Why Do Ponds and Lakes Freeze, but the Sea Hardly Ever Does?
― Chilled seawater doesn't stay at the surface. It sinks.
"Because the sea is salty" is only half the answer. Salt lowers the freezing point by just 2℃ or so. What really matters is that chilled seawater becomes heavier and sinks. So the sea can't start freezing until the water is cooled deep down, not just at the surface.
On a midwinter morning, a thin sheet of ice covers the pond in the park. A puddle by a nearby rice field has frozen white, too.
The same day, you go to the seaside. The waves are rolling as usual. The air is just as far below freezing, yet there is no ice anywhere.
If a child asked, "Why doesn't the sea freeze?", how would you answer? If you stop at "because it's salty", you leave out the more important half.
There are only two reasons the sea resists freezing
When something is dissolved in water, the water becomes harder to freeze. For ordinary seawater, freezing starts not at 0℃ but at about −1.9℃.
In a pond, chilled water stays on top and freezes there. In the sea, chilled water gets heavier and sinks, trading places with the water below. For the sea to freeze, the whole mixed layer has to be cooled through.
The first reason is well known. But a difference of 2℃ alone would let seas in cold regions freeze easily every winter. The second reason is the one that matters most.
A pond freezes from the "lid"; the sea must wait until the whole "pot" is cold
Fresh water has an odd property. As you cool it, it is heaviest at 4℃. Cooled below that, it actually gets lighter.
In a winter pond, the whole body of water first cools to around 4℃. If the surface water then cools further, it is lighter, so it doesn't sink. A thin cold layer forms on top, and only that layer reaches 0℃ and freezes. Only the "lid" needs to freeze.
Once salt is dissolved, things change. The more salt there is, the lower the temperature of greatest density becomes. When the salinity goes above about 2.5%, that temperature drops below the freezing point. Ordinary seawater is about 3.5% salt, well past this boundary.
So seawater keeps getting heavier as it cools, right up until it freezes. Water chilled at the surface sinks at once, and slightly warmer water rises from below. Move the slider in Figure 1 to see where this switch happens.
What happens before the sea freezes?
When cold wind chills the sea surface, the chilled water sinks and water from below rises. This swapping continues down to a depth where salinity or temperature changes sharply. The layer above that depth is called the "mixed layer", and it is said to be tens to hundreds of metres thick, depending on place and season.
In a pond, only a thin surface layer had to be cooled. In the sea, ice can't form at the surface unless water 100 metres deep is cooled all the way to −1.9℃. The amount of water to be chilled is vastly larger. Figure 2 sets a pond and the sea side by side.
Even so, the sea does freeze in the Arctic and Antarctic, and along the coast of the Sea of Okhotsk in Hokkaido. In the Sea of Okhotsk, fresh water from large rivers forms a low-salinity layer at the surface. Below this thin layer, salinity rises sharply, which is said to keep the mixing depth shallow. Less water has to be cooled, so Hokkaido's coast is considered one of the southernmost places in the Northern Hemisphere where drift ice is seen.
When seawater freezes, only the water molecules line up first to form ice, and most of the salt is pushed out. So sea ice (called "sea ice" in science, kaihyō in Japanese) is much less salty than the seawater it came from. The pushed-out, concentrated brine is heavy and sinks. This is thought to be one of the forces that drive the deep ocean currents of the whole planet.
Summary
The sea resists freezing for two reasons: salt lowers the freezing point to about −1.9℃, and chilled seawater sinks. A pond freezes once just a thin surface layer is cold. The sea can't form surface ice until the entire layer, mixed down to great depth, has been cooled through. "Because it's salty" is only the doorway. The real answer is "because chilled water sinks".
A pond freezes once its lid is cold.
The sea can't freeze until the whole pot is cold.
For how water being heaviest at 4℃ drives the seasonal changes inside a lake, see "Why Does a Lake's Water Turn Over, Top to Bottom, in Autumn?". For how salt lowers the freezing point, see "Why Does Salt on Ice Make It Colder Than 0℃?". And for why the sea is salty, see "Why Is Seawater Salty?".
- Prepare two identical cups. Put plain water in one, and 200 mL of water plus a heaped teaspoon of salt (about 7 g, roughly the saltiness of seawater) in the other.
- Put both in the freezer side by side and check the surface every 30 minutes. The fresh water should start freezing from the surface first.
- Once the salt water has frozen, lick a little of the half-melted ice. The first water to melt is salty, and the remaining ice may taste less salty than you'd expect.
A cup is shallow, so it can't reproduce the deep mixing of the sea. What it can show is two things: the freezing point drops, and salt is pushed out of the ice. When tasting anything from the freezer, use clean cups and edible salt.
For those who want more ― terms, formulas, and links to textbooksEach part is labelled with its level, from junior-high science to university specialist courses
- Junior highCovered in junior-high science
- High schoolCovered in high-school chemistry or earth science
- High school+Advanced high-school material, or a textbook sidebar topic
- UniversityNot taught in high school; university specialist courses (physical oceanography, physical chemistry)
- ResearchNot yet taught as settled fact even at university; being investigated by researchers right now
Junior highTerms: this phenomenon has names
- Freezing-point depression: the lowering of the temperature at which water starts to freeze (the freezing point) when something is dissolved in it.
- Temperature of maximum density: the temperature at which water is heaviest (has the greatest density). About 4℃ for fresh water.
- Mixed layer: the layer from the sea surface down to the depth where water is well stirred by sinking chilled water and by waves.
- Sea ice (kaihyō): ice formed from freezing seawater. Pieces that drift away from the shore are called drift ice (ryūhyō).
Junior highHigh schoolChecking with formulas: the freezing point of seawater, and the days needed to cool it
First, we work out how far the freezing point of seawater drops. Then we estimate how many days it takes to cool the mixed layer by 1℃.
| In symbols | ΔT ＝ Kf × m (and Q ＝ M × c × Δθ) |
| In words | Drop in freezing point = a constant set by the solvent × the number of dissolved particles (moles) per kg of water. Heat needed to cool = mass of water × specific heat × temperature drop |
| Where it comes from | Dissolved particles dilute the water molecules' "tendency to freeze", and the effect depends only on the number of particles, not on what the substance is (a property of dilute solutions). The heat formula comes from conservation of energy |
| Freezing-point depression constant of water, Kf | 1.86 ℃・kg/mol |
| Total dissolved particles (ions) in seawater, m | About 1.0 mol per kg of water |
| Mixed layer thickness (assumed here) | 100 m |
| Mass of 1 cubic metre of seawater | About 1000 kg |
| Specific heat of water, c | About 4200 J/(kg・℃) |
| Heat lost from the sea surface in a strong midwinter cold spell (assumed here) | 200 J per second per square metre |
| Drop in freezing point | 1.86 × 1.0 ＝ 1.86 (℃) |
| Mass of mixed-layer water under 1 square metre of sea surface | 100 × 1000 ＝ 100000 (kg) |
| Heat removed to cool that water by 1℃ | 100000 × 4200 ＝ 420000000 (J) |
| Time needed (seconds) | 420000000 ÷ 200 ＝ 2100000 (s) |
| Converted to days | 2100000 ÷ 86400 ≒ 24 (days) |
The freezing point is 0 − 1.86, about −1.9℃, which matches the real value for seawater. And a 100 m mixed layer takes more than three weeks to cool by just 1℃, even in a strong cold spell. The top few centimetres of a pond can freeze overnight. This difference in quantity is what lies behind "the sea is slow to freeze".
High schoolHigh school+Properties that depend on particle count, and density changed by salt
High schoolFreezing-point depression is one of the "colligative properties", along with boiling-point elevation and osmotic pressure. Its size depends on the number of particles, not on what is dissolved. Table salt splits into two ions in water, so it works about twice as strongly as the same number of moles of sugar.
High school+The two lines in Figure 1 are approximations that both fall roughly in proportion to salinity. The freezing point drops about 0.54℃ per 1% of salinity, and the temperature of maximum density drops about 2.2℃ per 1%. The latter has the steeper slope, so the two lines cross at about 2.5% salinity and about −1.3℃.
UniversitySea-ice formation and thermohaline circulation
In physical oceanography, sinking caused by cooling of the sea surface is treated as "convection", and the depth of the mixed layer is set by the strength of cooling, wind, and the structure of the salinity layers. In seas with low surface salinity, like the Sea of Okhotsk, a density step (the halocline) keeps convection shallow. Concentrated brine pushed out when sea ice forms sinks, and around Antarctica and elsewhere it becomes the source of deep water. This is part of the planet-wide ocean circulation called "thermohaline circulation".
📖 Derivations and further reading: Freezing-point depression (Wikipedia, in Japanese) ／ Thermohaline circulation (Wikipedia, in Japanese)
ResearchWhat is still not well understood
- How will deep ocean currents change if sea ice shrinks? If less sea ice forms, less concentrated brine sinks. How much the deep circulation will weaken is being studied through both observation and computer models.
- The future of sea ice in the Sea of Okhotsk. Drift ice is said to be decreasing over the long term, but year-to-year swings are large, and research continues on how changes in river water and wind will affect it.
- Brine channels inside the ice. Thin channels of salt water remain inside sea ice like a mesh. How their shape affects heat flow and the lives of organisms is not yet understood in detail.
In other words, this article too is "an explanation of what is understood so far". The link between sea ice and climate in particular is a field where the numbers may change.
Links to textbooks (by level)
| Level | Subject and unit | Where in this article |
|---|---|---|
| Junior high | Science: changes of state and density | Water is heaviest at 4℃; chilled water sinks |
| High school | Chemistry: properties of dilute solutions | Getting −1.9℃ from the freezing-point depression formula |
| High school+ | Earth science: structure of the ocean | Mixed layer; temperature of maximum density changing with salinity |
| University | Physical oceanography | Convection, halocline, thermohaline circulation |
| Research | Climate and sea-ice research | Sea-ice decline and deep circulation |
| ― | Connection to daily life | Winter seaside scenery, drift-ice tourism, freezer experiment |
- Wikipedia, "Freezing-point depression" (凝固点降下, in Japanese)
- Wikipedia, "Sea ice" (海氷, in Japanese)
- Wikipedia, "Drift ice" (流氷, in Japanese)
- Wikipedia, "Thermohaline circulation" (熱塩循環, in Japanese)
- National Astronomical Observatory of Japan (国立天文台), ed., Rika Nenpyō (理科年表, Chronological Scientific Tables), Maruzen Publishing (properties of seawater)
* This article is a popular-science explanation. The figures given are rough guides to help you understand the mechanisms. Mixed-layer thickness and heat lost from the sea surface vary greatly with place, season, and weather.