My Dishpan Climate Model

About two years ago, I discovered the concept of “dishpan climate models”, through Iain Stewart’s Climate Wars documentary. The experiment is pretty simple: a large bowl filled with water (representing one hemisphere of the Earth) with a block of ice in the middle (a polar region) rotates on a turntable with a Bunsen Burner (the Sun) heating it from one side. By injecting some dye into the water, you can see regular currents from heat transport and the Coriolis effect. Spencer Weart dug up some fascinating results from the days when dishpan climate models were the only sort available: researchers were able to simulate the Hadley circulation, Rossby waves, and the Gulf Stream.

I wanted to try this out for myself. Iain Stewart had made it look easy enough, and he got some really neat currents flowing. So one Saturday afternoon a friend and I got to work in my kitchen.

We started by figuring out how to rotate the bowl. My family doesn’t own a record player, so we couldn’t use that as a turntable. We tried to rig something up out of an old toy helicopter motor, but it wasn’t strong enough. Eventually we settled for a Lazy Susan which we spun by hand. It wasn’t a constant rotation, but it would have to do.

Then Antarctica, which consisted of a handful of ice cubes, kept floating away from the centre of the bowl. Soon the ice cubes melted and there were none left in the freezer. We filled a Ziploc bag with frozen corn, which wasn’t quite as buoyant, and used that for Antarctica instead.

Unsurprisingly, there was no Bunsen burner in my kitchen cupboard, so the Sun was represented by a paraffin candle that sort of smelled like cinnamon.

The only serious problem remaining was the dye. Every kind of dye we tried – food colouring, milk, food colouring mixed with milk – would completely homogenize with the water after just a few rotations, so all the currents were invisible.

The only liquid in my kitchen that wouldn’t mix with water was vegetable oil, so we dyed some of it blue and poured it in. This was a really really bad idea. The oil seemed to be attracted to the plastic bag keeping Antarctica together, so it all washed up onto the continent like some kind of awful petroleum spill in the Antarctic Ocean.

At that point, our climate model looked like this:

I would like to try this again some day, perhaps when I have access to a better laboratory than my kitchen. Any ideas for improvement (besides the obvious)? In particular, what kind of dye does work, and how does Antarctica stay together without being encased in plastic?


10 thoughts on “My Dishpan Climate Model

  1. You don’t want the liquids to mix??
    Surely that would be a poor model. Isn’t the idea to track the different temperatures in the one liquid? So maybe you need one liquid that changes colour depending on temperature.

    I want them to mix enough so that they flow in the same manner, but not so much that they’re completely homogeneous and I can’t see what currents are going on. It needs to be sort of cloudy and partially mixed. -Kate

  2. As a theoretical physicist subject to the Pauli effect, I’m probably not the best one to give advice. But you might try reading the original dishpan experiment papers, for example Fultz (1949).

    With respect to tracer injection, Fultz used organic dyes, including “a solution of carmine red in ammonia” (better for film photography) and “a water solution of crystal violet” (better for visual inspection). Instead of dye he also tried injecting tracer pellets of a synthetic resin (Velsicol) after adjusting the salinity of the water so the densities matched.

  3. This is very clever and I’m sharing it with my followers. I haven’t been to your site for a while. Looks like you have been busy.

    OBXCommonGround has a wider perspective (wildlife to climate change to other environmental issues). I publish my newsletter OBXCommonGood on Wednesdays so I’m running off. I’m always looking for credibility, authority, who to believe info.



  4. I’ve been too swamped with my thesis work to contribute much here for a while, but when I see improvisation For Science, I just can’t resist.

    You could try freezing a chopstick in Antarctica (label it South Pole for perfect grade-school science fair chic) and weighting one end of it as an anchor. If the dish is deep enough in the center it shouldn’t affect circulation too much. If your model requires sea ice (i.e. floating, as opposed to simply a temperature gradient – I’m guessing it doesn’t need much of it since you’re calling it Antarctica) you’ll have to find a different trick, since I don’t think this can be done without either increasing the effective weight of the ice or supporting it in the middle of the pan.

    If you need an easy source of constant rotation for your model, depending on the style of lazy susan you’re using (if it has a central pillar as opposed to multiple castors), you can use bicycle chains or pulleys to gear it down without too much of a hassle. A common – and surprisingly effective – method for this that you might have access to would be the good old fashioned LEGO gear train; this paper provides illustrations of a 243:1 gear ratio train. (My lab uses one that’s …quite a bit higher, I forget exactly how much, as a demonstration of gearing – one of those mini-motors can overpower an athletic human’s grip if it’s applied right.) You’ll lose quite a bit of speed doing this, but it will let your helicopter motor turn the lazy susan with ease.

    How effective was the candle? I can think of a few alternatives, but they’ll probably be less efficient or introduce airflow to the system.

    I can’t help much with the dye, sorry.

  5. Very impressive!

    There are a few things I would recommend that might help make your next experiment a bigger success.

    1. If you haven’t already done so, you may want to take a course in similitude to get the proportions right.

    2. For Antartica, how about a big block of ice that rests firmly on the bottom of the container. Fill a bucket with water and put it into your freezer until solid. Pick the bucket size for the ice block to fit your experiment. When ready or before the experiment, run water over the bucket and knock out the ice block. You can put the block back in the freezer until needed. You can also make several blocks of ice for additional experiments.

    3. You should not turn the platform by hand. That is too jerky and difficult to maintain the correct constant rotational speed. The container’s rotational speed is not arbitrary. It must be determined by similitude. Perhaps you could visit a hobby shop and get an inexpensive variable speed motor with speed reducing gears and a rubber wheel to drive your platform.

    4. The diameter of your circular water container relative to the diameter of ice block relative to the depth of water may be critical to establishing representative currents. Similitude should indicate the correct proportions.

    5. I believe that heat from a candle is too weak. How about an infrared or regular light bulb with reflector whose wattage would be determined by similitude.

    6. Injecting dye into your ocean can be a tricky issue, but there may be a way. You could design a continuous dye injection system. Here is an idea that I have no idea if it will work or not. How about something like an intravenous drip dye injection system. You connect a very fine hypodermic needle to a tube that feeds dye to it from a container. The tip of the needle is suspended just below the water surface, so that a tiny stream of dye flows steadily into the water just below the surface, leaving a thin trace of dye in the water as it passes by. You have two possibilities: The needle may be held stationary relative to the table, or it may be held stationary relative to the rotating container. In the first case you would see a circle of dye forming as the container turns. If the water is moving radially, you will see the circle’s radius changing as the container turns. If the water is moving circumferentially, this case will not readily show that motion, In the second case, if the water is not moving relative to the rotating container, the dye will just expand radially from the needle and form a perfect circular blob. If the water is moving radially, the dye will migrate radially from the needle, forming a radial line. If the water is moving circumferentially, the dye will move circumferentially from the needle, forming an arc. And if the water is moving both radially and circumferentially, the dye will indicate both components in the form of a spiral.

    Now I have just one non technical question.What does your mom think about you budding young scientists messing up her kitchen? :o)

    • @roger
      Her Mom laughed out loud describing it to me – and then told me I HAD to read the comments because you guys were SO funny egging her on with such good ideas.
      So who is going to do the math for the similitude and figure out the size of the bowl, the speed of rotation and how much ice for Antartica?
      Too much fun.

  6. Fultz’s legacy is revived at the University of Chicago , by Dr Noboru Nakamura.

    Also, Dr Satoshi Sakai of Kyoto University has developed a series of teaching materials on the web at . On its “index of experiments” page, Fultz-type experiment has a title “Riding on the jet stream” which corresponds to “baroclinic instability”.

  7. Ice cream for Antarctica, melted ice cream for the oceans, chocolate sauce for the dye. If it doesn’t work, dip it in liquid nitrogen and eat.

  8. Wow, I actually worked with various versions of these for several years.

    – Unfortunately the most important aspect is the hardest: you really need a constant angular velocity. Fortunately for me this problem always came solved. :-) The suggestions above are a good start. The tank also has to be level to avoid sloshing and ideally centered to keep a symmetrical domain. Sloshing is probably what is diffusing your dye quickly. A related issue is that before putting the dye in you should let the water spin up so that it’s mostly still in the tank’s frame of reference.
    – For Antarctica a cylindrical container of ice water (for the high heat capacity) should suffice. You will probably have to weigh it down. Outside heat should be superfluous, and a candle would never cut it anyway — for that you would need an immersion heater and a way to make it heat evenly around the edge. Ambient temp should be enough.
    – Food coloring should work, and for a basic setup without any flow sources or sinks, just dropping it in should work. If density differences between the dye and the water become an issue then you’re making progress! In that case let the water come to room temp along with the dye, thin the dye with water before using, possibly mix it with alcohol to compensate for the denser dye. Ideally you should be able to syringe a drop of dye into the middle of a room-temp water column and have it stay there.
    – As for scaling, a 60cm diameter tank with 25cm deep water (from memory) worked for me. Rotation rate is important.

    Here’s a contemporary source:

    Good luck and keep us updated!

  9. Kate, try a shallow metal pan if you can find one, and a plumbers propane torch like the ones manufactured by Bernzomatic. A plumbers torch should be relatively low cost. Another option would be to get a job as an assistant in your schools chemistry lab. That might allow you to get your hands on some equipment that you would be able to use for free if you can get the O.K. of a sympathetic professor. Is it possible that your school might already have all the equipement to perform this experiment?

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.