We now have an experiment that shows how a fresh, yellow inflow (representing the freshwater input into fjords close to their heads by rivers) flows over a initially stagnant pool of salt water. As the freshwater plume flows out of the fjord, it entrains more and more salt water from below, thus thickening and setting up a return flow that brings in more salt water from the reservoir (representing the open ocean) on the right.
We drop dye crystals to visualize the surface current going out of the fjord and the return flow going in, and draw the profiles on the tank to be able to discuss them later.
Here is a movie of the whole thing:
But there is more to see!
When tipping the tank to empty it, a lot of turbulence was created at the sill (see movie below). While a fjord typically isn’t tipped very often, what we see here is basically what tides do on the sill (see the waves that keep going back and forth over the sill after the tank is initially lifted? Those are exactly like tides). This could purposefully integrated in teaching rather than only happen by accident, those waves could be created just by surface forcing rather than by tipping the tank. That’s a very nice demonstration to explain high mixing rates in the vicinity of steep topography!
And then there is also the issue of very low oxygen concentrations in some Norwegian fjords, and one proposed solution is to bring the river inflow deep down into the fjord. The idea is that the less dense river water will move up to the surface again, thereby creating mixing and oxygenating the stagnant deep water that, in some cases, hasn’t been renewed in many years.
We model this by putting the inflow (the hose) down into the tank and see the expected behaviour. What we also see: Since the water has a quite strong downward component as it enters the fjord, it stirs up a lot of old dye from the bottom. So possibly something to be aware of since there might be stuff dumped into fjords that you might not necessarily want to stir up…
And last, not least, a bonus picture: This is how we measure temperature at GFI. You would think it should be possible to find a smaller thermometer that isn’t an old reversing mercury one? But in any case, this worked very well, too :-)
How will lee waves look differently if we are using the asymmetrical mountain instead of the symmetric one? And is symmetry actually important at all or are we just looking at different slopes downstream while the upstream slope doesn’t have an influence on the wave field?
There are a couple of reasons why I had not done that before:
It’s longer (1.5 m instead of the 1 m of the other mountain), therefore the tank is, relatively speaking, shorter. And since being close to the ends of the tank leads to weird interferences, this limits the distance over which observations can be made
Since it’s asymmetrical, pulling one way or the other would likely show different wave fields, so you couldn’t just run it back and forth and have students observe the same thing several times in a row
But then it would be really interesting to see what the difference would be, right?
I tried two different stratifications.
Weak stratification, shallow water
Since I just wanted a quick idea of what this mountain would do, I used leftover water I had prepared for the moving mountain experiment. Since there wasn’t a lot left, I ended up with 11.5 cm fresh water, but only 4 cm salt water at approximately 20 psu (since I stretched the 35 psu a little).
What I noticed: A LOT more mixing than with the other mountain! Stratification is pretty much destroyed after the first run, usually we run back and forth a lot. This can be for several reasons:
The water is very shallow, meaning mixing is happening over the whole water column. It might not actually be more mixing than in the other case, but since it’s affecting the whole water column, it might just seem like more because no clearly visible stratification is left above and below the layer which is mixed by the mountain?
The left side of the mountain was bent up a little (as in 2 or 3 cm), meaning that especially on the way back it was flapping up and down on the upstream side, doing a lot of mixing that wasn’t due to the shape of the mountain, just of bits of it being loose.
And the shape of the “reservoir” that is being built up upstream of the mountain is different to what I have observed before: Running in either direction, the reservoir didn’t built up smoothly, but as a hump that was pushed in front of the mountain. Maybe because the internal wave speed in this case was very close to the speed of the mountain, something like 7cm/s, so the disturbance created by the mountain couldn’t propagate upstream. Is that an upstream hydraulic jump we are seeing there?!
Since I was not satisfied with this at all, I ran a second experiment:
Strong stratification, deep water
First, I tried to set up the same stratification as for this lee wave experiment with the symmetrical mountain because I thought that would be easiest to compare. But I aborted that after having moved the mountain just a little because it was mixing so much that there stratification was destroyed completely and nothing could be seen. I ended up putting more dense water in and ended up with 12 cm pink (35 psu) and 4 cm clear freshwater. And this is what this looked like:
You now see a wave train with wave lengths longer than in the symmetrical case. Probably due to the longer length of the obstacle (even thought the waves are still shorter than the obstacle)? Or what sets the wavelength?
This time, with a faster internal wave speed of around 10cm/s while the mountain is still pulled with 7cm/s, we don’t see the “hump” in the upstream reservoir — the signal can propagate faster than the mountain and thus smoothes out.
So that is what I think is going on here. While the first experiment mainly showed effects of the stratification compared to previous experiments, the second one might provide some insight on the different slopes of the mountain, although I am not sure in what way. Do you see something I didn’t observe? How would you expect the different slopes to influence the lee waves?
I am so glad I tried this and I’m looking forward to thinking about this more! :-) Any insights you’d care to share with me?
This blog post is meant as guidelines if someone other than me might have to set up this demonstration at some point… Have fun! :-)
Setting up the stratification
If I am working fast and nothing goes wrong, this takes almost 2.5 hours. Make sure you have enough time to set this up! Filling the tank takes time and there is not much you can do to speed up the process once you’ve started…
Fill in what will end up being the top layer: 5 cm at 0 psu. For this, connect the tap to the bottom inlet in the left corner of the mountain with one of the hoses. When you are done, make sure to close the lock at the tank!
Move “mountain” over inflow to contain mixing to the volume underneath the mountain (better for your nerves, trust me)
Prepare the future bottom layers one by one (35 cm at 35 psu). We will need four full fillings of the 80l barrel (which doesn’t empty all the way because the tap is slightly elevated from the bottom, in case you were calculating ;-)), each with 2.8kg salt dissolved in it. To prepare that, connect the hose from the tap to the outlet of the barrel, put in the salt, put in the dye, use a paddle while you fill the barrel with water to stir. This way the salt will be pretty much dissolved by the time the barrel is full.
Note: Make sure the barrel is located high enough so that gravity will pull the water down in the tank from the barrel!
Note: When the barrel is filled, close the lock at the barrel before disconnecting the hose to reconnect it to the tank!
Fill in the bottom layers into the tank one by one. While one layer is slowly running into the tank, you have time to measure the salt for the next one.
Pulling the boat
Here is a sketch of the contraption that pulls the boat:
Put 4 or 5 gram in the little zip lock bag (called “weight” in the sketch above). This only works when the ship
Set up bumper to stop the ship before weights reach the floor (too much slack on the line, line might come off pulleys)
Stern rope on one of the tank’s braces is set up so the line is stretched as far as it can safely go
Check that there are marks on the tank which help measuring the speed of the boat (6 marks over 3 meters work well)
If there is suddenly too much friction in the system, check: Did the pulley on the left edge of the tank fall down? Did the rope come off the pulleys (sometimes happens if there was too much slack in the system, e.g. if the bag has been lifted or the bumper is too far left)
If the boat is moving a lot faster in the beginning than in the end, even though waves haven’t caught up with it, and it bothers you, move the two fixtures that hold the line at the ceiling closer together. Ideally, they should be in the same place, but this didn’t work for us because of tangling lines. Compromise between “constant” force and being able to run the experiment at all…
Ask students to observe:
Speed of the boat (actually take the time for a set distance)
Development of the boat’s speed over time, especially when waves are catching up with it
Generation of internal waves. Is there one, are there many? What are their wavelengths and speeds?
Generation of surface waves and their size relative to the internal waves. Why?
Below are movies of a couple of experiments which you could use in teaching instead of running the experiment for real (if for some reason running the experiment is not possible. But I would totally 100% recommend doing the experiment!). For a fun video, watch the one above (the ones below are cut to only show the tank so might be a little boring less exciting ;-))
Ship pulled with 5g in the bag
Ship pulled with 4g in the bag (for a repeat, see experiment 4!)
Ship pulled with 3g in the bag
Ship pulled with 4g in the bag (again, because we like repeat experiments ;-))
This blog post is meant as guideline if someone other than me might have to set up this demonstration at some point… Have fun! :-)
Lee waves are the kind of waves that can be observed downwind of a mountain in the clouds, or downstream of an obstacle in a current as a series of undulations with crests parallel to the disturbance.
Why move the mountain?
Students sometimes find it hard to imagine that a moving mountain should be equivalent to flow across a ridge. It helps to discuss how it would be really difficult to set up a flow in a tank: A huge amount of water would need to be moved without too much turbulence. Instead, it’s a lot easier to imagine the water is moving by moving a mountain through the tank, so the water is moving relative to it if not relative to the lab.
The size of the tank is 60×1.5×5 dm, so it can hold a total of 450l of water.
The mountain we use is 10.5 cm high and 1 m long and it’s symmetric, so pulling it either way shows similar lee waves (which is why I’ve always used it). There is a second, asymmetrical mountain on the shelf that I have never used*.
Setting up the stratification
The stratification that we’ve found works well is 10 cm at 35 psu (here dyed pink) and 9 cm at 0 psu. This leads to an internal wave speed of approximately ~11cm/s.
Prepare the dense water in a barrel that sits high enough so gravity will bring the water down into the tank (see picture below). For the 80l barrel, you need 2.8kg of salt and 1/3 tea spoon of dye MAX.
Elin’s GEOF213 class observing lee waves
You achieve the stratification by filling in the fresh water first through the bottom left inlet, moving the mountain over it, and filling in the dense water. That way the mixing is contained to the volume underneath the mountain which will be a lot better for your nerves (believe me!).
Moving the mountain
The system that pulls the mountain can go at two speeds: “fast” and “slow”, “slow” meaning 5m in 1:11min (7cm/s) and “fast” meaning 5m in 0:36min (14cm/s).
Here is where you run the mountain from:
Troubleshooting if the mountain doesn’t move:
you might be trying to pull the mountain in the wrong direction (into the wall)
the mountain might not be located on the sledge well. There is a tongue on the sledge that needs to sit in the groove in the mountain
the mountain might not be sitting well in the tank so an edge digs into the side
the belt that pulls the tank might not be tight enough (always make sure the two weights at both ends of the tank are actually hanging down to put tension on the belt!)
the belt might have come off the axle that drives it (the white plastic above the left end of the tank)
Elin’s GEOF213 class observing lee waves
As you see in the pictures above (or the movie below), there is a lot to observe!
Lee waves (not one, but a whole train!)
Different flow regimes: supercritical shooting down the lee side of the mountain, then a hydraulic jump, and then a normal flow
The reservoir upstream of the mountain that builds up as the mountain is moving
Even after the mountain has stopped, you see waves travelling through the tank and being reflected at the ends
Here is a movie of the lee wave experiment. Feel free to use it in teaching if you like! And let me know if you need the movie in a higher resolution, I am happy to share!
*Yes, this was true at the time of writing. But I am setting up that experiment as we speak. Write. Read. Whatever. Will post movies tomorrow!
A ship that is continuously pulled with a constant force suddenly slows down, stops, and then continues sailing as if nothing ever happened? What’s going on there? We will investigate this in a tank! And in order to see what is going on, we have dyed some of the water pink. Can you spot what is going on?
The phenomenon of “dead water” is probably well known to anyone sailing on strong stratifications, i.e. in regions where there is a shallow fresh or brackish layer on top of a much saltier layer, e.g. the Baltic Sea, the Arctic or some fjords. It has been described as early as 1893 by Fridtjof Nansen, who wrote, sailing in the Arctic: “When caught in dead water Fram appeared to be held back, as if by some mysterious force, and she did not always answer the helm. In calm weather, with a light cargo, Fram was capable of 6 to 7 knots. When in dead water she was unable to make 1.5 knots. We made loops in our course, turned sometimes right around, tried all sorts of antics to get clear of it, but to very little purpose.” (cited in Walker, J.M.; “Farthest North, Dead Water and the Ekman Spiral,” Weather, 46:158, 1991)
When observing the experiment, whether in the movie above or in the lab, the obvious focus is on the ship and the interface between the clear fresh water layer (the upper 5cm in the tank) and the pink salt water layer below. And yes, that’s where a large-amplitude internal wave develops and eats up all the energy that was going into propulsion before! Only when looking at the time lapse of the experiments later did I notice how much more was going on throughout the tank! Check it out here:
The setup for this experiment is discussed here and is based on the super helpful website by Mercier, Vasseur and Dauxois (2009). In the end, we ended up without the belt to reduce friction, and with slightly different layer depths than we had planned, but all in all it works really well!
When setting up the stratification for the Nansen “dead water” demo (that we’ll show later today, and until then I am not allowed to share any videos, sorry!), I went into a meeting after filling in layer 4 (the then lowest). When I came back, I wanted to fill in layer 5 as the new bottom layer. For this experiment we want the bottom four layers to have the same density (so we would actually only have one shallow top layer and then a deep layer below [but we can’t make enough salt water at a time for that layer, so I had to split it into four portions]), and I had mixed it as well as I could. But two things happened: a) my salinity was clearly a little fresher than the previous layer, and b) the water in the tank had warmed up and the new water I was adding with layer 5 was cold tap water. So I accidentally set up the stratification for salt fingering: warm and salty over cold and fresh! Can you spot the darker pink fingers reaching down into the slightly lighter pink water? How cool is this??? I am completely flashed. Salt fingering in a 6 meter long tank! :-D
The main reason why we went to all the trouble of setting up a quasi-continuous stratification to pull our mountain through instead of sticking to the 2 layer system we used before was that we were expecting to see a tilt of the axis of the propagating phase. We did some calculations of the Brunt-Väisälä frequency, that needs to be larger than the product of the length of the obstacle and the speed the obstacle is towed with (and it was, by almost two orders of magnitude!), but happy with that result, we didn’t bother to think through all the theory.
And what happened was what always happens when you just take an equation and stick the numbers in and then go with that: Unfortunately, you realize you should have thought it through more carefully.
Luckily, Thomas chose exactly that time to come pick me up for a coffee (which never happened because he got sucked into all the tank experiment excitement going on), and he suggested that having one mountain might not be enough and that we should go for three sines in a row.
Getting a new mountain underneath an existing stratification is not easy, so we decided to go for the inverse problem and just tow something on the surface rather than at the bottom. And just to be safe we went with almost four wavelengths… And look at what happens!
We are actually not quite sure if the tilting we observed was due to a slightly wobbly pulling of the — let’s use the technical term and go for “thingy”? — or because of us getting the experiment right this time, but in any case it does look really cool, doesn’t it? And I’ll think about the theory some more before doing this with students… ;-)
So how do we teach about the Coriolis force? The following is a shortened version of an article that Pierre de Wet and I wrote when I was still in Bergen, check it out here.
The Coriolis demonstration
A demonstration observing a body on a rotating table from within and from outside the rotating system was run as part of the practical experimentation component of the “Introduction to Oceanography” semester course. Students were in the second year of their Bachelors in meteorology and oceanography at the Geophysical Institute of the University of Bergen, Norway. Similar experiments are run at many universities as part of their oceanography or geophysical fluid dynamics instruction.
Rotating table with a co-rotating video camera (See Figure 1. For simpler and less expensive setups, please refer to “Possible modifications of the activity”)
Screen where images from the camera can be displayed
Solid metal spheres
Ramp to launch the spheres from
Tape to mark positions on the floor
Figure 1A: View of the rotating table. Note the video camera on the scaffolding above the table and the red x (marking the catcher’s position) on the floor in front of the table, diametrically across from where, that very instant, the ball is launched on a ramp. B: Sketch of the rotating table, the mounted (co-rotating) camera, the ramp and the ball on the table. C: Student tracing the curved trajectory of the metal ball on a transparency. On the screen, the experiment is shown as filmed by the co-rotating camera, hence in the rotating frame of reference.
About 45 minutes to one hour per student group. The groups should be sufficiently small so as to ensure active participation of every student. In our small lab space, five has proven to be the upper limit on the number of students per group.
In the demonstration, a metal ball is launched from a ramp on a rotating table (Figure 1A,B). Students simultaneously observe the motion from two vantage points: where they are standing in the room, i.e. outside of the rotating system of the table; and, on a screen that displays the table, as captured by a co-rotating camera mounted above it. They are subsequently asked to:
trace the trajectory seen on the screen on a transparency (Figure 1C),
measure the radius of this drawn trajectory; and
compare the trajectory’s radius to the theorized value.
The latter is calculated from the measured rotation rate of the table and the linear velocity of the ball, determined by launching the ball along a straight line on the floor.
In years prior to 2012, the course had been run along the conventional lines of instruction in an undergraduate physics lab: the students read the instructions, conduct the experiment and write a report.
In 2012, we decided to include an elicit-confront-resolve approach to help students realize and understand the seemingly conflicting observations made from inside versus outside of the rotating system (Figure 2). The three steps we employed are described in detail below.
Figure 2: Positions of the ramp and the ball as observed from above in the non-rotating (top) and rotating (bottom) case. Time progresses from left to right. In the top plots, the position in inert space is shown. From left to right, the current position of the ramp and ball are added with gradually darkening colors. In the bottom plots, the ramp stays in the same position, but the ball moves and the current position is always displayed with the darkest color.
Elicit the lingering misconception
1.a The general function of the “elicit” step
The goal of this first step is to make students aware of their beliefs of what will happen in a given situation, no matter what those beliefs might be. By discussing what students anticipate to observe under different physical conditions before the actual experiment is conducted, the students’ insights are put to the test. Sketching different scenarios (Fan (2015), Ainsworth et al. (2011)) and trying to answer questions before observing experiments are important steps in the learning process since students are usually unaware of their premises and assumptions. These need to be explicated and verbalized before they can be tested, and either be built on, or, if necessary, overcome.
1.b What the “elicit” step means in the context of our experiment
Students have been taught in introductory lectures that in a counter-clockwise rotating system (i.e. in the Northern Hemisphere) a moving object will be deflected to the right. They are also aware that the extent to which the object is deflected depends on its velocity and the rotational speed of the reference frame.
A typical laboratory session would progress as follows: students are asked to observe the path of a ball being launched from the perimeter of the circular, not-yet rotating table by a student standing at a marked position next to the table, the “launch position”. The ball is observed to be rolling radially towards and over the center point of the table, dropping off the table diametrically opposite from the position from which it was launched. So far nothing surprising. A second student – the catcher – is asked to stand at the position where the ball dropped off the table’s edge so as to catch the ball in the non-rotating case. The position is also marked on the floor with insulation tape.
The students are now asked to predict the behavior of the ball once the table is put into slow rotation. At this point, students typically enquire about the direction of rotation and, when assured that “Northern Hemisphere” counter-clockwise rotation is being applied, their default prediction is that the ball will be deflected to the right. When asked whether the catcher should alter their position, the students commonly answer that the catcher should move some arbitrary angle, but typically less than 90 degrees, clockwise around the table. The question of the influence of an increase in the rotational rate of the table on the catcher’s placement is now posed. “Still further clockwise”, is the usual answer. This then leads to the instructor’s asking whether a rotational speed exists at which the student launching the ball, will also be able to catch it him/herself. Ordinarily the students confirm that such a situation is indeed possible.
Confronting the misconception
2.a The general function of the “confront” step
For those cases in which the “elicit” step brought to light assumptions or beliefs that are different from the instructor’s, the “confront” step serves to show the students the discrepancy between what they stated to be true, and what they observe to be true.
2.b What the “confront” step means in the context of our experiment
The students’ predictions are subsequently put to the test by starting with the simple, non-rotating case: the ball is launched and the nominated catcher, positioned diametrically across from the launch position, seizes the ball as it falls off the table’s surface right in front of them. As in the discussion beforehand, the table is then put into rotation at incrementally increasing rates, with the ball being launched from the same position for each of the different rotational speeds. It becomes clear that the catcher need not adjust their position, but can remain standing diametrically opposite to the student launching the ball – the point where the ball drops to the floor. Hence students realize that the movement of the ball relative to the non-rotating laboratory is unaffected by the table’s rotation rate.
This observation appears counterintuitive, since the camera, rotating with the system, shows the curved trajectories the students had expected; circles with radii decreasing as the rotation rate is increased. Furthermore, to add to their confusion, when observed from their positions around the rotating table, the path of the ball on the rotating table appears to show a deflection, too. This is due to the observer’s eye being fooled by focusing on features of the table, e.g. cross hairs drawn on the table’s surface or the bars of the camera scaffold, relative to which the ball does, indeed, follow a curved trajectory. To overcome this latter trickery of the mind, the instructor may ask the students to crouch, diametrically across from the launcher, so that their line of sight is aligned with the table’s surface, i.e. at a zero zenith angle of observation. From this vantage point the ball is observed to indeed be moving in a straight line towards the observer, irrespective of the rate of rotation of the table.
To further cement the concept, the table may again be set into rotation. The launcher and the catcher are now asked to pass the ball to one another by throwing it across the table without it physically making contact with the table’s surface. As expected, the ball moves in a straight line between the launcher and the catcher, who are both observing from an inert frame of reference. However, when viewing the playback of the co-rotating camera, which represents the view from the rotating frame of reference, the trajectory is observed as curved.
Resolving the misconception
3.a The general function of the “resolve” step
Misconceptions that were brought to light during the “elicit” step, and whose discrepancy with observations was made clear during the “confront” step, are finally corrected in the “resolve” step. While this sounds very easy, in practice it is anything but. The final step of the elicit-confront-resolve instructional approach thus presents the opportunity for the instructor to aid students in reflecting upon and reassessing previous knowledge, and for learning to take place.
3.b What the “resolve” step means in the context of our experiment
The instructor should by now be able to point out and dispel any remaining implicit assumptions, making it clear that the discrepant trajectories are undoubtedly the product of viewing the motion from different frames of reference. Despite the students’ observations and their participation in the experiment this is not a given, nor does it happen instantaneously. Oftentimes further, detailed discussion is required. Frequently students have to re-run the experiment themselves in different roles (i.e. as launcher as well as catcher) and explicitly state what they are noticing before they trust their observations.
Possible modifications of the activity:
We used the described activity to introduce the laboratory activity, after which the students had to carry out the exercise and write a report about it. Follow-up experiments that are often conducted usually include rotating water tanks to visualize the effect of the Coriolis force on the large-scale circulation of the ocean or atmosphere, for example on vortices, fronts, ocean gyres, Ekman layers, Rossby waves, the General circulation and many other phenomena (see for example Marshall and Plumb (2007)).
Despite their popularity in geophysical fluid dynamics instruction at the authors’ current and previous institutions, rotating tables might not be readily available everywhere. Good instructions for building a rotating table can, for example, be found on the “weather in a tank” website, where there is also the contact information to a supplier given: http://paoc.mit.edu/labguide/apparatus.html. A less expensive setup can be created from old disk players or even Lazy Susans. In many cases, setting the exact rotation rate is not as important as having a qualitative difference between “fast” and “slow” rotation, which is very easy to realize. In cases where a co-rotating camera is not available, by dipping the ball in either dye or chalk dust (or by simply running a pen in a straight line across the rotating surface), the trajectory in the rotating system can be visualized. The method described in this manuscript is easily adapted to such a setup.
Lastly we suggest using an elicit-confront-resolve approach even when the demonstration is not run on an actual rotating table. Even if the demonstration is only virtually conducted, for example using Urbano & Houghton (2006)’s Coriolis force simulation, the approach is beneficial to increasing conceptual understanding.
When talking to the “general public” (which sometimes just means friends or relatives) about working in climate sciences, it is sometimes really difficult to explain what it is we do every day. I have described a very simple way of explaining how climate models work before. But while this might help provide a general idea of what a model does, it does not show us what climate models actually do. But there is a great tool out there that does exactly that!
The Monash simple climate model is a real climate model. When I was still in Kiel, almost 10 years ago, my sailing buddy Janine was working on implementing the first version of that model! And now the DKRZ (the German Climate Computing Center) hosts an web-based interface that lets anyone access the model.
You can build up the climate model step by step, adding representations of processes like ice albedo, clouds, or many other and then compare model runs including those processes with those runs without. You are even shown the difference between those two runs to see how properties like surface temperatures are affected by the process under investigation! And really awesome feature? The visualization of which processes are switched on and off. See below: On the left, in experiment A, all processes are switched on (and therefore shown in the picture on the top left). In Experiment B, on the right, almost all processes have been switched off, only incoming solar radiation and outgoing radiation are active. Looking at the temperatures below, this shows how Experiment B is only influenced by the sun and temperatures are the same along lines of constant latitude. In Experiment A, though, the temperatures are modified by many more processes, and therefore the distribution is a lot more messy.
Screenshot from http://mscm.dkrz.de, shared under CC BY-NC-SA
You can also look at different climate change scenarios, and you always get to see the CO2 forcing of the respective scenario. You can also compare scenarios with each other (see below). Doing this, you can vary parameters, too, to investigate their impact. You can always look at different model fields like surface and subsurface ocean temperatures, atmospheric temperatures, atmospheric water vapor or snow/ice cover.
Screenshot from http://mscm.dkrz.de, shared under CC BY-NC-SA
There are very nice video tutorials for a quick start, and puzzles where you can test how well you understand the model.
I absolutely love this tool, and I wish I was teaching anything related to ocean and climate so I could use it in my teaching. This opens up so many possibilities for inquiry-based learning. Or basically just interest-driven exploration, which would be so fun to initiate and then support! You should definitely check it out! http://mscm.dkrz.de/
Whenever I’m in a canoe or kayak, I love watching the two eddies that form behind the paddle when you pull it through the water. It looks kinda like this:
Flow around a paddle
Instead of pulling a paddle through more or less stagnant water, we could also use a stationary paddle in a flow. And that is the setup I want to discuss today: A stationary, round paddle perpendicular to an air flow.
A very cool feature of the paddle – which we know has to exist from the sketch above – is shown below: There is a point somewhere downstream of the paddle, where the direction of the air flow changes and a return flow towards the paddle starts. You can see that the threads on the stick I am placing in the return flow go partly towards, partly away from the paddle. So clearly the stick is in the right spot!
Visualizing the flow field behind a paddle with a threaded stick
Another visualization that my dad came up with below: Threads are pulled back towards the paddle in the return flow.
Visualizing the return flow behind a paddle with threads
Doesn’t it look awesome?
Visualizing the return flow behind a paddle with threads
Another way to visualize the change in flow direction is to take a rotor and move it from far downstream of the paddle towards the paddle and back.
Visualizing the change in flow direction by moving a rotor towards and away from a paddle blocking an air stream
All of this is shown in the movie:
Don’t you wish you had all this stuff to play with? :-)
(And do you now understand why I was so excited about the diving duck? :-))