Introduction: Predict Weather With a Cup of Coffee

Picture of Predict Weather With a Cup of Coffee

I learned this trick reading Backpacker Magazine years ago while waiting for an air taxi flight into the Nahanni River. It really works.

Make a cup of coffee. I like to add cream and sugar since it makes the bubbles easier to see but black coffee is fine. Hot tea or hot cocoa will work too.

Step 1: Watch the Bubbles

Picture of Watch the Bubbles

When you pour the coffee into your cup, watch the bubbles.

If the bubbles move to the edge of the cup rather quickly, that's a good sign. Expect clear skies for the next 12 hours.

If the bubbles hang around in the center of the cup, get out your rain gear. You can expect rain in 12 hours.

If the bubbles slowly move to the edge of the cup, you may get a bit of weather, but it should be clearing in a few hours.

If you've managed to make a cup without bubbles, flop a spoonful of coffee back into your cup and make some more bubbles.

Step 2: Theory

Picture of Theory

The theory behind this trick is that high pressure will push the bubbles to the edge, and high pressure indicates a period of sunny, calm weather. Low pressure won't move the bubbles and low pressure systems typically bring unsettled weather.


nivyawsome (author)2015-11-11

wow this amazing.

patatarium (author)2015-03-11

Amazing I gonna try it !

kelsey.bailey.391 (author)2014-10-10

My mom told me about this about a month ago. The few times I done it, it was right. Crazy how such a little everyday things works :)

McClay14 (author)2014-09-11

(removed by author or community request)

stickmop (author)2013-01-15

I don't think I want to know what gets predicted when a frog jumps in your coffee. Best just run for the hills. :-)

Brian H (author)2011-01-16

Adams had been to The End of Time Restaurant, so he naturally knew all.

P.S. The coffee trick works best if you can get exactly 42 bubbles ...

fazgard (author)Brian H2012-03-28

I wonder what the improbability factor of getting exactly 42 bubbles would be?

Brian H (author)fazgard2012-03-28

Infinity, of course, give or take 42,000,000,000,000,000,000,000.

finton (author)Brian H2012-12-26

Which, of course is 42 x 10^(42/2). The chances of this occurring randomly is, well, pretty Improbable - as a rough Guide.

mikeltv1 (author)2012-04-09

How exactly does this work? I drink tea every day and I never noticed this cool trick. I should try it.

stickmop (author)mikeltv12012-04-09

See Step 2. Brew and brood about it; perhaps you can raise a thunderboomer.

aidanjarosgrilli (author)2012-04-04

I've been using this method of predicting the weather for about two months now, and it hasn't failed oncen
It works 100% :)


On the other hand, I've had a couple of failures since I moved close to the shore of Lake Superior a couple of years ago. Seems like sometimes the predicted rain stays out over the lake and doesn't make it to shore.

aristide202 (author)2012-03-30

Is there any best temperature of coffee or temperature difference between coffee and air around to perform the observation^
You're supposed not to stir your coffee, is it ? Just leave it still and observe the phenomena. I don't catch the theory of atmosphere pressure pushing significantly
on coffee surface, Ok it does but that's sounds like atmosphere blowing more or less in the centre of the cup and why not the reverse , I mean some kind of blow from side to center or the like. What about influence of air % humidity?

stickmop (author)aristide2022012-03-30

Temperature differences haven't affected my observations I don't think.

Stirring tends to either move the bubbles to the outside edge or congregate them in the center vortex, so that's a no-no. You do run into problems when the coffee doesn't cooperate and you have to use your spoon to make bubbles. A dip and splash technique usually avoids excessive manual bubble movements.

Barometric pressure may be affected by humidity, so that could affect the bubble reading. To avoid that you could dispense the coffee into a clear wet bulb I suppose. Don't know, just slinging out ideas, but some of this stuff can make you go psychro (sic) on the meter scale.

killerjackalope (author)2012-03-30

Saw this a few days ago - we've been having an unusual heat spell due to a high pressure front that's moving away.

So far the predictions of the cup have been ringing true, I drink a lot of coffee...

The weather report agrees too, so it may be wrong yet, it's not right often...

Browncoat (author)2009-09-01

Neat!! I assume this would work with about any beverage...?

Browncoat (author)Browncoat2009-09-01

Just watched the bubbles in my wine & compared it to They matched! :)

gestault (author)Browncoat2012-03-29

I tried this experiment with Vodka but always needed a lot of data so, after a while into my lab time, I didn't care what the weather was...

doesnt work with jerger mister

LOL doesn't work if your are passed out the next day either.

Brian H (author)Browncoat2012-03-28

Wine is unreliable. Milk always predicts good weather, no matter if a hurricane is about to rip your house up by the roots.

Brian H (author)Browncoat2012-03-28

No. The unique hydro-pneumatic nature of coffee bubbles means that it far out-prognosticates tea, beer, or pop.

stickmop (author)2012-03-28

Let's get together and test the theory some more. I'll bring the coffee if you bring some of your key lime pie. :-)

stickmop (author)2012-03-28

LOL, I thing they have their bubbles crossed. Their prediction for the 2011 winter seems to have been off quite a bit too. I'm in the UP of Michigan and we had temps 40 degrees above normal. (AOL). Can't wait to see what the Old Farmer's Almanac comes up with. :-)

whiteoakart (author)stickmop2012-03-28

The Farmer's Almanac for 2011/2012 Winter season wrote the following for our Michigan and Great Lakes region:
Expect colder than average temperatures and above average snowfall.

I would have a hard time making up a more inaccurate forecast.

fazgard (author)2012-03-28

Time to write a quick mobile app where you could take a quick video of your coffee bubble movements in the morning and have it produce a forecast - then have it compare it to the actually forecast - then put that into a database that tracks multiple users coffee bubbles forecasts .vs. actual forecasts based and then ....
- Oh wait, I've removed the simplicity of it.... lol

dmt (author)2010-01-25

presumably this would work with other opaque beverages as well, yes?

stickmop (author)dmt2010-01-25

I can only vouch for coffee, since that's the beverage I usually imbibe every day at home or in the woods and I'm used to how the bubbles flow in my cuppa.

Different fluids may not react the same way. They may have different miscibility properties, different triple points, different electrical conductivity, different surface tensions - you get the idea. All those factors could change the capillary wave dynamics and throw your readings off by hours.

Isn't it amazing how much pseudo-science you can dig up in 5 minutes on wiki? :-)

fazgard (author)stickmop2012-03-28

YOu've brewed up a fine batch of it!

mewat (author)2010-08-13

I have a better system: raindrops on the surface of coffee: it is raining; waves on the surface: windy; reflexes on the surface: sunny; ice on the surface: very cold: you cannot see the cup: fog (or someone has stolen the cup)

stickmop (author)mewat2010-08-13

Ah, but the bubble technique will predict weather that's coming in 12 hours, so that you'll know whether to grab your umbrella, kite, sunglasses, ice skates or flashlight. Fewer surprises.

mewat (author)stickmop2010-08-28

Prediction are uncertain! My system tells the actual weather without errors

Musicman41 (author)mewat2012-03-28

Your system fails indoors.

dbombere (author)2012-03-28

Thanks for this! I love coffee, and I love bubbles!

david_tv (author)2012-03-25

Muchas gracias por compartir la información. Lo que sí me queda claro, es que me antojaron a tomar una taza de café. ¡Saludos!

stickmop (author)david_tv2012-03-26

Deben trabajar con una taza de Sanborns. ¡ Disfruta!

Brian H (author)2010-12-29

Pressure, huh? Someone should invent a small device to measure that; maybe they could call it, oh, let's see ... a "barometer" (pressure measurer).

stickmop (author)Brian H2011-01-16

It sounds like that's coming soon with tablet and smartphones. Motorola is putting a barometer in their Xoom so they can figure out what floor of a building you are on, in conjunction with the GPS. And soon someone will create an app and we'll be making regional forecasts without the need for the NWS.

But a cuppa is sometimes handier than your gizmo.

And Cheaper

amicus curiae (author)2010-12-30

I had a dream there were clouds in my coffee, clouds in my coffee etc :-)
Carly simon was ahead of us all:-)

Brian H (author)2010-12-29

Doesn't seem to work when I make bubbles in the toilet bowl, though. Too big ... ??


nutsandbolts_64 (author)2010-07-29

IT WORKS (of course). 6 am in the morning *pours some good stuff into the cup* lets see.... bubbles moving slowly to the side.... according to the instructable, when it does this, it means some rain will fall in about 12 hours... (11 hours later) IT'S RAINING lol

jolshefsky (author)2009-08-26

I'd love to see some empirical analysis of this. Even the notion that the speed of the bubbles is proportional to atmospheric pressure is pretty cool.

stickmop (author)jolshefsky2009-08-26

Good idea - let's go find a grant application. I bet NOAA would fund us. :-)

ARVash (author)stickmop2010-03-01

 a simple pressure chamber would suffice. 

beehard44 (author)ARVash2010-07-27

next thing you know, morning ritual in NOAA, everyone, field scientists, computer technician, meteorologist are all watching their cups.....

lol, I wonder if other stuff works....

killerjackalope (author)2009-08-27

This doesn't work with the coffee I have at the moment, it has the unique capacity to never have bubbles

It must be a good coffee then.

About This Instructable




Bio: Mops from sticks and rags. Cheap!
More by stickmop:Patching Holes in Sneaker ToesCurtain Cleats --- Cord Management 101Q&D Caster Cups
Add instructable to: