Why shouldn't you go near the white, foamy
part of a river?
"It looks shallow, the current looks gentle, it's probably fine" β the spot that looks like that can be the most dangerous place in the whole river. Let's set the hard science aside for a moment and start with the simple fact: such a place exists.
A river in summer. A little way upstream, a concrete ledge runs straight across the river, maybe 1β2 metres high. Water pours over it in a small waterfall.
Above the ledge, the water is calm, almost like a pond. The water spilling over doesn't look like much either. And right below the ledge, the surface is white with foam and slightly humped up. It doesn't even look deep. It's the kind of peaceful scene where you'd expect to see kids playing.
This "white, foaming patch right below the ledge" is what water-rescue professionals call a "drowning machine."
There are only two reasons this is dangerous
White foam is "water mixed with air β watered-down water." Watered-down water can't support your weight as well. Even strong swimmers, even people wearing a life jacket, can sink in it.
Below the ledge, the water forms a loop: "surface flows back to the ledge, only the bed flows on downstream." Just when you think you've drifted free, you're pulled back under the ledge again.
Because both happen at once, once you're in, getting out on your own is extremely hard. Let's look at each one with an everyday comparison.
Reason 1: foamy water is "watered-down water"
If you've ever swum in the sea, you've probably felt that the sea holds you up better than a swimming pool. In the Dead Sea, a famously salty lake in the Middle East, people can lie on their backs and just float.
That's because denser water pushes things up harder. Seawater, loaded with dissolved salt, weighs more than pool water for the same cupful. Heavier water pushes back harder against whatever's floating on it.
Now flip that around: what about "thinner" water? Water full of foam is packed with tiny bubbles of air. Air is far lighter than water, so foamy water weighs less for the same cupful β it's "thin" water. Thin water can't push things up as hard.
The key point is that the human body only just barely floats in plain water to begin with. Breathe in and you rise a little; breathe out and you sink a little. There's almost no margin. So even a small drop in water density is enough to sink you. A life jacket simply adds "floating material" to lift your body, so in thinned-out water it loses effectiveness too.
Water and air are both clear, so why does foam look white? Because countless tiny bubbles scatter light in every direction off their surfaces. It's the same reason clouds look white, shaved ice looks white, and clear glass turns white when ground into powder. In other words, "white and foaming" is the river telling you "this is thin water."
Reason 2: the water is spinning in a loop
Now, why can't you escape? Take another look at Figure 1.
Water falling over the ledge plunges hard into the riverbed. It then hugs the bed as it travels on a little, before rising back up β that's the "white, humped band." But the water that rises doesn't continue downstream; instead, it flows along the surface back toward the ledge. There it gets caught up in the falling water again, and is driven back down to the bed.
So below the ledge there's a loop of water: "bed flows downstream, surface flows back toward the ledge." It's like the water inside a washing machine.
Here's what happens if a person falls into it:
- Pushed down to the bed by the falling water (hard to float, because of the foam)
- Carried a few metres along the bed current, then rises up near the white band
- Just as they think "I can breathe!" β the surface current drags them back toward the ledge
- Pushed back under the falling water β back to step 1
One cycle takes a few seconds to maybe ten-odd seconds. The only chance to breathe is that brief moment in step 3. And desperately trying to "swim downstream" at that moment is usually like trying to walk against a moving walkway β it just burns through your strength.
Looking calm is the biggest trap of all
Nobody goes near a big waterfall or a raging rapid. But this kind of spot looks like a pond above the ledge, the falling water looks small, and downstream looks shallow too. And yet, when water rises even a little β say, the day after rain β the loop gets stronger. "The spot we played in yesterday" turns into something else entirely.
In the United States, deaths at these low ledges (dams) have been recorded in the hundreds, and the gap between how harmless they look and how deadly they are earned them the nickname "drowning machine." Japan's rivers also have countless such ledges β built to divert water for farming, or to protect the riverbed from erosion β and warnings about accidents near them are repeated again and again.
- A straight ledge or concrete structure running across the river
- Downstream of it, a white, foaming, humped band running in a straight line across the water
- Even though the water is supposedly flowing, branches or debris keep circling in the same spot (the clearest sign a loop exists)
- A "No Entry" or "No Swimming" sign. If locals never go in, there's a reason
So what should you do?
- Don't play near a ledge in a riverNot above it, not below it. Slip above, and you'll go straight over the ledge and into the loop.
- See white foam and debris circling the same spot? That's a "river brake."Stay away. Always wear a life jacket β though even with one on, you may not float in foamy water.
- If someone falls in, don't jump in after themReach out with a stick or branch. Throw something that floats (a cooler box, a bag of plastic bottles, a rope). Then call emergency services. According to U.S. agencies, about a quarter of deaths at these ledges are would-be rescuers. Remember: "reach, throw, never swim."
First, know this: escaping on your own, without help, is considered next to impossible. Every U.S. safety agency starts its explanation with that exact sentence.
Beyond that, the posture these agencies all recommend is to "tuck your chin to your chest and curl your knees up into a ball." Done well, this can sometimes let the current along the riverbed push you out of the loop (the blue arrows in Figure 1) β because the surface always flows back toward the ledge, so only the bed current heads downstream.
On the other hand, no agency says "don't try to float" or "dive down to the bottom on purpose." The advice is purely passive: "curl up, save your energy, and the current may carry you out."
Honestly, this isn't so much "know this and you'll be fine" as "even knowing this, it's very hard." Underwater, you lose all sense of up and direction. Which is exactly why staying away is everything.
Always wear a life jacket. "Foamy water is hard to float in" does not mean it's fine to skip one. Every safety agency requires wearing one. There's also a claim that wearing a life jacket makes it harder to escape the loop β but research exists that disputes this (see reference 5 below).
You can see the same thing in your kitchen
- Put a flat plate or cutting board in the sink and run the tap hard onto the middle of it
- Watch the water spread out thin from where it hits, then suddenly jump up into a "ring" a little way out
This ring is a very similar shape-change to the white band below a river ledge (fast, thin flow hitting slower, deeper water and suddenly rising). But that doesn't mean the same forces are at work: in a thin film like this, the water's viscosity and surface tension matter, and whether they're the main driver is still unsettled among researchers.
Summary
The white, foaming part of a river isn't dangerous because "the current is fast." It's dangerous because two things happen at once that you'd never guess just by looking: β foam thins the water so you can't float, and β‘ the water forms a loop that won't let you out.
White foam, and debris circling the same spot, over and over.
That's the river telling you, "don't come here."
The way a vortex spins faster as it's squeezed also shows up in the ferocious winds of a tornado. You can read more about that in this article.
Want to go deeper? Terms, equations, and where this sits in a textbookWe've labelled which level each part belongs to, from middle-school science to university-level courses
- MSCovered in middle-school science
- HSCovered in high-school "Basic Physics" / "Basic Chemistry"
- HS+Covered in high-school "Physics," or treated as advanced/sidebar material in textbooks
- UniNot covered in high school β university-level specialist subjects (hydraulics, fluid dynamics)
- ResearchNot even settled as "established fact" at university β something researchers are actively studying
MSTerminology: this phenomenon has names
- Weir / drop structure: what the main text calls the "concrete ledge." Built to divert water for irrigation or to stop the riverbed from eroding. A low one is called, in English, a low-head dam.
- Boil line: the main text's "white, humped band." "Boil" comes from the way the surface looks like it's boiling. This has nothing to do with the physics "Boyle's Law" (that pressure and volume are inversely related, named after Robert Boyle) β the English words just happen to sound alike, and in Japanese they sound identical too.
- Roller / reverse roller: the main text's "water loop" β specifically the part where the surface flow reverses back upstream.
- Hydraulic jump: the name for the phenomenon itself β the ring in your sink, or the white band below a ledge.
MSHSCheck it with an equation: in foamy water, how many kilograms' worth stays unsupported?
The main text's "dense water / thin water" is, in scientific terms, density (weight per unit volume). This is one place where a calculation gives a clear number.
Buoyant force = density of water Γ volume displaced
| Buoyant force | upward supporting force [kgf] |
| Density of water | 1.00 [g/cmΒ³] for plain water |
| Volume displaced | the part of the body submerged [L] |
You're pushed up by a force equal to the weight of water you displace (Archimedes' principle). Notice that neither "swimming ability" nor "effort" appears anywhere in this equation. It depends only on the density of the water and the volume of your body.
And air is roughly 1/800th the density of water. Just 10% foam by volume drops the water's density to 0.90. Since the human body's density is about 0.98, that alone is enough to make you denser than the water.
| Plain water (density 1.00) | 1.00 Γ 71 = 71 kgf β roughly balances a 70 kg body |
| 10% foam water (density 0.90) | 0.90 Γ 71 β buoyancy β 64 kgf β 6 kgf short, net downward |
| 30% foam water (density 0.70) | 0.70 Γ 71 β 50 kgf β 70 β 50 = 20 kgf short, net downward |
| Wearing a 7.5 kgf life jacket, in 0.70-density water | The jacket's own buoyancy also drops to ~5 kgf (Γ0.70) β still 15 kgf short |
β» A simplified estimate. Real outcomes vary a lot with posture, breathing, and turbulence, but the trend β "in foamy water, even a flotation device may not keep you up" β is widely shared in the rescue community.
In 30%-foam water, 20 kgf stays net downward. That's the same as trying to swim while holding a 20 kg weight β about two sacks of rice.
Even worse, 15 kgf stays unsupported even with a life jacket on, because the jacket's own buoyancy also drops in foamy water. That's what the last row of the table in β‘ is showing.
This is exactly where "I'm a strong swimmer, so I'll be fine" stops working. Swimming ability doesn't appear in the equation. Only density and volume do, and neither is something you can control. Which is why not going in is everything.
The force flowing water exerts on a body can also be calculated: F = Β½ Γ water density Γ speedΒ² Γ area.
| Water density | 1000 kg/mΒ³ |
| Current speed | 2 m/s (about walking pace) |
| Area exposed to flow | 0.3 mΒ² (standing waist-deep) |
| Substituting in | 0.5 Γ 1000 Γ 2 Γ 2 Γ 0.3 = 600 N |
| Converted to weight | 600 Γ· 9.8 β 61 kgf |
That's the force of one adult, pushing sideways, continuously. You can't stay standing. And because speed enters as a square, double the current speed and the force quadruples (2 Γ 2).
That's exactly why "it's only knee-deep, so it's fine" is a dangerous judgement. It's not depth that matters β it's speed. Even if a spot looks calm, the flow below a weir is faster than it looks.
HSUniReason 2, precisely: hydraulic jumps and the Froude number
HSFirst, the water speeds up as it falls over the ledge because potential energy converts into kinetic energy β this is basic physics.
UniWhat determines "why it suddenly humps up," though, is a quantity built from flow speed v and water depth h: Fr = v / β(g h), the Froude number. A hydraulic jump occurs where "supercritical flow" (fast and shallow, Fr > 1) transitions into "subcritical flow" (slow and deep, Fr < 1). Water just after falling over the ledge has a high Fr, and where it meets the slower water downstream, depth suddenly increases and a vortex forms. This is not taught in high school β it's university-level hydraulics and open-channel flow. Dam spillways are deliberately designed to trigger this hydraulic jump to safely dissipate the water's energy (a "stilling basin").
HS+The hydraulic jump is also an interesting example of where Bernoulli's principle breaks down. Bernoulli's principle assumes no energy is lost, but in a hydraulic jump, huge amounts of energy are lost to turbulence and foam, so it can't be applied directly (momentum conservation is used instead). Bernoulli's principle itself is often treated as advanced material in high-school textbooks.
HS+UniBonus: just how strong is flowing water?
A cubic metre of water weighs about a tonne. The force when water flowing at speed v hits an area A is roughly F β Β½ Ο vΒ² A Γ C (C is a shape-dependent coefficient, roughly 1). Water flowing at 2 m/s hitting a person's torso (about 0.3 mΒ²) gives Β½ Γ 1000 Γ 4 Γ 0.3 β 600 N, about as much force as a 60 kg body weighs.
That 0.3 mΒ² assumes "submerged to the waist"; below the knee it's roughly half that. The key point is that speed matters as a square β double the current speed and the force quadruples. But whether a person actually gets swept off their feet isn't decided by comparing this force to their body weight β it comes down to friction with the riverbed. Since buoyancy already lightens the body, people's feet can be swept out from under them by a force much smaller than "a force equal to their body weight." One experiment reported people starting to slip at a depth of 0.23 m and a flow speed of 1.8 m/s. UniThis equation (the drag equation) and coefficient C (the drag coefficient) are studied in fluid dynamics, but HS+deriving the form Ο vΒ² A from "the momentum carried by water per unit time" can be followed using nothing more than high-school physics momentum concepts.
ResearchWhat's still not fully understood
Everything so far has been "what's written in textbooks." But this phenomenon still has parts that aren't settled β a reminder that there's a world beyond the textbook too.
- Exactly how much buoyancy foam removes isn't precisely known. "You can't float in foamy water" is common knowledge in the rescue world, but there's very little actual data measuring air content at the moment someone is drowning. Foamy flow is opaque, and inserting a sensor changes the flow itself β measurement itself is difficult. Probes using fibre optics or electrical resistance have been developed, but field measurements on real rivers remain limited.
- Is a lab hydraulic jump the same as a real river's hydraulic jump? Hydraulics experiments use scaled-down models, but bubble size doesn't shrink to match the model scale (because water's surface tension doesn't scale down). This mismatch is called a scale effect, and whether model results can be applied directly to real rivers remains a live debate.
- The conditions under which a person can escape the circulating flow haven't been modelled yet. The advice "dive to the bottom to get out" is inferred from experience and the structure of the flow, not verified with studies that actually track human body movement. There's no formula yet that predicts, for a given flow and ledge height, whether escape is possible.
- Nobody has a full picture of how many dangerous weirs exist, or where. Low weirs are often too small to appear in national dam registries, and even in the US, surveys began from a state of "nobody even knows how many there are." In Japan too, a comprehensive hazard map is still a work in progress.
In other words, even this article is "an explanation based on what's currently known." Most scientific topics have this layer of "not yet known" underneath them.
Where this fits in textbooks (by level)
| Level | Subject / unit | Where in this article |
|---|---|---|
| MS | Science: density / buoyancy, water pressure / energy | The "dense water / thin water" discussion, Figure 2, why water speeds up at the ledge |
| HS | Basic Physics: force balance / buoyancy / mechanical energy | The Archimedes' principle equation, the buoyancy estimate table |
| HS | Basic Chemistry: density and mixtures | How average density works for water mixed with foam |
| HS+ | Physics: momentum / Bernoulli's principle (often advanced material) | The Β½ΟvΒ²A estimate of flow force; why Bernoulli fails at a hydraulic jump |
| Uni | Hydraulics / open-channel flow (civil engineering) | Froude number, supercritical/subcritical flow, hydraulic jumps, stilling basins |
| Uni | Fluid dynamics / multiphase flow | Gas-liquid two-phase flow, void fraction, drag coefficient |
| Research | Hydraulics / water-rescue research (unresolved) | Measuring void fraction in real rivers, scale effects, modelling escape conditions, comprehensive hazard surveys |
| β | Disaster prevention / safety education | How to spot one, Reach, Throw, Row, Don't Go, preventing secondary drownings |
- U.S. Army Corps of Engineers, Low-Head Dam Inventory β Safety Tips. https://nid.sec.usace.army.mil/lhdi/dam-basics/safety-tips (curling into a ball, Reach, Throw, Row, Don't Go, roughly 25% of deaths being rescuers)
- Ohio Department of Natural Resources, Lowhead Dam Safety ("escape without help is nearly impossible"). / Indiana DHS, Low-Head Dam Safety Tips.
- Minnesota DNR, The Drowning Machine (2012). Reduced buoyancy from air bubbles, and reduced flotation even with a life jacket.
- WΓΌthrich, D., Shi, R. & Chanson, H., Air entrainment in hydraulic jumps, Environmental Fluid Mechanics 22(4), 2022 / Wang, H. & Chanson, H., Canadian Journal of Civil Engineering 45(2), 2018 (measured void fraction inside hydraulic jumps). Chanson, H., Hydraulics of aerated flows: qui pro quo?, Journal of Hydraulic Research 51(3), 223β243, 2013.
- Gioia, G. et al., Residence time of buoyant objects in drowning machines, PNAS 108(15), 2011 (a study finding that wearing a life jacket barely changes time-to-escape).
- Jonkman, S. N. & Penning-Rowsell, E., Human Instability in Flood Flows, JAWRA 44(5), 1208β1218, 2008 (slipping at 0.23 m depth, 1.8 m/s flow speed).
- Duchesne, A. & Limat, L., Circular hydraulic jumps: where does surface tension matter?, J. Fluid Mech. 937, R2, 2022 (the debate over sink hydraulic jumps).
- National Police Agency of Japan, "Summary of Water-Related Accidents" (annual), and awareness materials on water-accident prevention from the Japan River Foundation (ε ¬η貑ε£ζ³δΊΊζ²³ε·θ²‘ε£).
β»This article is a general-audience science explainer. For actual safety decisions, follow on-site warning signs and the instructions of local authorities, fire services, and river management bodies. The figures given are approximations meant to help explain the mechanism.