the right of the wind direction. Nansen mentioned the puzzle to a colleague, who handed it to a graduate student named Vagn Walfrid Ekman. In 1905 Ekman published the answer, and it turned out to explain not just drifting ice but the structure of every ocean gyre on the planet, the location of the world's great fisheries, and why the Gulf Stream is fast and narrow while the California Current is slow and broad.

This lesson is the payoff chapter of physical oceanography. Three ideas stack on top of each other: the Coriolis effect, Ekman transport, and geostrophic balance. Get them in that order and the surface circulation of the ocean stops being a map you memorize and becomes something you can derive.

The Coriolis effect, quantified

Earth rotates once a day. Anything moving freely across its surface, observed from the rotating surface, appears to curve: to the right in the Northern Hemisphere, to the left in the Southern. Nothing is pushing it sideways. The ground beneath it is turning while it travels. That apparent deflection is the Coriolis effect.

Its strength is measured by the Coriolis parameter, written f, which equals 2 times Earth's rotation rate times the sine of the latitude. Earth's rotation rate is 7.292 times 10 to the minus 5 radians per second.

Worked values. At 45 degrees latitude, the sine is 0.707, so f equals 2 times 7.292e-5 times 0.707, about 1.03 times 10 to the minus 4 per second. At 30 degrees, the sine is 0.5, giving 7.29 times 10 to the minus 5. At the pole, the sine is 1, giving 1.46 times 10 to the minus 4. At the equator, the sine is 0, so f is exactly zero and there is no Coriolis deflection at all. That last value has huge consequences and we will use it repeatedly.

Now settle the question everyone asks. Does the Coriolis effect drain your bathtub in a particular direction? Compute it. Take water moving at 10 centimeters per second, 0.1 meters per second, at 45 degrees latitude. The Coriolis acceleration is f times v, which is 1.03e-4 times 0.1, about 1.0 times 10 to the minus 5 meters per second squared. Over the two seconds the water takes to cross a bathtub, the sideways displacement is roughly one half times that acceleration times time squared: 0.5 times 1.0e-5 times 4, which is 2 times 10 to the minus 5 meters. That is 0.02 millimeters. The shape of the drain, the way you pulled the plug, and a stray air current each beat it by several orders of magnitude.

Now rerun it for the ocean. Take a current of 0.1 meters per second running for a full day, 86,400 seconds. The sideways displacement is 0.5 times 1.0e-5 times 86,400 squared, which is about 37 kilometers. Same physics, same latitude, same speed. The only difference is time and distance. Coriolis is negligible in a sink and utterly dominant in an ocean, and now you can say precisely why.

A related number is the inertial period, the time a freely moving parcel takes to loop a full circle under Coriolis alone. It equals 2 pi divided by f, which works out to 12 hours divided by the sine of the latitude. At 45 degrees that is about 17 hours; at 30 degrees, exactly 24 hours; near the equator it becomes infinite.

Key idea: The Coriolis parameter is 2 times Earth's rotation rate times the sine of latitude, zero at the equator, and it matters for ocean-scale motions because deflection accumulates over long times and distances, not because it is strong.

Ekman transport: why the ice drifted sideways

Wind blowing over water drags the very top layer along by friction. That thin surface layer, once moving, is itself deflected by Coriolis. It then drags the layer beneath it, which moves a little slower and is deflected a little further. Repeat downward and you get the Ekman spiral: each successive layer moves slower and more rotated than the one above, like a deck of cards fanned into a corkscrew, until at a depth of roughly 100 to 150 meters the motion becomes negligible.

In the idealized theory, the surface layer itself moves at 45 degrees to the right of the wind in the Northern Hemisphere. That was Nansen's ice, roughly. But the result that matters is what you get when you add up the movement of the whole spiral. The vector sum of the entire wind-driven layer, called Ekman transport, is directed 90 degrees to the right of the wind in the Northern Hemisphere and 90 degrees to the left in the Southern.

Say that plainly, because it is the most useful single sentence in this module: the net movement of wind-driven surface water is at right angles to the wind. A steady northward wind along a Northern Hemisphere coast moves water due east. Everything else in this lesson and the next follows from that.

Key idea: Summed over the whole wind-driven layer, Ekman transport carries water 90 degrees to the right of the wind in the Northern Hemisphere and 90 degrees to the left in the Southern.

Building a hill in the middle of the ocean

Now apply Ekman transport to the real wind field. Over the North Atlantic, trade winds blow from the northeast in the tropics and westerlies blow from the southwest around 40 degrees north. Turn each wind 90 degrees to the right and you get Ekman transport pointing northward from the trade-wind belt and southward from the westerly belt. Both point toward the middle of the basin, near 30 degrees north.

Water piles up there. It cannot pile up indefinitely, but it does build a genuine hill: the sea surface at the center of a subtropical gyre stands roughly one meter higher than at its edges, spread across a couple of thousand kilometers. This is why satellite altimetry was such a breakthrough, because that hill is directly measurable from orbit.

A hill of water creates a pressure gradient force pushing water outward and downhill. But as soon as the water starts moving downhill, Coriolis deflects it to the right. It turns, and keeps turning, until the rightward Coriolis force exactly opposes the downhill pressure gradient force. At that point the water is no longer flowing downhill at all. It is flowing along the contours of the hill, with the high water on its right.

That balance is called geostrophic flow, and it is the reason ocean currents circle rather than spread out. In the Northern Hemisphere, geostrophic flow around a mound of water is clockwise. In the Southern Hemisphere it is counterclockwise. Those are the subtropical gyres: five of them, in the North and South Atlantic, North and South Pacific, and Indian Ocean, each one a closed loop of water circling its own hill.

Poleward of the westerlies the wind pattern reverses and creates depressions rather than mounds, producing the smaller, counter-rotating subpolar gyres of the northern North Atlantic and North Pacific.

Key idea: Ekman transport converges water into a low mound at the center of each subtropical gyre, and the geostrophic balance between the downhill pressure force and Coriolis makes the water circle that mound instead of draining off it.

Western intensification

Look at any map of surface currents and you will notice an asymmetry that no amount of staring explains on its own. The western boundary current of each gyre, the Gulf Stream in the North Atlantic, the Kuroshio in the North Pacific, the Brazil and Agulhas currents in the south, is narrow, fast, deep, and warm. The eastern boundary current, the Canary, California, Peru, and Benguela currents, is broad, slow, shallow, and cool. The gyre is lopsided.

Property	Western boundary (Gulf Stream)	Eastern boundary (California Current)
Width	About 100 kilometers	About 1,000 kilometers
Speed	1 to 2.5 meters per second	Under 0.1 meters per second
Depth of flow	To 2,000 meters or more	A few hundred meters
Temperature	Warm, poleward-flowing	Cool, equatorward-flowing
Productivity	Low	High, because of upwelling

The cause is that the Coriolis parameter is not constant: it grows with latitude. Water circulating within a gyre must conserve its total spin, and because the planetary contribution to that spin changes as a parcel moves north or south, the return flow cannot be symmetric. The mathematics, worked out by Henry Stommel in 1948, requires the poleward return branch to be squeezed into a narrow, swift jet against the western edge of the basin. It is a genuine consequence of Earth's sphericity and rotation, and it applies to every ocean, in both hemispheres.

The numbers are worth carrying. Oceanographers measure volume transport in sverdrups, where 1 sverdrup is one million cubic meters per second. Every river on Earth combined delivers roughly 1.2 sverdrups to the ocean. The Gulf Stream carries about 30 sverdrups through the Florida Straits, growing to over 100 sverdrups further downstream as it entrains surrounding water. A single current moves about twenty-five times the world's entire river discharge past Miami, and roughly a hundred times it off Cape Hatteras. The largest current of all is the Antarctic Circumpolar Current, unobstructed by any continent, at well over 130 sverdrups.

Key idea: Because the Coriolis parameter increases with latitude, gyre return flow is compressed into narrow, fast western boundary currents, while eastern boundary currents are broad and sluggish.

How we know: ducks, shoes, drifters, and satellites

Currents are hard to measure because you cannot see them. Four approaches, in increasing sophistication, gave us the map.

Accidental drifters. In 1990 a storm knocked containers holding about 61,000 Nike shoes off a ship in the North Pacific. In 1992 another spill released roughly 28,800 plastic bath toys. Beachcombers reported where and when each landed, and oceanographers Curtis Ebbesmeyer and James Ingraham used the reports to test circulation models directly. It is one of the great examples of turning an accident into data.

Deliberate drifters. The Global Drifter Program maintains an array of satellite-tracked surface buoys, each with a drogue at 15 meters so it follows water rather than wind, reporting position and temperature continuously.

Moorings and profilers. Anchored instruments measure velocity at fixed depths for years, and Argo floats give the density field from which geostrophic currents can be computed.

Satellite altimetry. Because geostrophic flow is set by the slope of the sea surface, measuring that surface to a few centimeters from orbit yields global current maps. This is how we learned that the ocean is full of eddies tens to hundreds of kilometers across, which carry a large share of its kinetic energy.

Key idea: Surface currents are mapped from drifting objects, tracked buoys, moorings, and above all satellite altimetry, which converts sea-surface slope into geostrophic velocity.

Common misconceptions

• "Coriolis determines which way your sink drains." Over a bathtub the deflection is about 0.02 millimeters. Drain geometry beats it by orders of magnitude.
• "Surface currents flow in the direction the wind blows." Net Ekman transport is 90 degrees off the wind, and geostrophic currents flow along sea-surface contours, not downwind.
• "Coriolis is a force pushing objects sideways." Nothing pushes. It is the apparent curvature seen from a rotating reference frame.
• "Gyres are symmetric rings." They are strongly lopsided, with narrow fast currents on the western side of every basin.
• "The Gulf Stream is a jet of water crossing to Europe intact." It meanders, sheds rings and eddies, and broadens into the North Atlantic Current well before reaching Europe.

Recap

• The Coriolis parameter is 2 times Earth's rotation rate times the sine of latitude: about 1.03e-4 per second at 45 degrees, zero at the equator.
• Coriolis is negligible in a sink and dominant in the ocean because deflection accumulates over long times and distances.
• The Ekman spiral sums to a net transport 90 degrees right of the wind in the Northern Hemisphere, left in the Southern.
• Converging Ekman transport builds a sea-surface hill about one meter high at the center of each subtropical gyre.
• Geostrophic balance between the pressure gradient and Coriolis makes water circle the hill, clockwise in the north.
• Because f increases with latitude, western boundary currents are narrow, fast, and deep; eastern ones are broad and slow.
• Transport is measured in sverdrups; the Gulf Stream carries 30 to over 100, against about 1.2 for all rivers combined.

Sources

1. National Ocean Service. (n.d.). Ocean currents education tutorial. NOAA. oceanservice.noaa.gov
2. National Ocean Service. (n.d.). What is the Gulf Stream? NOAA. oceanservice.noaa.gov
3. Encyclopaedia Britannica. (n.d.). Coriolis force. britannica.com
4. Encyclopaedia Britannica. (n.d.). Ocean current. britannica.com
5. NASA Science. (n.d.). Ocean surface topography and currents. sealevel.nasa.gov
6. Wikipedia contributors. (n.d.). Ekman transport. Wikipedia. en.wikipedia.org

The big picture

Stand on a beach in northern Peru in the tropics, at a latitude where the water ought to be bath-warm, and wade in. It is about 16 degrees Celsius and it will take your breath away. Offshore, that cold water supports one of the largest single-species fisheries in the history of the world. Just a few hundred kilometers west, in the open Pacific at the same latitude, the water is 26 degrees, sapphire blue, and nearly empty of life.

The difference is upwelling, and by now you have everything you need to derive it. The previous lesson gave you Ekman transport at right angles to the wind. Apply that one rule at a coastline and the world's great fisheries appear on the map exactly where they actually are. Then, in the second half of this lesson, we turn from the wind-driven surface to the density-driven deep, and follow a parcel of water on a journey that takes it a thousand years and most of the way around the planet.

Coastal upwelling

Picture a north-south coastline on the eastern side of an ocean basin in the Northern Hemisphere, say California. The wind blows from the north, southward along the coast. Ekman transport moves the surface water 90 degrees to the right of the wind, which is westward: straight offshore.

Surface water is being carried away from the coast. Something must replace it, and it cannot come from the land. So water is drawn upward from below, from perhaps 100 to 200 meters down, and that is coastal upwelling. The water arriving is cold, because it comes from beneath the thermocline. It is also loaded with the nitrate and phosphate that Lesson 5 showed accumulate at depth as sinking organic matter decays. Cold nutrient-rich water arriving in full sunlight is the recipe for explosive phytoplankton growth.

Reverse the wind and you reverse the process: coastal downwelling pushes surface water against the coast and forces it down, capping the region with warm, nutrient-poor water.

The mirror-image rule holds in the Southern Hemisphere. Off Peru, the wind blows from the south, northward along the coast, and Ekman transport is 90 degrees to the left, again offshore. This is why all four of the great eastern boundary upwelling systems, the California, Canary, Peru or Humboldt, and Benguela systems, sit on the eastern side of ocean basins with equatorward winds. Not one is an accident of geography; each is Ekman transport plus a coastline.

Key idea: Winds blowing equatorward along an eastern ocean boundary drive Ekman transport offshore, and cold nutrient-rich water rises to replace it, creating the world's most productive fishing grounds.

Working an upwelling calculation

Upwelling looks dramatic in its consequences and is astonishingly gentle in its mechanics. Let us get the number.

Ekman transport per meter of coastline equals the wind stress divided by the product of water density and the Coriolis parameter. Take a moderate alongshore wind giving a stress of 0.1 newtons per square meter, a density of 1,027 kilograms per cubic meter, and a latitude of 35 degrees where f is 8.4 times 10 to the minus 5 per second.

Transport equals 0.1 divided by (1,027 times 8.4e-5). The denominator is 0.0863. So transport is about 1.16 square meters per second, meaning 1.16 cubic meters of water per second moves offshore for every meter of coastline.

Scale it up. Along 1,000 kilometers of coast, that is 1.16 times 10 to the 6th cubic meters per second, about 1.2 sverdrups: roughly the entire river discharge of the planet, moved offshore by one steady wind.

Now get the vertical speed. If that offshore transport is fed by water rising within a band about 50 kilometers wide, the upwelling velocity is 1.16 square meters per second divided by 50,000 meters, which is 2.3 times 10 to the minus 5 meters per second. Convert to a day: multiply by 86,400 seconds and you get about 2 meters per day.

Two meters a day. A swimmer could out-climb it by a factor of thousands. But it runs continuously for months, so water from 120 meters down reaches the surface in about two months, and it keeps arriving. Upwelling is not violent. It is relentless, and that is enough.

Key idea: Upwelling velocities are only a few meters per day, but sustained for months they resupply the entire sunlit layer with deep nutrients.

Equatorial upwelling and the divergence rule

There is a second place upwelling happens for free, and it follows from the sign flip in the Coriolis parameter. The trade winds blow from east to west along the equator. Just north of the equator, Ekman transport is 90 degrees to the right of that wind, which is northward. Just south of the equator, transport is 90 degrees to the left, which is southward. Surface water moves away from the equator on both sides at once.

That is a divergence, and it must be fed from below. The result is a band of equatorial upwelling running across the Pacific, visible in satellite chlorophyll images as a green ribbon along the equator in an otherwise blue ocean. The same logic, applied around Antarctica where westerlies drive northward transport, produces the Antarctic Divergence and its upwelling of nutrient-rich deep water.

The general rule is worth stating once: wherever Ekman transport diverges, water upwells; wherever it converges, water downwells. Gyre centers are convergences, which is why they are deserts. Equatorial and coastal upwelling zones are divergences, which is why they are gardens.

The economic weight of this is large. Coastal upwelling zones make up well under 1 percent of the ocean's surface area, and yet they yield on the order of 20 percent of the world's wild fish catch.

Key idea: Ekman divergence upwells and Ekman convergence downwells, which is why the equator and eastern boundaries are productive while gyre centers are barren.

The deep circulation: how water sinks

Everything so far has been wind-driven and confined to roughly the upper few hundred meters. Below that, a slower and vastly larger circulation is driven by density, meaning by temperature and salinity together. That is why it is called the thermohaline circulation.

Water sinks only where it becomes denser than everything beneath it, and that requires two conditions from earlier lessons. First, it must be cold, which means high latitude. Second, there must be no strong pycnocline blocking the way, which is again only true at high latitudes. Add the salt from brine rejection during sea-ice formation and you have the recipe.

Two regions dominate.

• The northern North Atlantic. In the Nordic and Labrador Seas, warm salty water carried north by the Gulf Stream system gives up its heat to the winter atmosphere. It has kept its high salinity while losing its warmth, so it becomes very dense and sinks, forming North Atlantic Deep Water at roughly 2 to 4 degrees and salinity 34.9.
• The Antarctic margins. On the Weddell and Ross Sea shelves, sea ice forms and rejects brine into water already at the freezing point. The result, Antarctic Bottom Water near minus 0.5 degrees, is the densest water in the world ocean and slides down the continental slope to fill the deepest basins everywhere.

Notice what is missing. The North Pacific makes no deep water at all, because its surface is too fresh: heavy precipitation and river runoff keep a low-salinity cap in place that cooling alone cannot overcome. Sinking is a chemistry problem as much as a temperature problem.

Key idea: Deep water forms only in the northern North Atlantic and around Antarctica, where cold temperatures, high salinity, and weak stratification coincide; the North Pacific is too fresh at the surface to sink.

The global conveyor and its timescale

Follow the water. North Atlantic Deep Water sinks and flows southward at depth through the Atlantic. Near Antarctica it joins the Antarctic Circumpolar Current, mixes with Antarctic Bottom Water, and is distributed eastward into the Indian and Pacific basins. There it slowly rises, mixed upward by tides and winds and drawn up in the Southern Ocean, and eventually returns toward the Atlantic as warmer upper-ocean flow. This loop is often called the global conveyor belt, and it links every ocean basin into a single circulation.

Worked timescale. Suppose the deep ocean holds about 1.3 times 10 to the 18th cubic meters and deep water forms at roughly 20 sverdrups, which is 2 times 10 to the 7th cubic meters per second. The time to replace all of it is 1.3e18 divided by 2e7, which is 6.5 times 10 to the 10th seconds. Divide by about 3.15 times 10 to the 7th seconds per year and you get roughly 2,000 years. Independent radiocarbon dating of deep water gives ages of order 1,000 to 1,500 years for the oldest deep Pacific water. Two very different methods agree on the order of magnitude, and that is the number to carry: the deep ocean turns over on a timescale of about a thousand years.

That single figure explains a great deal you have already met. It explains why the deep Pacific has the lowest oxygen and the highest nutrients, having been away from the surface longest. It explains why the Pacific carbonate compensation depth is shallower, since accumulated respired CO2 makes the oldest water most corrosive. And it explains a hard truth about climate: heat and carbon put into the deep ocean today will not come back out for centuries, and carbon we emit now will still be working its way through that loop long after everyone reading this is gone.

Key idea: The deep ocean overturns on a timescale near a thousand years, which sets the age gradients in oxygen, nutrients, and carbonate chemistry and locks in centuries of climate commitment.

The Atlantic overturning and how much to worry

The Atlantic limb of this system, the Atlantic Meridional Overturning Circulation or AMOC, carries warm water north near the surface and cold water south at depth, and it moves an enormous amount of heat northward, on the order of 1.3 petawatts. Since 2004 the RAPID mooring array at 26.5 degrees north has measured its strength continuously at roughly 17 sverdrups, with large variability.

Here is where a careful teacher has to slow down. It is often said that the Gulf Stream is the only reason northwest Europe is mild, and that a shutdown would freeze Britain. That overstates a real effect. Studies show that a substantial part of the mild European winter comes from prevailing westerly winds blowing across an ocean that stores summer heat, and from the atmosphere's own heat transport, not from ocean heat transport alone. The overturning matters, but it is one term in a sum.

What we can say with confidence is this. Freshening the northern North Atlantic, whether by melting Greenland ice or by increased rainfall, makes surface water lighter and works against sinking. That is straightforward physics from Lesson 6. Paleoclimate records show abrupt North Atlantic climate shifts, including the Younger Dryas, associated with disruptions of this circulation. The IPCC assessment concludes it is very likely that the AMOC will weaken over the twenty-first century, while judging a complete collapse before 2100 to be unlikely. Honest summary: a weakening is expected, a collapse is a low-probability high-impact tail risk, and the observational record is still too short to settle the trend cleanly.

Key idea: The AMOC carries about 1.3 petawatts of heat northward and is very likely to weaken this century, but attributing all of Europe's mild climate to it, or predicting imminent collapse, goes beyond the evidence.

Common misconceptions

• "Upwelling means water rushing up from the abyss." It rises at a few meters per day from a few hundred meters at most, and its power lies in persistence, not speed.
• "Upwelling happens where wind blows toward shore." The opposite. Alongshore equatorward winds drive Ekman transport offshore, which is what forces water up.
• "The conveyor is a single narrow river of water." It is a broad, diffuse, three-dimensional overturning with mixing everywhere; the belt cartoon is a simplification.
• "Deep water forms in the North Pacific too." It does not. The North Pacific surface is too fresh for cooling alone to make it dense enough to sink.
• "If the AMOC weakened, Europe would enter an ice age." Ocean heat transport is one contributor among several, and assessments put full collapse this century as unlikely.

Recap

• Equatorward winds on eastern boundaries drive offshore Ekman transport, forcing cold nutrient-rich water upward.
• A 0.1 newton per square meter wind stress at 35 degrees gives about 1.16 square meters per second of offshore transport and roughly 2 meters per day of upwelling.
• Ekman divergence upwells and convergence downwells, so the equator and coasts are productive and gyre centers are deserts.
• Coastal upwelling zones are under 1 percent of ocean area but yield about 20 percent of the wild fish catch.
• Deep water forms only in the northern North Atlantic and around Antarctica; the North Pacific is too fresh.
• Deep ocean volume divided by about 20 sverdrups gives an overturning time near 2,000 years, and radiocarbon gives 1,000 to 1,500.
• The AMOC moves about 1.3 petawatts northward, is measured near 17 sverdrups at 26.5 degrees north, and is very likely to weaken this century.

Sources

1. National Ocean Service. (n.d.). What is upwelling? NOAA. oceanservice.noaa.gov
2. National Ocean Service. (n.d.). What is the global ocean conveyor belt? NOAA. oceanservice.noaa.gov
3. Encyclopaedia Britannica. (n.d.). Thermohaline circulation. britannica.com
4. Fox-Kemper, B., Hewitt, H. T., Xiao, C., et al. (2021). Chapter 9: Ocean, cryosphere and sea level change. In Climate change 2021: The physical science basis (IPCC AR6 WGI). ipcc.ch
5. Woods Hole Oceanographic Institution. (n.d.). Ocean learning hub: ocean circulation. whoi.edu
6. Wikipedia contributors. (n.d.). Atlantic meridional overturning circulation. Wikipedia. en.wikipedia.org

The big picture

On a calm, windless morning in California, clean 3-meter swells arrive at the beach at regular 18-second intervals. There is no storm anywhere nearby. There is no wind. The waves come from a gale that blew itself out five days ago, ten thousand kilometers away in the Southern Ocean. They crossed an entire hemisphere, sorting themselves as they went, and arrived in an orderly procession.

Everything in that paragraph is calculable, and this lesson gives you the tools. Waves are the most quantitative part of introductory oceanography, and the formulas are short. Two of them do almost all the work, one for deep water and one for shallow. Learn when each applies and you can predict swell speed, wavelength, arrival time, breaker depth, and the direction sand moves along a beach.

Start with the single most important thing to understand about a wave, which is that the water does not travel with it. A wave is a moving disturbance passing through water that mostly stays put. Watch a gull sitting on the swell: it rises, moves slightly forward, sinks, moves slightly back, and returns to almost exactly where it started while wave after wave rolls beneath it. What travels across the ocean is energy, not water.

Anatomy and orbital motion

Name the parts. The crest is the high point, the trough the low point. Wave height, written H, is the vertical distance from trough to crest. Wavelength, L, is the distance from one crest to the next. The period, T, is the time for two successive crests to pass a fixed point, and it is the easiest property to measure from a pier with a stopwatch. Steepness is H divided by L, and it matters more than height alone.

Beneath the surface, water particles move in circles, called orbits. At the surface the orbital diameter equals the wave height. With depth the orbits shrink rapidly, and by a depth of half a wavelength they are negligible. That depth is the wave base: below it, the wave might as well not exist.

Wave base explains a lot. A submarine 100 meters down rides out a hurricane in perfect calm. A 10-second wave has a wavelength of about 156 meters, so its wave base is 78 meters, and sediment on a 90-meter-deep seafloor is undisturbed by even a big storm swell. Wave base is also the dividing line for the two formulas that follow.

Key idea: Waves transmit energy, not water; particles move in orbits that shrink with depth and vanish below the wave base at half a wavelength.

Deep-water waves: the workhorse formula

A wave is a deep-water wave when the water depth is greater than half its wavelength, so the orbits never touch bottom. Almost every wave in the open ocean qualifies.

For deep-water waves, speed depends only on period. In metric units, with g equal to 9.81 meters per second squared, gravity divided by 2 pi is about 1.56, giving two formulas you should memorize:

Speed in meters per second is about 1.56 times the period in seconds. Wavelength in meters is about 1.56 times the period squared.

Worked example. A 10-second swell. Speed is 1.56 times 10, which is 15.6 meters per second, about 56 kilometers per hour. Wavelength is 1.56 times 100, which is 156 meters. Check for consistency: speed should equal wavelength divided by period, and 156 divided by 10 is 15.6. It works.

Now a 20-second swell. Speed is 31.2 meters per second, about 112 kilometers per hour, and wavelength is 1.56 times 400, which is 624 meters. Doubling the period doubles the speed and quadruples the wavelength. Long waves are fast waves, and this fact organizes the whole subject.

Key idea: In deep water, speed is about 1.56 times the period and wavelength is about 1.56 times the period squared, so longer waves travel faster.

Dispersion and the sorting of swell

A storm generates a chaotic mess of waves with many periods at once, which sailors call sea. As that mess leaves the storm, the long-period waves outrun the short-period ones. This sorting by speed is dispersion, and the smooth, regular, long-period waves that arrive at a distant coast are called swell. Swell is simply sea that has been filtered by a long journey.

Worked example. A storm 5,000 kilometers away generates waves of 20-second and 8-second period. The 20-second wave has a phase speed of 31.2 meters per second; 5,000,000 meters divided by 31.2 gives about 160,000 seconds, or 44.5 hours. The 8-second wave travels at 12.5 meters per second; 5,000,000 divided by 12.5 is 400,000 seconds, or 111 hours. The long swell arrives roughly 67 hours, nearly three days, before the short one.

One honest complication that separates a real forecast from a textbook one. In deep water, the energy of a wave group travels at the group speed, which is exactly half the phase speed. Individual crests appear at the back of a group, march forward through it, and vanish at the front. So the swell energy from that 20-second storm actually arrives after about 89 hours, not 44.5. When surf forecasters compute arrival times they use group speed, and if you ever compute one and find the swell shows up twice as fast as reality, this is why.

Key idea: Dispersion sorts storm waves by period so long swell arrives first, but wave energy travels at group speed, half the phase speed, in deep water.

How big can a wave get?

Three factors set wave size: wind speed, wind duration, and fetch, the uninterrupted distance over which the wind blows. Limit any one and you limit the waves. A 40-knot wind across a 2-kilometer lake makes chop, not swell, because fetch is tiny. When speed, duration, and fetch are all sufficient, the sea stops growing: energy input from the wind is matched by energy lost to breaking whitecaps, and you have a fully developed sea. For a 20-knot wind that takes about 10 hours and 140 kilometers of fetch, producing an average height near 1.5 meters. For 40 knots it takes roughly 42 hours and 1,300 kilometers, giving average heights near 8.5 meters.

Two reporting conventions matter. Significant wave height, the number in every marine forecast, is the average height of the highest one third of waves, which corresponds roughly to what an experienced observer estimates by eye. And because wave energy is proportional to the square of height, a 4-meter wave carries four times the energy of a 2-meter wave, not twice. That square law is why storm damage rises so sharply with wave height.

Rogue waves, defined as waves more than twice the significant wave height, are real. On 1 January 1995 an instrument on the Draupner platform in the North Sea recorded a wave of about 25.6 meters in a sea whose significant height was around 12 meters, ending a long argument about whether sailors' accounts were exaggerations.

Key idea: Wave size is limited by whichever of wind speed, duration, or fetch runs out first, and energy scales with the square of height.

Shallow-water waves, shoaling, and breaking

A wave becomes a shallow-water wave when the depth is less than about one twentieth of its wavelength. Now the orbits are flattened by the bottom, and the speed no longer depends on period at all. Instead:

Speed equals the square root of gravity times depth.

Worked example. In 5 meters of water, speed is the square root of 9.81 times 5, which is the square root of 49, about 7.0 meters per second. In 20 meters, it is the square root of 196, about 14 meters per second. Shallower means slower, and that single fact drives everything that happens at a coast.

Follow a swell into the shallows. The period cannot change, since crests keep arriving at the same rate. But as the depth falls the wave slows down, so the wavelength must shorten. The energy has to go somewhere, and it goes into height. The wave grows and steepens. This is shoaling.

The wave breaks when the crest outruns the base it is standing on. Two equivalent criteria are used. The wave becomes unstable when steepness H divided by L exceeds about 1 in 7, and in practice a wave breaks where the water depth is about 1.3 times the wave height, which is usually written as a height-to-depth ratio near 0.78.

Worked example. A 2-meter wave breaks in a depth of 2 divided by 0.78, which is about 2.6 meters. A 4-meter wave needs about 5.1 meters. That is why big-wave surf breaks far offshore over a reef while small summer surf breaks near the sand.

Bottom slope determines the breaker type. On a gentle slope you get spilling breakers, foaming down the front over a long distance. On a moderately steep slope you get plunging breakers, the curling barrels surfers want. On a very steep beach you get surging breakers that rush up the face without truly breaking.

Key idea: Shallow-water speed is the square root of gravity times depth, so waves slow, shorten, and steepen as they shoal, breaking where depth is about 1.3 times their height.

Refraction, longshore transport, and the river of sand

Because shallow-water speed depends on depth, a wave crest arriving at an angle to shore has one end in shallower, slower water than the other. That end lags, and the crest pivots. This is wave refraction, and it bends wave crests to become more nearly parallel to the shoreline.

Refraction has a consequence that shapes coastlines. Off a headland the water is shallower, so waves refract toward it and their energy converges there. In a bay the water is deeper, so waves refract away and energy spreads out. Headlands are attacked and eroded; bays are sheltered and accumulate sand. Refraction is the reason irregular coastlines slowly straighten themselves.

Refraction is never complete, so waves usually still strike the beach at a small angle. Each breaking wave rushes up the sand at that angle and drains straight back down under gravity, moving each grain a small distance along the shore in a zigzag. Repeat this millions of times a day and you have a longshore current in the surf zone and longshore transport of sediment. A beach is not a static pile of sand. It is a river of sand, continuously flowing along the coast.

That reframing is the whole of coastal engineering. Build a groin or a jetty across the flow and sand accumulates on the updrift side and starves the downdrift side, so protecting one beach erodes the next town's. Build a seawall and you protect the property behind it while the beach in front of it disappears, because the wall reflects wave energy downward and scours the sand away. Dam a river and you cut off the sediment supply that fed the beach in the first place. None of these outcomes are surprises; they follow from the physics in this section.

A final safety note that this physics explains. Water piled onshore by breaking waves must return seaward, and it does so through narrow, fast channels called rip currents, which cause most surf-zone rescues. A rip pulls you offshore, not under. The correct response is to swim parallel to the beach until you are out of the narrow channel, then come in.

Key idea: Refraction concentrates energy on headlands and disperses it in bays, and the residual angle drives a longshore river of sand that every coastal structure interrupts.

Common misconceptions

• "Waves carry water across the ocean." They carry energy. Water particles orbit and return nearly to where they began.
• "Big waves are dangerous because they are tall." Energy scales with height squared, so doubling height quadruples the force delivered.
• "A rip current pulls you under." It pulls you seaward in a narrow channel. Swim parallel to shore to exit it.
• "Swell speed is the speed the storm energy travels." In deep water, energy travels at group speed, half the phase speed of individual crests.
• "A seawall saves the beach." It protects what is behind it and typically accelerates loss of the beach in front of it.

Recap

• Waves transmit energy; orbital motion decays to nothing below wave base at half a wavelength.
• Deep water: speed is about 1.56 times period, wavelength about 1.56 times period squared.
• Dispersion sorts swell so long periods arrive first, but energy moves at group speed, half the phase speed.
• Wave size is capped by wind speed, duration, or fetch; energy is proportional to height squared.
• Shallow water: speed is the square root of gravity times depth, independent of period.
• Shoaling waves slow, shorten, and steepen, breaking where depth is about 1.3 times the height.
• Refraction focuses energy on headlands and shelters bays; the residual angle drives longshore transport.
• Groins, jetties, seawalls, and dams all interrupt the longshore river of sand with predictable consequences.

Sources

1. National Ocean Service. (n.d.). Waves. NOAA. oceanservice.noaa.gov
2. National Ocean Service. (n.d.). What is a rip current? NOAA. oceanservice.noaa.gov
3. National Ocean Service. (n.d.). What is longshore drift? NOAA. oceanservice.noaa.gov
4. U.S. Geological Survey. (n.d.). Coastal change hazards. usgs.gov
5. Wikipedia contributors. (n.d.). Wind wave. Wikipedia. en.wikipedia.org

The big picture

On the morning of 26 December 2004, at a beach in Thailand, a ten-year-old English girl named Tilly Smith watched the sea pull back much further than it should have and start to froth. Two weeks earlier her geography teacher had shown her class a video about tsunamis. She recognized what she was seeing, told her parents, and the beach was cleared. Everyone on it survived. That day the Sumatra-Andaman earthquake, magnitude 9.1, generated a tsunami that killed roughly 228,000 people around the Indian Ocean, in many places because nobody knew what a retreating sea meant.

This lesson covers the two longest waves in the ocean. Tsunamis have wavelengths of a hundred kilometers or more. Tides have wavelengths of thousands. Both are so long compared with ocean depth that they behave as shallow-water waves everywhere on Earth, even over the deepest trench, and that single fact governs almost everything about them. If you understood the shallow-water speed formula from the previous lesson, you already have the key.

What a tsunami is, and what it is not

A tsunami is a wave generated by the sudden vertical displacement of a large volume of water. The usual cause is a subduction-zone earthquake in which the seafloor lurches upward or downward by meters over hundreds of kilometers, lifting or dropping the entire water column above it. Submarine landslides, volcanic flank collapses, and, rarely, asteroid impacts can do the same. In 1958, a rockslide into Lituya Bay in Alaska produced a wave that stripped trees to a height of 524 meters, the highest run-up ever recorded.

It is not a tidal wave. It has nothing whatever to do with the tides. It is also not a wind wave, and the difference is not one of size but of kind.

Property	Storm wave	Tsunami
Wavelength	100 to 300 meters	100 to 200 kilometers
Period	5 to 20 seconds	10 minutes to 2 hours
Deep-ocean height	1 to 15 meters	Often under 1 meter
Speed in 4,000 m of water	Roughly 20 to 30 meters per second	About 200 meters per second
Wave type	Deep-water wave	Shallow-water wave everywhere
Water involved	Upper tens of meters	The entire water column, seafloor to surface

The last row is the one that kills people. A storm wave is a disturbance in the top of the ocean. A tsunami moves the whole column, which is why it does not break and drain away like surf but arrives as a fast, prolonged, relentless flood that keeps coming for minutes.

Key idea: A tsunami is generated by sudden displacement of the whole water column, so unlike a wind wave it involves the entire ocean depth and arrives as a sustained flood rather than a breaking crest.

Working the tsunami numbers

Check the shallow-water condition first. A wave is a shallow-water wave when depth is less than about one twentieth of the wavelength. A tsunami wavelength of 180 kilometers gives a threshold of 9 kilometers. The deepest trench on Earth is under 11 kilometers and the average ocean is under 4. So yes: a tsunami is a shallow-water wave over essentially the entire ocean, and its speed is set by depth alone.

Speed. In water 4,000 meters deep, speed is the square root of 9.81 times 4,000, which is the square root of 39,240, about 198 meters per second. Convert: multiply by 3.6 to get about 713 kilometers per hour. A tsunami crosses the deep ocean at the cruising speed of a jetliner.

Wavelength. If the period is 15 minutes, which is 900 seconds, the wavelength is speed times period: 198 times 900 is about 178,000 meters, or 178 kilometers. That is consistent with the assumption we started from, so the calculation is self-consistent.

Why ships never notice. Take an open-ocean amplitude of 0.5 meters spread over 178 kilometers of wavelength. The steepness is 0.5 divided by 178,000, roughly 1 in 356,000. Rising half a meter over 89 kilometers is a slope no ship's crew could possibly detect. In 2004, vessels in deep water felt nothing at all while the wave passed beneath them at 700 kilometers per hour.

Travel time. Sri Lanka lies roughly 1,600 kilometers from the 2004 rupture. At 713 kilometers per hour that is about 2.2 hours, and the wave did indeed arrive about two hours later. This is exactly the arithmetic a warning center performs, and it is why warnings are possible at all: seismic waves travel through rock at several kilometers per second, roughly twenty times faster than the tsunami crosses water, so an earthquake is detected long before the wave arrives anywhere distant.

Why it grows at the coast. In 10 meters of water the speed is the square root of 98.1, about 9.9 meters per second, or 36 kilometers per hour. The front of the wave slows by a factor of twenty while the back is still racing, so an enormous length of wave compresses into a short one and the energy goes into height. A useful approximation called Green's law says height grows roughly as the fourth root of the depth ratio. From 4,000 meters to 10 meters, that ratio is 400, whose fourth root is about 4.5, turning a 0.5-meter open-ocean wave into roughly 2.2 meters at the shoreline. Real run-ups are often far larger still, because bays funnel and focus the energy and because the water keeps arriving; in the 2011 Tohoku tsunami, local run-up exceeded 30 meters in some inlets.

Key idea: Tsunami speed is the square root of gravity times depth, about 700 kilometers per hour in the deep ocean, and slowing at the coast compresses the wave and converts its length into height.

Warning, and what to actually do

Modern warning rests on two systems. Seismic networks locate and size an earthquake within minutes and flag those capable of generating a tsunami. Then DART buoys, each with a pressure recorder on the seafloor sensitive enough to detect a sea-level change of about a centimeter in kilometers of water, confirm whether a wave was actually produced and how big it is. Confirmation matters, because evacuating a coastline for a wave that never comes destroys public trust in the next warning.

The practical rules are short. If you are on a coast and feel a strong earthquake, do not wait for an official warning; a local tsunami can arrive in ten minutes. If the sea withdraws unusually far, that is the trough of the wave arriving first, and the crest is minutes behind. If you hear a loud roar from the ocean, go. In every case, move inland and to high ground, and stay there, because the second or third wave is often larger than the first and the danger lasts for hours.

Key idea: Seismic detection plus DART buoy confirmation gives distant coasts hours of warning, but for a local tsunami the ground shaking or a receding sea is the only warning you will get.

Tides: why the Moon beats the Sun

Tides are the rise and fall of sea level caused by the gravitational pull of the Moon and Sun on a rotating Earth. Here is the point that most explanations skip, and it is the interesting part.

Gravitational attraction falls off as the inverse square of distance. But tides are not caused by gravity; they are caused by the difference in gravity across the width of Earth, pulling the near side harder than the center and the center harder than the far side. That differential, the tide-generating force, falls off as the inverse cube of distance. One extra power of distance changes everything.

Worked comparison. The Sun is about 27 million times more massive than the Moon. It is also about 389 times farther away. The ratio of their tide-generating forces is the mass ratio divided by the cube of the distance ratio. Compute 389 cubed: 389 times 389 is 151,300, times 389 again is about 5.89 times 10 to the 7th. Now divide 2.7 times 10 to the 7th by 5.89 times 10 to the 7th and you get about 0.46.

So the Sun's tidal influence is roughly 46 percent of the Moon's, even though its gravitational pull on Earth is about 178 times stronger. Distance cubed beats mass. This one calculation explains why we speak of lunar tides with a solar modulation rather than the other way round.

Key idea: The tide-generating force varies as the inverse cube of distance, not the inverse square, which is why the Moon produces about twice the tidal effect of the vastly more massive Sun.

Two bulges, and a 50-minute lag

The differential pull produces two bulges of water, not one. On the side of Earth facing the Moon, lunar gravity is stronger than average and water is pulled toward the Moon. On the opposite side, lunar gravity is weaker than average, so relative to the accelerating Earth the water there is effectively left behind, producing a second bulge pointing away. Earth rotates through both bulges, which is why most coasts get two high tides a day rather than one.

Now the timing. While Earth spins once, the Moon has moved along its orbit, so a point on Earth must rotate a little further to catch up. A lunar day is 24 hours and 50 minutes rather than 24. Divide by two bulges and the interval between successive high tides is about 12 hours 25 minutes, and high tide arrives roughly 50 minutes later each day. If you have ever noticed the tide table shifting by nearly an hour a day, that is the Moon's orbital motion, visible from a pier.

The Sun adds its 46 percent, and the two combine according to their alignment. When the Sun, Earth, and Moon line up, at new moon and full moon, the two tidal effects reinforce and you get spring tides, with the largest range. The word has nothing to do with the season; it means to spring forth. When the Moon is at first or last quarter, the Sun and Moon pull at right angles, partly cancelling, giving neap tides with the smallest range. Because this depends on the Moon's phase, spring and neap tides alternate about every 7.4 days, with springs roughly twice a month.

Two further cycles fine-tune this. The Moon's orbit is elliptical, so when it is at perigee, its closest approach, tides are amplified; a spring tide coinciding with perigee produces the unusually high water people call a king tide. And the Moon's declination, its angle above or below the equatorial plane, makes the two daily bulges unequal.

Key idea: Two bulges plus a 24-hour 50-minute lunar day give high tides every 12 hours 25 minutes, and solar alignment produces spring tides at new and full moon and neap tides at the quarters.

Why real tides do not match the simple picture

Everything above is equilibrium theory, which assumes a water-covered Earth with bulges free to follow the Moon. Reality is different in three ways, and the differences are what make tide tables necessary.

First, continents are in the way. The bulges cannot travel freely around the planet; they are broken into separate basins that each slosh in their own way.

Second, the tide is itself a shallow-water wave and it cannot keep up. To follow the Moon at the equator, a bulge would need to travel at about 465 meters per second. In 4,000 meters of water, the shallow-water speed is only about 198 meters per second. The bulge is chasing a target it can never catch, so what actually exists is a forced wave sloshing within each basin.

Third, Coriolis deflects that sloshing, so in each basin the tide rotates around one or more amphidromic points, nodes where tidal range is essentially zero and around which the high-water crest sweeps like the hand of a clock. Tidal range then grows with distance from the amphidromic point. This is dynamic theory, and it is why tides at two ports a hundred kilometers apart can differ by hours and meters.

The result is three tidal patterns worth recognizing:

• Semidiurnal: two highs and two lows per day of roughly equal height. Typical of the Atlantic coast of North America and much of Europe.
• Diurnal: one high and one low per day. Found in parts of the Gulf of Mexico and Southeast Asia.
• Mixed semidiurnal: two highs and two lows of noticeably unequal height, driven by lunar declination. Typical of the Pacific coast of North America.

Basin resonance produces the extremes. The Bay of Fundy, in Canada, has a natural sloshing period close to the 12 hour 25 minute tidal period, so each tidal push arrives in step with the previous one, like a child pumping a swing. The result is the largest tidal range on Earth, reaching about 16 meters. Meanwhile the nearly enclosed Mediterranean has a range under 0.3 meters in most places. Where a large tidal range meets a shallowing river mouth, the flood tide can steepen into a breaking wall of water called a tidal bore, as on China's Qiantang River and the Amazon.

Finally, tides drive tidal currents: flood current on the rise, ebb current on the fall, and brief slack water between. In a narrow passage these can exceed 5 meters per second, faster than most boats, and they matter enormously to navigation, to larval transport, and to the small but growing tidal-energy industry at sites like La Rance in France and Sihwa Lake in South Korea.

Key idea: Real tides are forced waves rotating around amphidromic points within basins, which is why patterns and ranges vary from under 0.3 meters in the Mediterranean to 16 meters in the Bay of Fundy.

Common misconceptions

• "A tsunami is a tidal wave." It has no connection to tides. It is generated by displacement of the seafloor or a landslide.
• "A tsunami is a single towering breaking wave." It is usually a series of fast, prolonged surges, and the first is often not the largest.
• "The Sun barely affects tides because it is so far away." It contributes about 46 percent of the Moon's effect, which is exactly what makes spring and neap tides.
• "There is one tidal bulge, on the side facing the Moon." There are two, and the far-side bulge is why most coasts get two high tides a day.
• "Spring tides happen in spring." They happen twice a month, at new and full moon, all year round.

Recap

• Tsunamis are generated by sudden vertical displacement of the whole water column and are shallow-water waves everywhere.
• Speed is the square root of gravity times depth: about 198 meters per second, or 713 kilometers per hour, in 4,000 meters.
• Open-ocean steepness near 1 in 356,000 makes a tsunami undetectable to ships, while shoaling converts length into height.
• Seismic waves outrun tsunamis, which makes distant warning possible; local tsunamis give only the shaking as warning.
• The tide-generating force scales as the inverse cube of distance, so the Sun contributes about 46 percent of the Moon's effect.
• Two bulges and a 24-hour 50-minute lunar day give highs every 12 hours 25 minutes, shifting about 50 minutes daily.
• Spring tides occur at new and full moon; neap tides at the quarters.
• Real tides rotate around amphidromic points, giving semidiurnal, diurnal, and mixed patterns and ranges from 0.3 to 16 meters.

Sources

1. National Ocean Service. (n.d.). Tides and water levels education tutorial. NOAA. oceanservice.noaa.gov
2. National Ocean Service. (n.d.). What is a tsunami? NOAA. oceanservice.noaa.gov
3. U.S. Geological Survey. (n.d.). Tsunamis and tsunami hazards. usgs.gov
4. Encyclopaedia Britannica. (n.d.). Tide. britannica.com
5. NOAA Tides and Currents. (n.d.). Tide predictions and water level data. tidesandcurrents.noaa.gov
6. Wikipedia contributors. (n.d.). Amphidromic point. Wikipedia. en.wikipedia.org

The big picture

Here is a comparison that should stop you. All the phytoplankton in the ocean, added together, amount to well under one percent of the plant matter on Earth. Yet they carry out roughly half of all photosynthesis on the planet, producing about as much organic carbon each year as every forest, grassland, and crop field combined. At least half the oxygen in the breath you just took came from the sea.

How can so little biomass do so much work? Because it turns over almost instantly. A redwood holds its carbon for centuries. A diatom lives a few days, divides, is eaten, and is replaced. The ocean's standing stock is tiny and its throughput is enormous. Once you have that picture, marine biology stops being a catalogue of organisms and becomes a study of flows.

This lesson follows the carbon. It enters at the surface, where light and nutrients allow photosynthesis, passes through a food web with brutal inefficiency, and then some small fraction of it sinks into the deep ocean and stays there. That last process, the biological pump, is the reason the atmosphere has the carbon dioxide concentration it does rather than a much higher one.

Primary production and what limits it

Primary production is the creation of organic matter from inorganic carbon, almost always by photosynthesis. Written in plain text, the reaction is: carbon dioxide plus water, driven by light energy, yields sugar plus oxygen. Gross primary production is the total made; net primary production, or NPP, is what remains after the organism's own respiration, and NPP is what is available to everything else. Global marine NPP runs about 45 to 50 billion tonnes of carbon per year.

Two ingredients limit it, and which one binds depends entirely on where you are.

Light. Seawater absorbs light exponentially, and it absorbs colors unequally. Red light is essentially gone by 15 meters. Blue penetrates furthest, which is why the open ocean is blue and why a diver's cut looks green at depth. The euphotic zone is conventionally the layer receiving at least 1 percent of surface light, reaching perhaps 150 to 200 meters in the clearest open ocean and only a few meters in turbid coastal water.

Worked example. Suppose each 10 meters of water absorbs half the remaining light. After 10 meters you have 50 percent, after 20 meters 25 percent, after 30 meters 12.5 percent. Each additional 10 meters halves it again, so after 70 meters you have 0.5 to the seventh power, which is 0.0078, or about 0.78 percent. The euphotic zone in this water is therefore about 70 meters deep. Notice how fast the decline is: you lose more light in the first 10 meters than in the next 60 combined.

Two related depths matter. The compensation depth is where a phytoplankton cell's photosynthesis exactly equals its own respiration, so it just breaks even. Deeper than that it is losing ground. But cells get mixed up and down by turbulence, so what really matters is the critical depth: the depth above which total photosynthesis in the mixed column equals total respiration in it. Harald Sverdrup formalized this in 1953, and it explains the timing of blooms.

Nutrients. Where light is abundant year-round, in the tropics, the limit is nitrate and phosphate, kept below the permanent thermocline by the stratification of Lesson 6. In roughly a third of the ocean, the true limit is iron.

Key idea: Marine production needs light from above and nutrients from below, and because the two come from opposite directions, productivity peaks only where a physical process brings them together.

Why blooms happen when they do

Put light limitation and nutrient limitation on the same map and the seasonal patterns fall out.

Region	Limiting factor	Seasonal pattern
Polar	Light, almost always	One intense burst during the brief summer of continuous daylight
Temperate	Light in winter, nutrients in summer	A large spring bloom and a smaller autumn bloom
Tropical open ocean	Nutrients, always	Low and nearly constant year-round
Coastal upwelling	Rarely limited	High whenever the upwelling wind blows

The temperate spring bloom is worth walking through, because it ties three modules together. Through winter, storms mix the water column deeply, so nutrients are plentiful at the surface but cells spend most of their time below the critical depth in darkness. In spring the sun strengthens and a seasonal thermocline forms, trapping cells in the well-lit upper layer. Now they have both light and the nutrients winter mixing supplied, and the population explodes. Within weeks they consume the nutrients, the bloom crashes, and grazing zooplankton finish it. In autumn, cooling and storms break down the thermocline, returning some nutrients to a still-lit surface, which produces a smaller second bloom before winter darkness ends it.

Key idea: The temperate spring bloom begins when a seasonal thermocline traps cells above the critical depth in water that winter mixing already stocked with nutrients.

Who the plankton are

Plankton means drifters: organisms that cannot swim effectively against a current, regardless of size. Some jellyfish are plankton; some plankton are bacteria.

Phytoplankton, the photosynthesizers, come in a few important groups. Diatoms build ornate glass boxes of opaline silica, dominate cold nutrient-rich water, and are the workhorses of upwelling zones and spring blooms. Dinoflagellates swim with two flagella, prefer warmer stratified water, and include the species responsible for harmful algal blooms and red tides. Coccolithophores armor themselves with calcite plates and can bloom so densely that satellites see the water turn milky turquoise. And the smallest, discovered only in 1988, is Prochlorococcus, a cyanobacterium about 0.6 micrometers across. There are something on the order of 10 to the 27th of them in the ocean, making it plausibly the most abundant photosynthetic organism on Earth, and nobody knew it existed until forty years ago.

Zooplankton, the drifting animals, are dominated by copepods, millimeter-scale crustaceans that are, by most counts, the most abundant multicellular animals on the planet. Krill, larger shrimp-like crustaceans, anchor the Southern Ocean food web. Zooplankton that spend their whole lives drifting are holoplankton; those that are only temporarily planktonic, such as larval crabs, fish, and sea stars, are meroplankton, and for many bottom-dwelling species this larval phase is the only chance they get to travel.

Every night, a large fraction of the zooplankton and small fishes of the ocean swim upward hundreds of meters to feed in the dark surface waters, then descend again before dawn to hide from visual predators. This diel vertical migration is the largest synchronized movement of biomass on Earth, and it happens every single day. It was discovered accidentally during the Second World War, when sonar operators found a false seafloor at 300 to 500 meters that rose toward the surface at dusk and sank at dawn. They called it the deep scattering layer, and it turned out to be alive.

Key idea: Plankton are defined by drifting rather than size, and their daily vertical migration is the largest recurring movement of animal biomass on the planet.

Trophic efficiency: why big fish are expensive

Energy passes up a food web badly. As a rule of thumb, only about 10 percent of the energy at one trophic level becomes biomass at the next. The rest is spent on respiration, movement, and reproduction, or is simply not eaten.

Worked example. Phytoplankton are level 1, copepods level 2, small fish level 3, and a large predatory tuna level 4. To build 1 kilogram of tuna requires about 10 kilograms of small fish, which required about 100 kilograms of copepods, which required about 1,000 kilograms of phytoplankton. A single kilogram of tuna represents a tonne of plankton production.

Now push it. Some open-ocean predators sit closer to level 5, adding another factor of ten and putting 10 tonnes of phytoplankton behind each kilogram. This is not a curiosity; it is the reason that fisheries in upwelling zones, where the catch is anchovies at level 2 or 3, are enormously more productive per unit area than fisheries targeting large predators, and it is the arithmetic behind Lesson 13 on fishery collapse.

It also explains a geographic pattern. Coastal upwelling food webs are short, often just diatoms to anchoveta, so a large fraction of primary production reaches something people eat. Open-ocean food webs are long, with tiny cells eaten by tiny grazers eaten by slightly larger grazers, so almost nothing survives to a harvestable size.

Key idea: Roughly 10 percent of energy transfers between trophic levels, so a kilogram of top predator embodies a tonne or more of primary production and short food webs are far more productive for fisheries.

The biological pump

Not everything gets eaten. Dead cells, fecal pellets, mucus, and aggregated debris clump together into marine snow, which drifts downward at typically 10 to 100 meters per day. As it falls, bacteria consume most of it, releasing the carbon and nutrients back into the water. That is remineralization, and it is why nutrients are high and oxygen is low at mid-depths, as Lesson 5 described.

But the fraction that survives the fall is enormously important. Carbon that sinks below the permanent thermocline is out of contact with the atmosphere for the roughly thousand-year overturning time from Lesson 8. Carbon that reaches the sediment and gets buried is out of circulation for millions of years. This whole process, moving carbon from the surface to the deep by biology, is the biological pump.

The numbers are worth carrying. Of the roughly 48 billion tonnes of carbon fixed at the surface annually, on the order of 10 billion tonnes sinks past 100 meters as export production. Most of that is remineralized in the upper kilometer. Only about 0.2 billion tonnes, well under one percent of the original production, is finally buried in sediments. The pump is leaky at every stage, and it still matters more than almost anything else.

Here is why. Model studies indicate that if the biological pump stopped entirely, so that surface and deep ocean carbon simply equilibrated, atmospheric carbon dioxide would rise by something like 150 to 200 parts per million. That is comparable to the entire increase humans have caused since the industrial revolution. The pump is not a minor flux. It is a substantial part of the reason the pre-industrial atmosphere held 280 parts per million rather than something far higher.

One refinement keeps the picture honest. Organisms that build calcium carbonate shells, such as coccolithophores and foraminifera, run a partial counter-pump: making carbonate releases carbon dioxide to the surrounding water even as sinking shells carry carbon downward. The net effect of the biological pump is the soft-tissue pump minus this carbonate counter-pump, and getting that balance right is an active research problem.

Key idea: The biological pump exports roughly 10 billion tonnes of carbon below 100 meters each year and buries about 0.2 billion, and without it atmospheric CO2 would be perhaps 150 to 200 parts per million higher.

Common misconceptions

• "Kelp and seaweed produce most marine oxygen." Nearly all marine photosynthesis is done by microscopic drifting phytoplankton, not by visible seaweeds.
• "Clear tropical water is the most productive." Clarity indicates few cells. The most productive water is cold, greenish, and nutrient-rich.
• "Plankton means small." It means unable to swim against currents. Some plankton are meters across.
• "Carbon that sinks is permanently removed." Most is remineralized within a kilometer and returns to the surface within about a thousand years; only a tiny fraction is buried.
• "Adding iron to the ocean is a proven climate fix." Iron fertilization does trigger blooms, but how much carbon actually reaches the deep ocean is uncertain and the ecological side effects are poorly understood.

Recap

• Phytoplankton are under one percent of Earth's plant biomass but perform about half of global photosynthesis.
• Global marine net primary production is about 45 to 50 billion tonnes of carbon per year.
• Light attenuates exponentially; the euphotic zone is the 1 percent light level, up to about 200 meters in the clearest water.
• The compensation depth is where a cell breaks even; the critical depth governs whether a mixed column can bloom.
• Polar production is light-limited, tropical open-ocean production is nutrient-limited, and temperate seas bloom in spring and autumn.
• Diatoms, dinoflagellates, coccolithophores, and Prochlorococcus dominate the phytoplankton; copepods dominate the zooplankton.
• Trophic efficiency near 10 percent means a kilogram of tuna sits on about a tonne of phytoplankton.
• The biological pump exports about 10 billion tonnes of carbon below 100 meters per year and buries about 0.2 billion.

Sources

1. National Ocean Service. (n.d.). How much oxygen comes from the ocean? NOAA. oceanservice.noaa.gov
2. National Ocean Service. (n.d.). What are plankton? NOAA. oceanservice.noaa.gov
3. NASA Earth Observatory. (n.d.). What are phytoplankton? earthobservatory.nasa.gov
4. Woods Hole Oceanographic Institution. (n.d.). Ocean learning hub: the ocean and carbon. whoi.edu
5. Wikipedia contributors. (n.d.). Biological pump. Wikipedia. en.wikipedia.org

The big picture

On 17 February 1977, the submersible Alvin descended 2,500 meters to the Galapagos Rift. The scientists aboard were geologists looking for warm water seeping from young volcanic crust. What their lights found was a dense crowd of animals: white clams the size of dinner plates, blind crabs, and red-tipped tube worms taller than a person, packed around shimmering vents in a place where, according to everything then understood, there should have been almost nothing. The expedition had brought no biologist. They preserved specimens in vodka from the ship's stores because they had not packed enough formalin.

That dive rewrote a rule everyone had believed: that all life on Earth ultimately runs on sunlight. This lesson tours the ocean's living realms, from the sunlit surface to the trench floor, and it ends at those vents, because they are the best possible illustration of how much room the ocean still has to surprise us.

Marine ecologists divide the ocean two ways at once. The pelagic realm is the open water column; the benthic realm is the seafloor and everything living in or on it. Each is subdivided by depth, and depth is the master variable, because it sets light, pressure, temperature, and food supply all at once.

The pelagic zones

Zone	Depth	Conditions	Life
Epipelagic	0 to 200 m	Sunlit, warm, seasonal	All photosynthesis; most familiar fish, whales, seabirds
Mesopelagic	200 to 1,000 m	Twilight; too dim to photosynthesize	Lanternfish, squid, the deep scattering layer
Bathypelagic	1,000 to 4,000 m	Total darkness, near 4 C, high pressure	Anglerfish, gulper eels; sparse, slow, patient
Abyssopelagic	4,000 to 6,000 m	Dark, cold, 400 to 600 atmospheres	Very sparse; specialists
Hadopelagic	Below 6,000 m	Trenches only, up to 1,100 atmospheres	Amphipods, snailfish; recorded past 8,000 m

The mesopelagic, the twilight zone, deserves particular attention because it is where most of the ocean's animal biomass probably lives and because we badly underestimated it. Acoustic surveys published in 2014 revised the global biomass of mesopelagic fishes upward by roughly a factor of ten, to something on the order of 10 billion tonnes. An ecosystem larger than every fishery on Earth combined was sitting a few hundred meters below the surface, mis-measured for decades because the fish were dodging our nets.

The adaptations down there follow from one constraint: it is dark, and food is scarce and unpredictable. Bioluminescence is nearly universal; in some surveys about three quarters of observed deep-sea animals produce their own light, using it to lure prey, startle predators, signal mates, or hide. That last use is the cleverest. In the twilight zone a predator looking upward sees silhouettes against the faint downwelling light, so many mesopelagic animals carry rows of belly photophores tuned to match the light above them, erasing their own shadow. This is counterillumination. Other adaptations include enormous upward-pointing tubular eyes, dark or red pigmentation (red is effectively invisible where no red light exists), enormous expandable jaws and stomachs for the rare large meal, and metabolisms slowed to a crawl. A deep anglerfish may wait months between meals.

Key idea: Depth controls light, pressure, temperature, and food supply at once, and the twilight zone holds most of the ocean's fish biomass hidden behind bioluminescence and counterillumination.

The benthic realms

The seafloor spans the same depth range plus one zone that has no pelagic equivalent: the intertidal, alternately submerged and exposed twice a day.

The intertidal is the most physically brutal habitat in the ocean and the easiest place to see ecology in action. Walk down a rocky shore at low tide and you see sharp horizontal bands of different species. Joseph Connell's classic experiments on Scottish barnacles in 1961 explained the pattern with a rule that has held up broadly: the upper limit of a species is usually set by physical stress, meaning how long it can survive drying and heat, while the lower limit is usually set by biology, meaning competition and predation from species that thrive in the wetter conditions below. Zonation is a compromise between being cooked and being outcompeted.

Below that lie the subtidal shelf, then the bathyal slope, the vast abyssal plain, and the hadal trenches. Benthic animals divide into infauna, living within the sediment, and epifauna, living on top of it, and they feed either by filtering particles from the water as suspension feeders or by processing sediment as deposit feeders. On the abyssal plain, where the only food is the thin rain of marine snow from Lesson 11, deposit feeders dominate and life is sparse, small, and extremely slow-growing. Some deep-sea corals and sponges are thousands of years old, which is exactly why bottom trawling in those habitats causes damage that cannot recover on a human timescale.

Key idea: Intertidal zonation reflects physical stress setting upper limits and competition setting lower ones, while abyssal benthic life is sparse and slow because its only food is marine snow.

Coral reefs: an oasis built on a partnership

Coral reefs occupy well under one percent of the ocean's area and support something like a quarter of all marine species. That density is a paradox, because reefs grow best in warm, clear, nutrient-poor tropical water, which is exactly the water Lesson 6 identified as a biological desert.

The resolution is a partnership. Reef-building corals are animals, close relatives of jellyfish, that host single-celled photosynthetic algae called zooxanthellae inside their own tissues. The algae get shelter, carbon dioxide, and nutrients from the coral's waste. The coral gets sugars, in some cases up to about 90 percent of its energy budget. This symbiosis lets a reef recycle nutrients internally instead of depending on the surrounding water, so it can build a dense, productive ecosystem in a nutrient desert. A reef is an oasis precisely because it does not need the water around it.

Because the algae need light, reef-building corals need shallow, clear water, generally within about 50 meters of the surface. They also need warmth, roughly 23 to 29 degrees, normal salinity, and hard substrate. Darwin's classification still holds: fringing reefs grow directly against a shore, barrier reefs are separated from land by a lagoon, and atolls are rings enclosing a lagoon where the island has subsided away, exactly as Lesson 2 described.

Now the vulnerability. The partnership is temperature-sensitive. When water stays roughly 1 degree above the normal summer maximum for several weeks, the symbiosis breaks down and the coral expels its algae. The white calcium carbonate skeleton shows through the now-transparent tissue, which is coral bleaching. A bleached coral is not dead; it is starving, and if temperatures fall soon enough it can take up new algae and recover. If the heat persists, it dies.

The metric used operationally is Degree Heating Weeks, which accumulates how far above the local summer maximum the water has been and for how long. Around 4 degree heating weeks, significant bleaching is expected; around 8, widespread bleaching with mortality. NOAA's Coral Reef Watch program forecasts this from satellites, and the record since the 1980s shows global-scale bleaching events recurring at shortening intervals, with severe events affecting the Great Barrier Reef repeatedly since 2016.

Key idea: Reefs thrive in nutrient-poor water because corals host photosynthetic zooxanthellae internally, and that same partnership collapses into bleaching when water stays about a degree above the summer maximum for weeks.

Chemosynthesis: life without the Sun

At a mid-ocean ridge, seawater seeps into cracks in hot young crust, is heated to several hundred degrees, leaches metals and sulfide from the rock, and jets back out. At black smokers, fluid emerges at up to about 400 degrees Celsius, kept liquid rather than boiling by the immense pressure, and precipitates dark metal sulfide particles on contact with near-freezing seawater, building chimneys.

The energy source is chemical. Bacteria and archaea oxidize the hydrogen sulfide in the vent fluid, using the released energy to fix carbon dioxide into sugars. In plain text: carbon dioxide plus oxygen plus hydrogen sulfide yields carbohydrate plus sulfur plus water. This is chemosynthesis, and structurally it is photosynthesis with a chemical bond substituted for a photon. The organisms doing it are primary producers in exactly the same sense a diatom is.

Around them the community is built. The signature animal is the giant tube worm, Riftia pachyptila, which grows to two meters and is among the fastest-growing marine invertebrates known. It has no mouth, no gut, and no anus. Instead it houses symbiotic chemosynthetic bacteria in a specialized organ called the trophosome, and its red plume, colored by hemoglobin, absorbs hydrogen sulfide and oxygen from the water and delivers both to its bacteria. The worm is essentially a life-support system for its microbes.

Vent communities are ephemeral. A vent may run for years or decades and then shut off, killing everything that depends on it, so vent animals disperse as larvae across enormous distances to find the next one. Related chemosynthetic communities live at cold seeps, where methane and sulfide leak from sediments without the heat, and at whale falls, where a single carcass on the abyssal floor supports a succession of communities for decades, eventually including sulfide-dependent species living off the decaying bones.

The implications reach beyond oceanography. If life can run on chemical energy from rock and water, then the requirements for life elsewhere are far less restrictive than sunlight, and the subsurface oceans of Europa and Enceladus become plausible places to look. Vents are also a serious candidate for where life on Earth began.

Key idea: Chemosynthetic microbes oxidize hydrogen sulfide to fix carbon, supporting entire vent, seep, and whale-fall ecosystems that never use sunlight at all.

Common misconceptions

• "Coral is a plant, or a rock." Coral is an animal that hosts photosynthetic algae inside its tissues and builds a carbonate skeleton.
• "A bleached coral is dead." It has expelled its algae and is starving. Recovery is possible if the heat stress ends soon enough.
• "Nothing lives in the deep sea." The mesopelagic alone may hold on the order of 10 billion tonnes of fish, more than all fisheries combined.
• "Vent animals eat the hot fluid." They depend on microbes that oxidize hydrogen sulfide; the giant tube worm has no gut at all.
• "Deep-sea animals are crushed by pressure if they have no armor." Pressure is not a problem for bodies without gas spaces, since liquids and tissues barely compress.

Recap

• The pelagic realm runs from the epipelagic to the hadopelagic; the benthic realm adds the intertidal.
• The mesopelagic twilight zone may hold about 10 billion tonnes of fish, roughly ten times earlier estimates.
• Bioluminescence is nearly universal in the deep, and counterillumination erases silhouettes seen from below.
• Intertidal upper limits are set by physical stress and lower limits by competition and predation.
• Reefs cover under one percent of the ocean but support about a quarter of marine species.
• Zooxanthellae supply corals with up to about 90 percent of their energy, allowing reefs to thrive in nutrient-poor water.
• Bleaching occurs when water stays roughly 1 degree above the summer maximum for weeks, tracked as Degree Heating Weeks.
• Chemosynthesis oxidizes hydrogen sulfide to fix carbon, powering vent, seep, and whale-fall communities without sunlight.

Sources

1. National Ocean Service. (n.d.). Corals education tutorial. NOAA. oceanservice.noaa.gov
2. National Ocean Service. (n.d.). What is coral bleaching? NOAA. oceanservice.noaa.gov
3. National Ocean Service. (n.d.). What is a hydrothermal vent? NOAA. oceanservice.noaa.gov
4. Woods Hole Oceanographic Institution. (n.d.). Dive and Discover: hydrothermal vents. divediscover.whoi.edu
5. NOAA Coral Reef Watch. (n.d.). Satellite coral bleaching heat stress monitoring. coralreefwatch.noaa.gov
6. Encyclopaedia Britannica. (n.d.). Coral reef. britannica.com

The big picture

On 2 July 1992 the Canadian fisheries minister stood in a hotel in St. John's, Newfoundland, and announced an indefinite moratorium on fishing northern cod. Outside, fishermen tried to break down the door. The Grand Banks had supported cod fishing for roughly five hundred years, since John Cabot's crew reported in 1497 that they could lower baskets over the side and haul them up full. In a single generation of industrial trawling, the spawning stock fell to roughly one percent of its historical level. About 30,000 people lost their work in one announcement. More than thirty years later, the stock has still not recovered to anything like its former abundance.

This lesson is about how that happens, and it is the most quantitative lesson in the biology module, because the collapse of a fishery is at bottom an arithmetic failure. A population that is harvested faster than it regrows will decline, and the rate at which it regrows is itself a function of its size. Get that function wrong, or get the stock estimate wrong, or refuse to act on the estimate for political reasons, and the decline accelerates rather than levels off.

The stakes are large. Global wild capture has been roughly flat near 90 million tonnes per year since the late 1980s, despite continuing increases in fishing effort and technology. Aquaculture has grown to overtake wild capture for aquatic animals. Fish supply a substantial share of animal protein worldwide, and hundreds of millions of people depend on fisheries and aquaculture for their livelihoods.

The surplus-production model

Start with the simplest useful model of a fish population. Let N be the stock biomass, r the intrinsic growth rate, and K the carrying capacity, the biomass the environment supports with no fishing. The logistic growth model says:

The growth rate equals r times N times the quantity 1 minus N divided by K.

Read the two factors. The r times N part says a bigger population grows faster, because there are more breeders. The 1 minus N over K part says growth slows as the population approaches its ceiling, because food, space, and predation bite. Multiply them and the growth curve is a hump: near zero when N is tiny, near zero again when N is at K, and maximum somewhere in between.

Where in between? At exactly half of carrying capacity. That maximum growth is the maximum sustainable yield, or MSY, and it equals r times K divided by 4.

Worked example. Take a stock with a carrying capacity of 1,000,000 tonnes and an intrinsic growth rate of 0.4 per year. MSY equals 0.4 times 1,000,000 divided by 4, which is 100,000 tonnes per year, achieved when the stock is held at 500,000 tonnes. In principle you can take 100,000 tonnes every year forever.

Notice what MSY implies and what people often get wrong about it. A stock fished at MSY is deliberately held at half its unfished size. That is not a damaged population; it is a managed one, and it produces more harvestable surplus than an untouched one would. Fishing is not inherently destructive. The problem is what happens on the other side of the hump.

Key idea: Logistic growth peaks at half of carrying capacity, so maximum sustainable yield equals r times K over 4 and is achieved by deliberately holding the stock at K/2.

Why MSY is a knife edge

Now work the failure. Keep the same stock and suppose managers, under political pressure and with an optimistic stock assessment, set the quota at 120,000 tonnes instead of 100,000.

At 500,000 tonnes of biomass, the surplus is 100,000. You take 120,000. The stock falls by 20,000, to 480,000.

Next year, compute the surplus at 480,000: 0.4 times 480,000 times (1 minus 0.48), which is 0.4 times 480,000 times 0.52, or about 99,800. You take 120,000 again. Now the deficit is about 20,200 and the stock falls further.

Keep going. At 400,000 tonnes the surplus is 0.4 times 400,000 times 0.6, which is 96,000, so the annual deficit has grown to 24,000. At 300,000 tonnes the surplus is 0.4 times 300,000 times 0.7, which is 84,000, and the deficit is 36,000. At 200,000 tonnes the surplus is 0.4 times 200,000 times 0.8, which is 64,000, and the deficit is 56,000.

Look at what just happened. The deficit did not stay constant at 20,000. It grew every single year, because as the stock falls below K/2 its ability to replace itself falls too. A modest, fixed overharvest produces an accelerating collapse. That is the central and deeply counterintuitive result of this lesson, and it explains why fisheries do not decline gently and then stabilize. They look fine, then they look slightly worse, then they are gone.

Three real-world features make it worse than the model.

• Catch stays high while the stock falls. Better sonar, bigger nets, and targeting of remaining spawning aggregations mean landings can hold steady even as abundance crashes, so the fishery feels healthy right up until it is not.
• Assessments are uncertain. You cannot count fish. Stock size is estimated from catch rates and surveys with wide error bars, and the historical tendency has been to err optimistically.
• Depensation. At very low densities, reproduction can fail disproportionately because spawners cannot find each other or because a different species takes over the habitat. Below some threshold, the population may not rebound even if fishing stops entirely, which is one explanation for why northern cod remains depressed decades after the moratorium.

Key idea: A constant overharvest produces an accelerating decline because surplus production falls as the stock falls, and uncertainty, technology, and depensation all push the real world further toward collapse than the model alone predicts.

Two collapses

Northern cod, Newfoundland. Foreign factory trawlers arrived on the Grand Banks in the 1950s and 1960s, and catches spiked. Canada extended its exclusive economic zone to 200 nautical miles in 1977 and then expanded its own trawler fleet, expecting to harvest sustainably. Scientific assessments through the 1980s suggested trouble; quotas were repeatedly set above what the science advised because coastal communities had nowhere else to go. By 1992 the spawning biomass had fallen to roughly one percent of historical levels and the moratorium followed. The lesson is not that anyone failed to understand the model. It is that the model gave uncomfortable answers and the institutional pressure to discount them was overwhelming.

Peruvian anchoveta. This was, at its peak around 1970, the largest single-species fishery ever prosecuted, landing over 12 million tonnes in a year, roughly a fifth of the entire world catch. It depended completely on the upwelling of Lesson 8. In 1972 a strong El Nino suppressed that upwelling: the thermocline deepened, warm nutrient-poor water replaced cold nutrient-rich water, production crashed, and the anchoveta both died and dispersed. Landings fell to about 2 million tonnes. A stock already fished at very high intensity had no buffer against a natural fluctuation it had survived many times before. The lesson here is different and equally important: heavy fishing does not just reduce a stock, it removes the resilience that lets a stock absorb environmental variability.

Key idea: Cod collapsed because quotas were set above scientific advice for social reasons; anchoveta collapsed because heavy fishing removed the buffer that would have carried it through an El Nino.

The wider picture

The FAO's periodic assessment finds that the fraction of assessed marine stocks fished within biologically sustainable levels has fallen steadily, from about 90 percent in the mid-1970s to roughly 62 percent in the most recent assessment, meaning something over a third are now fished unsustainably. Several patterns run alongside that headline.

Fishing down the food web. As large predators are depleted, fleets shift to smaller, lower-trophic-level species. The average trophic level of global landings has declined. This can look like stable total tonnage while the ecosystem underneath is being fundamentally altered.

Shifting baselines. Each generation of scientists and fishers takes the abundance they encountered early in their careers as normal, so the accumulated loss is invisible. Colonial accounts of the Caribbean describe sea turtles so numerous that ships navigated by their noise. Nobody alive has a memory that would flag today's numbers as catastrophic.

Bycatch and habitat damage. Non-target species, including turtles, sharks, and seabirds, are caught and usually discarded dead. Bottom trawling drags heavy gear across the seafloor, flattening structural habitat that in deep-water settings may take centuries to millennia to recover, as Lesson 12 noted. Lost or abandoned gear continues to fish as ghost gear.

Illegal, unreported, and unregulated fishing undermines every quota by making the actual removals unknown.

Key idea: Beyond single-stock overfishing, fisheries reshape ecosystems through trophic downgrading, bycatch, habitat destruction, and unreported catch, while shifting baselines hide the cumulative loss.

What actually works

This is not a hopeless subject, and it would be dishonest to teach it as one. Stocks do rebuild when four things are in place.

Catch limits that follow the science. In the United States, the Magnuson-Stevens Act was amended to require annual catch limits with accountability measures, and dozens of stocks previously declared overfished have been rebuilt since 2000. The mechanism is not clever; it is simply setting the limit at or below the scientific advice and enforcing it.

Aligned incentives. Where fishers hold a secure long-term share of the catch, rather than racing everyone else to grab fish before they are gone, the incentive shifts from maximizing today's haul to maximizing the stock's future value. Well-designed rights-based systems have reduced the race to fish, though they raise real questions about equity and consolidation.

Gear and area measures. Turtle excluder devices, circle hooks, bird-scaring lines, and seasonal closures cut bycatch substantially at modest cost. Protected areas, covered in Lesson 16, provide refuges where fish grow large and fecund.

Aquaculture, with caveats. Farmed production now exceeds wild capture for aquatic animals, and farming shellfish and seaweed can be genuinely low-impact or even beneficial. But farming carnivorous fish such as salmon requires wild-caught fishmeal, so it can transfer pressure rather than relieve it, and it brings problems of disease, escapes, effluent, and mangrove clearance. Aquaculture is part of the answer and not a substitute for managing wild stocks.

Key idea: Stocks rebuild when catch limits follow scientific advice, incentives reward long horizons, gear reduces bycatch, and aquaculture supplements rather than replaces well-managed wild fisheries.

Common misconceptions

• "A sustainably fished stock is one left at its natural size." Maximum sustainable yield is obtained at about half of carrying capacity, where surplus production is greatest.
• "Steady landings mean a healthy stock." Improving technology can hold catch steady while abundance collapses, which is exactly what happened to cod.
• "If we stop fishing, the stock always comes back." Depensation and ecosystem shifts can prevent recovery; northern cod has stayed depressed for decades.
• "Aquaculture removes pressure from wild stocks." Farming carnivorous species consumes wild-caught fishmeal and can shift pressure onto forage fish.
• "The anchoveta collapse was purely natural, caused by El Nino." El Nino was the trigger, but heavy fishing had already removed the buffer that let the stock survive earlier El Ninos.

Recap

• Logistic growth peaks at half of carrying capacity, giving MSY equal to r times K divided by 4.
• For K of 1,000,000 tonnes and r of 0.4, MSY is 100,000 tonnes per year at a stock of 500,000 tonnes.
• Harvesting 120,000 tonnes from that stock produces a deficit that grows each year, from 20,000 to 56,000 by the time biomass halves again.
• Uncertain assessments, improving technology, and depensation make real collapses faster than the model predicts.
• Northern cod fell to about one percent of historical biomass and was closed in 1992; it has not recovered.
• Peruvian anchoveta peaked above 12 million tonnes and collapsed in 1972 when El Nino met an over-fished stock.
• Roughly a third of assessed stocks are now fished unsustainably, alongside bycatch, habitat damage, and unreported catch.
• Rebuilding works where catch limits follow science, incentives are long-term, and gear reduces bycatch.

Sources

1. Food and Agriculture Organization of the United Nations. (n.d.). The state of world fisheries and aquaculture (SOFIA). fao.org
2. NOAA Fisheries. (n.d.). Sustainable fisheries and stock status. fisheries.noaa.gov
3. Encyclopaedia Britannica. (n.d.). Commercial fishing. britannica.com
4. Wikipedia contributors. (n.d.). Collapse of the Atlantic northwest cod fishery. Wikipedia. en.wikipedia.org
5. Wikipedia contributors. (n.d.). Maximum sustainable yield. Wikipedia. en.wikipedia.org

The big picture

Here is a number that reframes the entire climate problem. Of all the extra energy trapped by added greenhouse gases since 1971, roughly 91 percent has gone into the ocean. The atmosphere, which is what we actually live in and measure obsessively, has taken about one percent. The land and the melting of ice take most of the rest.

Global warming is, in a very literal accounting sense, ocean warming. And that is not merely a bookkeeping curiosity. It means the ocean sets the pace: how fast the surface warms, how much warming is already committed regardless of future emissions, how sea level responds, and how the year-to-year climate everyone experiences swings between wet and dry. This lesson works through three consequences: heat, El Nino, and sea level.

Why water dominates: the heat capacity calculation

Water has an unusually large specific heat capacity, about 4,184 joules per kilogram per degree for fresh water and near 3,990 for seawater. Air is about 1,005. So per kilogram, water stores roughly four times as much heat per degree. But the more important factor is mass, and here the comparison becomes lopsided.

Worked comparison. The atmosphere has a mass of about 5.15 times 10 to the 18th kilograms. Its total heat capacity is that mass times 1,005 joules per kilogram per degree, which is about 5.2 times 10 to the 21st joules per degree.

Now ask how deep a layer of ocean has the same heat capacity. Ocean area is about 3.6 times 10 to the 14th square meters. A layer of depth d has mass 3.6e14 times d times 1,027 kilograms, and heat capacity that mass times 3,990, which works out to about 1.47 times 10 to the 21st times d joules per degree. Set that equal to the atmosphere's 5.2 times 10 to the 21st and solve: d is about 3.5 meters.

The top three and a half meters of the ocean hold as much heat as the entire atmosphere from the sea surface to space. Different textbooks quote 2.5 to 3.5 meters depending on the constants used, but the conclusion does not depend on the details. The atmosphere is, thermally, a thin skin on top of an enormous reservoir.

Two consequences follow immediately. First, the ocean buffers climate: coastal places have mild winters and cool summers because the water beside them changes temperature reluctantly, while continental interiors swing wildly. Second, the ocean has thermal inertia. It absorbs heat slowly and it will release it slowly, so even if emissions stopped today the ocean would continue warming the atmosphere toward equilibrium for centuries. Some warming is already in the bank.

A scale check. Earth's current energy imbalance is roughly 0.9 watts per square meter over 5.1 times 10 to the 14th square meters of surface, which is about 4.6 times 10 to the 14th joules every second. Divide by the roughly 6.3 times 10 to the 13th joules released by the Hiroshima bomb and you get about 7 such energy releases per second, continuously, more than 90 percent of which lands in the ocean. That is the scale of what the water is absorbing.

Key idea: The top three and a half meters of ocean match the whole atmosphere's heat capacity, so the ocean has absorbed about 91 percent of excess heat and its thermal inertia commits us to further warming.

El Nino and La Nina: the tropical Pacific swings

Now to the largest year-to-year climate signal on the planet, and one you can derive from what you already know.

The normal state. Trade winds blow east to west across the tropical Pacific. They drag surface water westward, piling warm water in the western Pacific until sea level near Indonesia stands roughly half a meter higher than near Peru. Because the warm layer has been pushed west, the thermocline is deep in the west, perhaps 150 to 200 meters, and shallow in the east, perhaps 40 to 50 meters. With the thermocline that shallow off South America, coastal upwelling taps cold, nutrient-rich water, and the anchoveta fishery of Lesson 13 flourishes. Warm water in the west feeds rising air, convection, and heavy rain over Indonesia and northern Australia, while cool water in the east sits under descending air and coastal desert. This loop of rising air in the west and sinking air in the east is the Walker circulation.

El Nino. Every two to seven years, the trades weaken or even reverse in the western Pacific. The mound of warm water is no longer held up, and it slides eastward across the basin. The thermocline flattens: it shoals in the west and deepens in the east. Now the upwelling off Peru is still occurring, but it is drawing up warm, nutrient-poor water from above the deepened thermocline, so productivity collapses and so does the fishery. The convection that used to sit over Indonesia moves east into the central Pacific, and with it the rainfall. Indonesia and Australia dry out and burn; coastal Peru and Ecuador flood. The name comes from Peruvian fishermen, who noticed the warm water tended to appear around Christmas and called it El Nino, the boy child.

La Nina. The opposite phase: unusually strong trades, an unusually steep thermocline, unusually cold water in the east, and an intensified version of the normal pattern.

This is a genuinely coupled oscillation. The winds move the water and the water's temperature pattern controls the winds, so neither is simply the cause. That coupling is what makes it oscillate rather than settle.

The effects are global, propagated through the atmosphere as teleconnections: shifted jet streams, altered North American winter storm tracks, suppressed Atlantic hurricane activity during El Nino and enhanced activity during La Nina, drought in southern Africa and southeast Asia, and reliable spikes in global mean surface temperature during strong El Nino years, as in 1997 to 1998 and 2015 to 2016. When someone points to an unusually warm year, checking the ENSO phase is the first thing a careful analyst does.

Key idea: El Nino is a coupled ocean-atmosphere oscillation in which weakened trades let warm water slide east, flattening the thermocline, killing Peruvian upwelling, and shifting rainfall and storms worldwide.

Sea level rise, component by component

Global mean sea level has risen roughly 21 to 24 centimeters since 1880, and the rate is accelerating. Tide gauges give about 1.4 millimeters per year for the early twentieth century. Satellite altimetry since 1993 gives an average near 3.4 millimeters per year, and recent years exceed 4. That acceleration is itself one of the clearest signals in the climate record.

The rise has four physical sources, and separating them is essential because they behave differently.

Component	Approximate share of recent rise	Mechanism
Thermal expansion	Roughly a third to 40 percent	Warming seawater occupies more volume
Mountain glaciers	About a fifth	Land ice melting and running to the sea
Greenland ice sheet	About a fifth	Surface melt plus faster outlet glaciers
Antarctic ice sheet	Roughly a tenth, and growing	Ice shelf thinning by warm ocean water beneath

Worked thermal expansion. The thermal expansion coefficient of seawater is about 2 times 10 to the minus 4 per degree in the warm upper ocean. Suppose a 1,000-meter column warms by 0.1 degrees. The height increase is the coefficient times the column depth times the temperature change: 2e-4 times 1,000 times 0.1, which is 0.02 meters, or 2 centimeters. A tenth of a degree, spread over a kilometer of water, raises sea level two centimeters without a single ice cube melting. Now notice the leverage: the same warming applied to a 2,000-meter column gives 4 centimeters. Depth of penetration matters as much as the temperature change, which is why the stratification changes from Lesson 6 feed directly into sea level projections.

The sea ice question. Melting sea ice does not raise sea level. Floating ice already displaces its own weight of water, so when it melts, the meltwater exactly fills the volume the ice displaced. You can verify this in a glass: a floating ice cube melts and the level does not change. Sea ice loss matters enormously for albedo, ecosystems, and circulation, but not for sea level. Land ice is entirely different, because it adds water that was not previously in the ocean at all. If Greenland melted completely it would add about 7.4 meters; the Antarctic ice sheet holds about 58 meters.

Why your coast is not average. Global mean sea level rise is an average, and local rise can differ substantially. Land subsidence, from groundwater extraction or from ongoing rebound after the last ice age, can double or reverse the local signal. Ocean circulation redistributes water, so a weakening AMOC raises sea level along the northeastern coast of North America. And a melting ice sheet's own gravitational pull relaxes, so sea level actually falls near a shrinking ice sheet and rises more than average far away, a pattern called a sea-level fingerprint.

The immediate impact is not usually a dramatic inundation. It is that the baseline has moved, so the same storm surge or high tide now reaches further. Coastal nuisance flooding on sunny high-tide days has become several times more frequent at many United States locations over the past few decades, which is what a rising baseline looks like in practice.

Key idea: Sea level rise combines thermal expansion with melting land ice, is accelerating past 4 millimeters per year, varies regionally because of subsidence and gravity, and matters mainly by raising the baseline under every storm and tide.

Marine heatwaves

One more consequence deserves naming. A marine heatwave is a period of at least five days when sea surface temperature exceeds the local 90th percentile for the time of year. They have become substantially more frequent and longer as the surface has warmed and stratification has strengthened, concentrating added heat in a thinner layer.

The 2013 to 2016 northeast Pacific event nicknamed the Blob is the best-studied case. It reorganized the ecosystem: harmful algal blooms closed crab fisheries, seabirds died in enormous numbers, and species distributions shifted hundreds of kilometers. Marine heatwaves are also the direct driver of the coral bleaching from Lesson 12.

Key idea: Marine heatwaves, defined as five or more days above the local 90th percentile, are lengthening and intensifying, and they drive bleaching, fishery closures, and rapid ecosystem reorganization.

Common misconceptions

• "Global warming is mainly atmospheric warming." About 91 percent of the excess energy has gone into the ocean; the atmosphere holds roughly 1 percent.
• "Melting sea ice raises sea level." Floating ice already displaces its own weight, so its melting adds essentially no volume. Land ice is the problem.
• "El Nino is caused by warm water moving east, or by the winds weakening." Neither alone. It is a coupled oscillation in which each drives the other.
• "Sea level rises equally everywhere." Local rates vary widely because of subsidence, ocean circulation, and the gravitational fingerprints of melting ice.
• "If emissions stopped, the ocean would stop warming immediately." Thermal inertia means the ocean would keep taking up heat and adjusting for centuries.

Recap

• The top about 3.5 meters of ocean has the same heat capacity as the entire atmosphere.
• Roughly 91 percent of excess heat since 1971 has entered the ocean; the atmosphere holds about 1 percent.
• Earth's energy imbalance of about 0.9 watts per square meter is roughly 4.6e14 joules per second.
• In the normal Pacific state, trades pile warm water west, deepening the thermocline there and allowing Peruvian upwelling.
• In El Nino, weakened trades let warm water slide east, flattening the thermocline, ending productive upwelling, and shifting global rainfall.
• Sea level has risen 21 to 24 centimeters since 1880 and now exceeds 4 millimeters per year.
• Thermal expansion of a 1,000-meter column warmed by 0.1 degrees raises sea level about 2 centimeters.
• Melting sea ice does not raise sea level; Greenland holds about 7.4 meters and Antarctica about 58.

Sources

1. NASA Global Climate Change. (n.d.). Vital signs: sea level. climate.nasa.gov
2. National Ocean Service. (n.d.). Is sea level rising? NOAA. oceanservice.noaa.gov
3. Pacific Marine Environmental Laboratory. (n.d.). What is El Nino? NOAA. pmel.noaa.gov
4. Fox-Kemper, B., Hewitt, H. T., Xiao, C., et al. (2021). Chapter 9: Ocean, cryosphere and sea level change. In Climate change 2021: The physical science basis (IPCC AR6 WGI). ipcc.ch
5. National Centers for Environmental Information. (n.d.). Global ocean heat content. NOAA. ncei.noaa.gov
6. Encyclopaedia Britannica. (n.d.). El Nino. britannica.com

The big picture

In 2007 and again in 2008, the Whiskey Creek Shellfish Hatchery on Netarts Bay in Oregon began losing nearly all of its oyster larvae. The animals would spawn, begin to build their first tiny shells, and then simply fail. The owners checked for disease, for bacteria, for equipment faults, and found nothing. Working with oceanographers, they eventually established the cause: on days when the wind drove coastal upwelling, the water they were pumping into the hatchery came from a few hundred meters down, and its chemistry had become corrosive enough to prevent larval shell formation.

That water was not polluted in any conventional sense. It was ordinary deep Pacific water, already rich in respired carbon dioxide from Lesson 5, brought to the surface by the ordinary upwelling of Lesson 8, and pushed across a chemical threshold by the extra carbon dioxide humans have added to the atmosphere. The hatchery now monitors carbonate chemistry continuously and buffers its intake water. It is generally regarded as the first commercial operation clearly damaged by ocean acidification, and it happened decades earlier than most people expected.

This lesson works the chemistry properly, because acidification is a subject where the arithmetic does most of the explaining and where the loose verbal version misleads badly. It is sometimes called the other carbon dioxide problem: same cause as global warming, entirely different mechanism, and not solved by anything that only addresses temperature.

The ocean sink, and its price

The ocean has absorbed roughly a quarter to a third of all the carbon dioxide humans have emitted. Over the recent decade the ocean sink has taken up on the order of 26 percent of total anthropogenic emissions each year. That is an enormous service. Without it, atmospheric carbon dioxide would already be far higher than the roughly 420 parts per million we observe, and warming would be correspondingly worse.

But the carbon does not vanish when it enters the water. It reacts. Acidification is simply the invoice for the sink, and unlike warming it is essentially unavoidable chemistry: you cannot dissolve carbon dioxide into seawater without changing seawater's chemistry.

Key idea: The ocean absorbs about a quarter of human carbon emissions, which slows warming, and acidification is the direct chemical consequence of that absorption.

The reaction chain, step by step

Recall from Lesson 5 that carbon dioxide entering seawater does not simply dissolve; it reacts. Written in plain text, the sequence is:

1. Gaseous CO2 dissolves to become aqueous CO2.
2. CO2 plus H2O gives H2CO3, carbonic acid.
3. H2CO3 gives up a proton: H2CO3 becomes H+ plus HCO3-, bicarbonate.
4. Bicarbonate can give up another: HCO3- becomes H+ plus CO3 2-, carbonate.

Step 3 is where the acidity comes from. Every molecule of carbon dioxide that dissolves and reacts releases hydrogen ions, and more hydrogen ions means lower pH by definition.

Now the step most explanations skip, and it is the one that actually matters for shells. Those newly released hydrogen ions do not just sit there. They react with the carbonate ions already present, running step 4 backwards:

H+ plus CO3 2- gives HCO3-.

Combine that with step 2 and you get the single net reaction that summarizes ocean acidification:

CO2 plus H2O plus CO3 2- gives 2 HCO3-.

Read that carefully, because it contains the whole problem. Adding carbon dioxide to seawater consumes carbonate ions. It does not merely make the water more acidic; it strips out the very building block that corals, oysters, pteropods, and coccolithophores need to make calcium carbonate. Total dissolved inorganic carbon goes up while the carbonate fraction of it goes down. That is why a change of one tenth of a pH unit, which sounds trivial, is a serious biological event.

Key idea: Added CO2 releases hydrogen ions, and those ions convert carbonate to bicarbonate, so acidification simultaneously lowers pH and removes the carbonate that calcifying organisms need.

The pH arithmetic, worked

pH is defined as the negative base-ten logarithm of the hydrogen ion concentration in moles per liter. Because it is logarithmic, small-looking pH changes are large concentration changes, and this trips people up constantly. Let us do it explicitly.

Pre-industrial surface ocean pH was about 8.2. Today it is about 8.1.

Step 1. Hydrogen ion concentration at pH 8.2 is 10 to the minus 8.2, which is 6.31 times 10 to the minus 9 moles per liter.

Step 2. At pH 8.1 it is 10 to the minus 8.1, which is 7.94 times 10 to the minus 9 moles per liter.

Step 3. The ratio is 7.94 divided by 6.31, which is 1.26.

So a drop of 0.1 pH units means the hydrogen ion concentration has risen by 26 percent. Not 1 percent. Not 0.1 percent. Twenty-six percent, in about two centuries.

There is a shortcut worth learning. Since pH is a base-ten log, the factor change in hydrogen ions is simply 10 raised to the size of the pH drop. For a drop of 0.1, that is 10 to the 0.1, which is 1.26, a 26 percent increase. For a drop of 0.3, it is 10 to the 0.3, which is 2.0, a doubling. For a drop of 1.0, it is a tenfold increase.

Projection. Under a high-emissions pathway, surface pH is projected to fall by roughly a further 0.3 units by 2100, reaching about 7.8. From the pre-industrial 8.2 that is a total drop of 0.4, and 10 to the 0.4 is 2.5. Hydrogen ion concentration would be about 150 percent higher than pre-industrial, meaning two and a half times as concentrated.

One more piece of context that matters more than any of these numbers: the rate. Geological records show the ocean has been more acidic in the deep past, but the transitions took thousands to millions of years, giving organisms and the carbonate buffer time to adjust through weathering and sediment dissolution. The current change is happening over roughly two centuries, and by most estimates it is faster than anything in at least tens of millions of years. During the Paleocene-Eocene Thermal Maximum about 56 million years ago, often used as the closest analogue, the carbon release took on the order of thousands of years. Ours is taking hundreds.

Key idea: A 0.1 unit pH decline is a 26 percent rise in hydrogen ions, because the factor change equals 10 raised to the pH drop, and the current rate of change likely exceeds anything in tens of millions of years.

Saturation state: the number biologists actually use

Organisms building shells do not care about pH directly. They care about whether the water around them favors making calcium carbonate or dissolving it, and that is measured by the saturation state, written as the Greek letter omega.

Omega is the product of the calcium ion and carbonate ion concentrations divided by the solubility product for the mineral in question. Interpret it simply:

• Omega greater than 1: the water is supersaturated, and calcium carbonate is stable and can be built.
• Omega equal to 1: exact equilibrium, the saturation horizon.
• Omega less than 1: the water is undersaturated and corrosive, and unprotected calcium carbonate dissolves.

Because calcium concentration barely varies, omega is essentially proportional to the carbonate ion concentration, which is exactly what the net acidification reaction consumes. Adding carbon dioxide lowers omega directly.

Crucially, calcium carbonate comes in two mineral forms with different solubilities. Calcite is the less soluble form, used by coccolithophores and most foraminifera. Aragonite is about 50 percent more soluble, and it is what corals and pteropods and many mollusc larvae build with. So the aragonite saturation horizon, the depth where omega for aragonite equals 1, always sits shallower than the calcite horizon, and aragonite builders are in trouble first.

Tropical surface water today has an aragonite omega near 3 to 4, down from something closer to 4.5 before industrialization; reef calcification declines measurably as omega falls even while it remains above 1. As carbonate is consumed, the saturation horizon rises through the water column, in some regions by 100 to 300 meters already, which means water that was chemically hospitable is becoming corrosive without anything else about it changing.

Key idea: Saturation state omega, essentially proportional to carbonate ion concentration, determines whether shell material is stable; aragonite dissolves more readily than calcite, so corals and pteropods are affected first.

Who is affected, and where first

Three regions cross thresholds earliest, and each for a reason you can now derive.

Polar oceans. Cold water dissolves more gas, so it takes up more carbon dioxide, and cold water naturally holds less carbonate. Southern Ocean and Arctic surface waters are already close to aragonite saturation and are projected to become seasonally undersaturated within decades.

Eastern boundary upwelling systems. The water rising off California, Peru, and Namibia has been away from the surface for decades to centuries, accumulating respired carbon dioxide the whole time. It is already low in pH before any human carbon is added. The anthropogenic increment simply pushes it across the line, which is precisely what happened at Whiskey Creek.

Depth generally. Deep water is colder and richer in respired CO2, so omega falls with depth everywhere, and the horizon shoals as acidification proceeds.

The biological picture is genuinely mixed, and honesty requires saying so. Larval and juvenile stages of oysters, clams, and sea urchins are consistently among the most vulnerable, because they must build a first shell fast from a small energy budget. Pteropods, the swimming snails that are a keystone food in polar food webs, show visible shell dissolution in undersaturated water. Corals calcify more slowly. But some seagrasses and fleshy algae grow faster with more dissolved carbon dioxide, some coccolithophore strains adapt, and some fish are unaffected at realistic levels while others show behavioral changes that remain debated. The overall expectation is not that everything dies but that the winners and losers change, and ecosystems reorganize around organisms that happen to tolerate the new chemistry.

It also is not free to build a shell in low-omega water. Many organisms can still calcify below omega of 1 by spending metabolic energy to pump ions, but that energy comes out of growth, reproduction, or immune function. The damage often shows up as a smaller, weaker animal rather than as a dissolved one.

Key idea: Polar waters and upwelling coasts cross thresholds first, and the biological effect is usually a higher energetic cost of calcification rather than outright dissolution.

How we know

The evidence here is unusually clean. Two long-running open-ocean time series, one near Hawaii and one near Bermuda, have measured seawater carbonate chemistry monthly for several decades. Both show surface pH declining steadily at a rate close to 0.002 units per year, and both show dissolved inorganic carbon rising in step with the atmospheric carbon dioxide measured independently at Mauna Loa. The ocean signal tracks the atmospheric signal with the timing and magnitude that the chemistry predicts. There is no plausible alternative explanation, and the mechanism was worked out from first principles long before the trend was measurable.

Key idea: Multi-decade time series near Hawaii and Bermuda show surface pH falling at about 0.002 units per year in lockstep with rising atmospheric CO2, exactly as the chemistry predicts.

Common misconceptions

• "The ocean is becoming acidic." It is becoming less alkaline. Surface pH is about 8.1 and will remain above 7 under any plausible scenario; acidification names the direction of change.
• "A 0.1 pH change is negligible." Because pH is logarithmic, it is a 26 percent increase in hydrogen ion concentration.
• "Shells dissolve in the ocean today." Most surface water is still supersaturated. The usual harm is the rising energetic cost of building a shell, and undersaturation so far mainly at depth and in polar and upwelled water.
• "Acidification is just another word for ocean warming." They share a cause and nothing else. Removing heat would not fix acidification; only reducing CO2 would.
• "The ocean has been more acidic before, so this is fine." It has, but over thousands to millions of years, which allowed weathering and sediment buffering to keep pace. The rate is the problem.

Recap

• The ocean has absorbed roughly a quarter of human CO2 emissions, and acidification is the chemical price of that service.
• The net reaction is CO2 plus H2O plus CO3 2- gives 2 HCO3-, so added CO2 consumes carbonate ions.
• Surface pH has fallen from about 8.2 to 8.1, which is a 26 percent rise in hydrogen ions since 10 to the 0.1 equals 1.26.
• A further 0.3 unit drop would double hydrogen ions again, reaching about 150 percent above pre-industrial.
• Saturation state omega, proportional to carbonate ion concentration, determines whether carbonate is stable; below 1 it dissolves.
• Aragonite is about 50 percent more soluble than calcite, so corals and pteropods are affected before coccolithophores.
• Polar waters and eastern boundary upwelling systems cross thresholds first, as at the Whiskey Creek hatchery.
• Time series near Hawaii and Bermuda show pH falling about 0.002 units per year in step with atmospheric CO2.

Sources

1. National Ocean Service. (n.d.). What is ocean acidification? NOAA. oceanservice.noaa.gov
2. NOAA Ocean Acidification Program. (n.d.). Ocean acidification: the other carbon dioxide problem. oceanacidification.noaa.gov
3. Pacific Marine Environmental Laboratory. (n.d.). Ocean carbon and acidification data. NOAA. pmel.noaa.gov
4. Canadell, J. G., Monteiro, P. M. S., Costa, M. H., et al. (2021). Chapter 5: Global carbon and other biogeochemical cycles and feedbacks. In Climate change 2021: The physical science basis (IPCC AR6 WGI). ipcc.ch
5. Woods Hole Oceanographic Institution. (n.d.). Ocean learning hub: ocean acidification. whoi.edu
6. Wikipedia contributors. (n.d.). Ocean acidification. Wikipedia. en.wikipedia.org

The big picture

Almost everyone has seen the phrase Great Pacific Garbage Patch and pictured a floating island of bottles you could walk across. That picture is wrong in a way that matters. Sail through the middle of it and the water looks, to the naked eye, like ordinary open ocean. Tow a fine mesh net for an hour and you bring up a soup of fragments: shredded confetti, fishing line, opaque flecks smaller than a grain of rice, mixed with plankton. The patch covers something on the order of 1.6 million square kilometers and contains an estimated 79,000 tonnes of plastic, which sounds enormous until you divide it out and realize it is spread thinner than a few dozen kilograms per square kilometer.

That is worse news, not better. A raft you could see could be towed away. A dilute suspension of particles the size of plankton, distributed through the top of an ocean basin, cannot be filtered out without also removing the ecosystem. Understanding the actual physical form of the problem is what separates a useful response from a satisfying gesture.

This final lesson is about what humans do to the ocean beyond warming and acidification, what legal and practical tools exist to change that, and how a person actually goes about working on any of it.

Plastic

Somewhere between roughly 8 and 12 million tonnes of plastic enter the ocean each year, most of it from land, carried by rivers and by poor waste management rather than dumped from ships. Once at sea, sunlight and wave action embrittle and fragment it. Particles under 5 millimeters are called microplastics; they are either manufactured small, as with microbeads and industrial pellets, or produced by the breakdown of larger items and by fibers shed from synthetic textiles in laundry.

Where does it collect? You already know. The convergence zones at the centers of the five subtropical gyres, from Lesson 7, are exactly where Ekman transport piles surface water, and buoyant debris piles up with it. The garbage patches are not a separate phenomenon; they are the wind-driven circulation doing precisely what the physics says it does, applied to something that floats.

A detail that reorders priorities: a large fraction of the mass in the North Pacific patch is derelict fishing gear, nets, lines, and floats, rather than consumer packaging. Abandoned gear continues to catch animals for years as ghost gear, and it is the single most lethal category of marine debris for large animals.

The harms come in three forms. Entanglement drowns or maims turtles, seals, and cetaceans. Ingestion fills stomachs with indigestible material, so animals starve while feeling full; albatross chicks on remote Pacific atolls are fed plastic by parents who mistake it for squid. And plastics both leach their own additives and sorb other pollutants from seawater onto their surfaces, though how much of that transfers into animals in practice remains genuinely uncertain.

Two facts should shape any response. First, most plastic does not stay at the surface: it sinks, and microplastics have been found in deep-sea sediment, in Arctic sea ice, and in the bodies of amphipods from the hadal zone. Surface cleanup, by definition, reaches only the visible minority. Second, the flow is ongoing. Removing what is already there while the input continues is like bailing a boat without patching it. The interventions with the best evidence behind them are upstream: waste collection and management in the regions with the highest leakage, reduction of unnecessary single-use production, gear-marking and retrieval programs for fisheries, and design changes such as filters on washing machines. Open-ocean collection systems attract attention and funding out of proportion to the fraction of the problem they can address.

Key idea: Ocean plastic accumulates in gyre convergence zones as a dilute suspension of fragments rather than a solid raft, much of it derelict fishing gear, and upstream reduction beats offshore cleanup on the evidence.

Nutrients, eutrophication, and dead zones

The second great pollution problem is invisible, dissolved, and in some ways more damaging.

Fertilizer applied to farmland, treated and untreated sewage, and atmospheric deposition from combustion all deliver nitrogen and phosphorus to rivers and then to the coast. Recall from Lesson 5 that these are precisely the nutrients that limit marine production. Delivering them in excess does exactly what you would predict: an enormous algal bloom.

Then follow the carbon. The bloom is not eaten fast enough, so much of it dies and sinks. Bacteria decompose it, and decomposition consumes oxygen. If the water column is stratified, as it is over much of the shelf in summer because of solar heating and a freshwater cap from river discharge, then the bottom water is cut off from the atmosphere and cannot be resupplied. Oxygen falls. Below about 2 milligrams per liter the water is hypoxic, and mobile animals flee while sessile ones suffocate. The result is a dead zone.

The chain is worth stating as a sequence, because every link is a place to intervene: excess nutrients, then a bloom, then sinking organic matter, then bacterial respiration, then oxygen depletion under stratification, then hypoxia.

The best-documented example is in the northern Gulf of Mexico, fed by nitrogen from the Mississippi River basin, which drains a huge fraction of American farmland. Its summer extent has averaged on the order of 15,000 square kilometers in recent decades, with the largest measured event exceeding 22,000 square kilometers in 2017, against a management target of 5,000. Worldwide, more than 400 coastal systems have been documented as experiencing eutrophication-driven hypoxia, with the Baltic Sea hosting the largest.

Two clarifications keep this precise. First, dead zones are seasonal and reversible in a way that many marine problems are not: cut the nutrient load and they shrink, as demonstrated in the Black Sea when fertilizer use collapsed in the 1990s. Second, coastal eutrophic hypoxia is distinct from the open-ocean deoxygenation now being observed as the ocean warms, which arises because warm water holds less oxygen and because stronger stratification, from Lesson 6, slows resupply to the interior. The two share the word hypoxia and almost nothing else.

Key idea: Excess nutrients drive blooms whose decay strips oxygen from stratified bottom water, producing seasonal dead zones that shrink when nutrient loads fall.

Other stressors, briefly but honestly

Oil spills are dramatic and episodic; the 2010 Deepwater Horizon blowout released roughly 4.9 million barrels into the Gulf of Mexico over 87 days. Chronic small-scale inputs from runoff and shipping are less visible and cumulatively large. Underwater noise from shipping, seismic surveys, and sonar has raised background sound levels in ways that interfere with the acoustic communication whales depend on across ocean basins. Ship strikes kill large whales. And commercial interest in mining polymetallic nodules from the abyssal plain raises the problem noted in Lesson 4: those nodules grow at millimeters per million years and the communities on them are correspondingly slow, so any disturbance is effectively permanent.

Key idea: Beyond plastic and nutrients, chronic oil inputs, ship noise, vessel strikes, and prospective deep-sea mining act on timescales that range from continuous to effectively irreversible.

Marine protected areas: what works and what does not

A marine protected area is a defined region where human activity is restricted for conservation. Roughly 8 percent of the ocean now carries some MPA designation. But that headline number conceals the thing that actually matters, which is protection level. Only about 3 percent is fully or highly protected, meaning extractive activity is genuinely prohibited. Many designated areas permit commercial fishing, and some permit nearly everything; these are often called paper parks, and their measured biological effect is close to zero.

Where protection is real, the effect is large and well documented. A frequently cited synthesis of no-take reserves found on average several times higher fish biomass inside reserves than in comparable fished areas, along with higher density, larger individual size, and higher species richness. Larger fish matter disproportionately because egg production rises steeply with body size, so a reserve full of big old females exports far more larvae than its area would suggest. That export, plus adult movement across the boundary, produces spillover that can improve catches in the surrounding fishery. A well-placed reserve is not only a conservation tool; it can be a fisheries management tool.

The international target, adopted in the Kunming-Montreal Global Biodiversity Framework in December 2022, is to protect 30 percent of land and ocean by 2030, commonly shortened to 30 by 30. The honest assessment is that the area target is achievable and the protection-quality target is much harder, because it is politically easy to designate remote water that nobody fishes and politically hard to restrict productive grounds close to home. Large remote MPAs such as Papahanaumokuakea in the northwestern Hawaiian Islands and the Ross Sea MPA in Antarctica are genuine achievements and also, in part, the low-hanging fruit.

Key idea: Real no-take protection reliably multiplies fish biomass and exports larvae and adults to surrounding waters, but most designated area is weakly protected, so coverage statistics overstate actual protection.

Who owns the ocean?

The legal architecture matters because it determines who can act. The United Nations Convention on the Law of the Sea, adopted in 1982 and in force since 1994, divides ocean space into zones measured from a coastal state's baseline.

Zone	Extent from baseline	Coastal state rights
Territorial sea	Up to 12 nautical miles	Full sovereignty, subject to innocent passage
Contiguous zone	Up to 24 nautical miles	Enforcement of customs, immigration, sanitation law
Exclusive economic zone	Up to 200 nautical miles	Exclusive rights to fisheries, energy, and seabed resources
High seas	Beyond the EEZ	No state; open to all, governed by treaty and flag state

Roughly two thirds of the ocean by area lies beyond national jurisdiction. That is the classic setting for a tragedy of the commons: a resource everybody can use and nobody owns, where the benefit of taking one more fish accrues to the taker and the cost of depletion is spread across everyone. Regional fisheries management organizations govern particular stocks by agreement, the International Maritime Organization regulates shipping and pollution, and the International Seabed Authority controls mineral activity on the deep seabed, but for decades there was no comprehensive framework for conserving biodiversity in the high seas.

The agreement on Biodiversity Beyond National Jurisdiction, adopted in June 2023 and usually called the High Seas Treaty, was designed to fill that gap. It provides mechanisms for establishing marine protected areas on the high seas, requires environmental impact assessment for planned activities, and addresses the sharing of marine genetic resources. It required 60 ratifications to enter into force, a threshold reached in 2025; because ratification continues, check the United Nations treaty pages for the current list of parties before citing its status. It is the most significant change in ocean governance in a generation, and like every treaty its value will depend entirely on implementation and enforcement.

Key idea: UNCLOS gives coastal states exclusive rights out to 200 nautical miles and leaves about two thirds of the ocean as high seas, a commons the 2023 BBNJ agreement was written to protect.

Working in ocean science

A course like this should end by telling you honestly how people actually get into this field, because the popular image is misleading. Most oceanographers are not scuba divers. Many go to sea rarely. A large share of the work is programming, statistics, instrument engineering, and writing.

The realistic pathways are broader than a single track.

• Research. A doctorate is the standard entry, usually through a physics, chemistry, biology, geology, or engineering undergraduate degree rather than an undergraduate oceanography major. Institutions include universities, Woods Hole, Scripps, MBARI, and national laboratories.
• Technical and operational. Marine technicians, hydrographic surveyors, ROV and AUV pilots, ship's officers, and data managers. These are skilled, well-paid, in demand, and reachable through associate degrees, certifications, and apprenticeships rather than doctorates.
• Government and management. In the United States, NOAA, the USGS, the Environmental Protection Agency, state coastal and fisheries agencies, and the uniformed NOAA Corps. Equivalent agencies exist in most maritime countries.
• Policy, law, and communication. Fisheries policy, marine spatial planning, treaty work, journalism, museum and aquarium education. Often reached from a science background plus a law or policy degree.
• Private sector. Offshore wind and energy, subsea cables, environmental consultancy, insurance and shipping risk, ocean technology startups, and aquaculture. This sector is growing fastest and is the one students most often overlook.

If you want the single highest-leverage advice: learn to program and learn statistics. The bottleneck in modern ocean science is not collecting data; Argo floats, satellites, moorings, and autonomous vehicles produce more than anyone can analyze. The bottleneck is people who can handle large messy datasets competently. A biologist who can write clean code is far more employable than one who cannot, in every one of the pathways above.

The practical steps are unglamorous and effective. Get sea time or field time however you can, including as a volunteer or a deckhand. Apply for undergraduate research programs, which in the United States include the National Science Foundation's Research Experiences for Undergraduates and Sea Grant internships. Learn one instrument well. Read job listings from the organizations you admire two years before you plan to apply, and reverse-engineer the qualifications. And recognize that ocean science is competitive and often modestly paid relative to the training required; the people who thrive are the ones who find the work itself worth doing.

Key idea: Ocean careers span research, technical operations, government, policy, and a fast-growing private sector, and quantitative and programming skills are the most reliable differentiator in all of them.

Common misconceptions

• "The garbage patch is a floating island." It is a dilute suspension of fragments across a huge area, much of it derelict fishing gear, and much more plastic sinks than floats.
• "Dead zones are permanent." Most are seasonal and shrink when nutrient loading falls, as the Black Sea demonstrated.
• "Coastal dead zones and open-ocean deoxygenation are the same problem." One is driven by nutrient runoff, the other by warming and stratification.
• "8 percent of the ocean is protected." About 8 percent is designated; only about 3 percent is fully or highly protected, and weak designations show little measurable benefit.
• "You need to be a diver to be an oceanographer." Most of the field is data analysis, instrumentation, chemistry, and modeling. Many leading oceanographers have never done research dives.

Recap

• Roughly 8 to 12 million tonnes of plastic enter the ocean yearly, accumulating in the same gyre convergences that Ekman transport creates.
• The Pacific patch spans about 1.6 million square kilometers with roughly 79,000 tonnes of mostly fragmented debris and derelict gear.
• Upstream waste reduction outperforms offshore cleanup because most plastic sinks and the input flow continues.
• Eutrophication runs from excess nutrients to blooms to bacterial decay to hypoxia below 2 milligrams per liter under stratification.
• The Gulf of Mexico dead zone averages roughly 15,000 square kilometers each summer against a 5,000 target.
• About 8 percent of the ocean is designated as protected but only about 3 percent is fully or highly protected.
• UNCLOS grants a 200-nautical-mile EEZ and leaves about two thirds of the ocean as high seas, addressed by the 2023 BBNJ agreement.
• Ocean careers run from research to technical operations to policy to industry, and programming plus statistics is the common differentiator.

Sources

1. National Ocean Service. (n.d.). What is the Great Pacific Garbage Patch? NOAA. oceanservice.noaa.gov
2. National Ocean Service. (n.d.). What is a dead zone? NOAA. oceanservice.noaa.gov
3. National Ocean Service. (n.d.). What is a marine protected area? NOAA. oceanservice.noaa.gov
4. United Nations. (n.d.). Oceans and the law of the sea. Division for Ocean Affairs and the Law of the Sea. un.org
5. U.S. Bureau of Labor Statistics. (n.d.). Occupational outlook handbook: geoscientists. bls.gov
6. Woods Hole Oceanographic Institution. (n.d.). Careers and academic programs. whoi.edu