American Scientist

Paired videos recently debuted on two well-known science YouTube channels—Derek Muller’s Veritasium and Destin Wilson Sandlin’s Smarter Every Day. I saw the title, “The Truth about Toilet Swirl,” and I was intrigued. So I clicked. The setup of the videos was clever: They were meant to be played in synchronization, like some sort of science tango. Even more intriguing, one video had been filmed in the northern hemisphere and the other in the southern hemisphere so that Muller and Sandlin could find out whether water going down a drain really spins in opposite directions on either side of the equator. They were investigating the Coriolis effect.

You can imagine my excitement as a meteorologist. The Coriolis effect causes storms to rotate counterclockwise in the northern hemisphere and clockwise in the southern hemisphere. And now there were stereoscopic videos on a topic involving my field of study that didn’t consist of jokes made at the expense of the weatherperson. Plus, most videos about the Coriolis effect involve spinning metal death machines known as merry-go-rounds, so I was excited to see something new. And then I watched.

Muller and Sandlin somehow made the water draining out of a kiddie pool in the southern hemisphere rotate in the opposite direction from water draining out of a kiddie pool in the northern hemisphere. They claimed the Earth’s rotation moved the liquid. Honestly, my first thought was incredulity, followed by many more thoughts best described as expressing a healthy dose of skepticism. There was no way Earth’s rotation was moving water in something as tiny as a 1.5 meter kiddie pool by way of the Coriolis effect. Or was there?

So I did what any self-respecting scientist would do. I consulted my Facebook friends. Was I correct that this experiment was wrong? Or was I missing something? And boy did they respond. What commenced was the nerdiest comment chain in recent memory: We discussed the forces at play within the kiddie pool. We estimated key variables to determine the likelihood that the experiment was real and possible. We attempted to figure out the deflection of the water from the north and south ends of the pool by the time it reached the middle. Basically, we geeked out. Hard. And after discussing the experiment with more meteorologists than have any right to be following what I write on Facebook, I came to a conclusion. Contrary to my highly negative initial reaction, the YouTube sci-comm all-stars’ rationale was correct. So were they right? Well, I don’t know. Maybe. Let me explain.

To meteorologists, when one discusses the effect of the Earth’s rotation on a fluid, or something that acts pretty much like a fluid (like the atmosphere), we immediately think of not only the Coriolis effect but something called the Rossby number. This dimensionless number reflects a ratio comparing a fluid’s inertia to the force of Earth’s rotation. Or simply put, the Rossby number attempts to answer whether the effects of Earth’s rotation are a major factor in determining how air or water move in a given location. In general, if the Rossby number is greater than 1, then the Earth’s rotational effect is negligible, meaning that the fluid can spin whatever way it wants to. If it is less than 1, Coriolis wins out.

Specifically, the Rossby number compares the horizontal velocity of the system (that is, the rate at which it spins) to the size of the system combined with the effect of the Earth’s rotation (the latter in the form of a Coriolis parameter). For instance, consider a regular low-pressure weather system. Its size is pretty large, potentially spanning half the United States, and wind speeds are much lower than those of hurricanes and tornadoes. In its Rossby number equation, we can expect to see a small number in the numerator and a larger number in the denominator. Thus, the low-pressure system’s Rossby number would be lower than 1, indicating that its spin is influenced by the Earth’s rotation.

A tornado, however, has a high horizontal velocity (meaning it spins really fast) and a relatively small size. Its Rossby number is large. The Earth’s rotation should not affect the way tornadoes spin. And in fact you can see tornadoes rotating in counterclockwise and clockwise directions in either hemisphere.

But back to the experiment. This is a kiddie pool—tiny compared to a hurricane or a tornado. How can the Coriolis effect win out over other forces on such a small scale? Let’s put some basic figures into the Rossby number calculation to see. If the radius of a pool is 1 meter, the Coriolis parameter at a latitude of 40 degrees north is roughly .0001 radians per second. We can conclude that simply getting the Rossby number below 1 would require the horizontal velocity to be lower than .0001 meters per second (roughly 1.2 feet per hour). That is incredibly still. In contrast, your toilet or sink drains much faster and thus is not affected by the Coriolis effect, even without the jets inside toilets shooting water in a certain direction.

This is why, as Muller explained, the only way to get the Earth’s rotation to influence water on such a small scale would be to reduce all other forces to nearly nothing: The water had to be basically motionless. Muller and Sandlin set up their experiment on correct premises. All things being equal, if you make sure that the water is motionless and no other forces are introduced during the process, Coriolis would be the big winner in the battle royal of forces acting on the liquid draining from that kiddie pool. And for Muller and Sandlin, it worked! They tried their experiment three times in each hemisphere. In each case the water rotated clockwise in the southern hemisphere and counterclockwise in the northern hemisphere. Proof! Right?

Well, not so fast. The YouTube experiments were actually based on previous ones done in a laboratory setting. In those highly controlled settings, scientists at MIT in the 1960s were able to show that Coriolis could work on a draining tub. In fact, I have been told that graduate students at MIT still do this experiment today in one of their classes. The major difference between the past examples and the current YouTube version is that one was done in a lab with a fine control over outside forces and the other uses a kiddie pool set up on a plywood platform in a garage or sheltered patio. In the video experiments any number of things may have introduced an outside force that could swamp the comparatively tiny influence from the Earth’s rotation. Temperature differences in the water could create currents; tiny bumps in the texture of the kiddie pool or plywood could guide the flow; the way the valve released the water could steer the movement; and, especially for the outdoor experiment, a slight breeze could push the water along. A systematic error could have led to the same consistent results seen in the videos.

This is not to say these YouTube experiments did not work. It’s just that a lot of other things could have affected the results besides the Earth’s rotation. Of course, regular folks do not have access to the type of laboratories and equipment needed to control for all other factors and allow the Earth’s rotation to determine the outcome. And part of the purpose of these videos is to show viewers an experiment they can recreate on their own.

Maintaining that degree of accessibility while satisfying skeptics like me would involve putting together a larger sample size of experiments similar to the ones made by Muller and Sandlin, done in different locations across both hemispheres. Then we could get a better idea of whether the setup shown in “The Truth about Toilet Swirl” did in fact allow the Coriolis effect to shine. Until then, I’ll remain unconvinced that what we saw truly resulted from the Coriolis effect.

Ultimately, of course, this is nitpicking. The videos were educational, and the explanation of what the Coriolis effect means for air or water moving on Earth was on the money, regardless of whether I completely accept the experiment’s setup.

Therefore, I guess the only thing left to do is to assemble an army of kiddie-pool Coriolis experimenters to figure this out once and for all. If you bring the pool, I’ll supply the water (void in California). The Earth will provide the spin.

Correction: This post originally said that the Coriolis parameter at a latitude of 40 degrees north is roughly .0001 meters per second squared. At a latitude of 40 degrees north, the Coriolis parameter is about .0001 radians per second.