Discuss: Flowchart on drinkable water in the world

March 22 was World Water Day, and TreeHugger posted this graphic on drinking water that is available in the world. The main point is that a very small percentage of water in the world is actually drinkable. It’s definitely a story worth telling, but the graphic doesn’t work at all. Even as a simple presentation of percentages (from UN-Water Statistics), it’s confusing.

How can we improve this graphic to tell the story more clearly? Discuss.

[Thanks, Donald]


  • For starters: I think it would be more clear if the bars were proportional to their percentage, so that you get some visual clues of their proportions…

  • I think it would be clearer if the percentage on each level would relate to the total water amount on earth. The numbers are jumping up and down, the bars are constantly getting smaller.

    • @David: I definitely agree that it’s only confusing that the percentage points jump up and down, especially since there are two “which leaves…”-bars that should be put into relation to their “parent”.

      I think one of the main problems of such comparisons is that the ratio “a-to-b” (when a+b=1) is extremely small/large. Graphically 1% vs 99% looks almost the same as 0.1% to 99.9% and so on. I think this is why the UN-Water Stats graphs don’t make an impression.

      Personally I like to use metaphors when trying to show differences between very large and very small numbers. It’s easier to see the differences between a dinosaur, a cat and an ant than 100, 10 and 1.

      Water on earth = keg of beer (10 liters)
      2.5% fresh water = half-pint of beer (2.5 dl)
      30% access = 4 shots of beer (7.5 cl)
      1% human use = the stale beer at the bottom of your glass you never bother to finish (0.75 ml)
      8% domestic use = small droplet of beer (60 nl)

  • Is that final 0.08 meant to be a percentage?

    • You’re right to question it. The final 0.08% figure is representing ‘the total fresh water readily accessible to humans’ less 70% for irrigation and 22% for industry = 8%. So 0.08 would be the decimal representation of this just as a number meaning 8% should be the displayed value. Massively flawed graphic.

      • Dude right about that it says it’s drawing from 1%, so 0.08% is right — the fraction is 0.0008.

  • Why not go for a “Square Pie” approach but this time not having 100 individual squares but 10 000 individual squares (a 100×100 matrix). This should keep a better sens of proportion and still have 8 square for domestic use.

    On top of this, I would rework the color palette in order no to jum between two colors that don’t convey information.

    I’ll have a go at this in the coming days and post-it on my blog.

    • Good idea, but you’d need a 1000×1000 matrix…

      0.025 * 0.3 * 0.01 * 0.08 = 6e-6

      not so sure it’d be easy to tell the difference between the 6 dots of drinking water and the 75 dots of accessible fresh water, but surprise me ! ;)

      • I agree with you, I would therefore break it into 2 square pies. The first to present Fresh water prorpotion against total water and then indicating a zoom to plot the value for domestic use water.

  • It’s essentially a 1000-unit beginning and a 1-unit end (0.08% being a little over 1/1000th of 75%). So starting with 1000 droplets zoomed out, and then zooming in a few times as the percentage drops. Color-code the droplets, of course.

    Another way of doing it would be ‘the size of Spain’ or ‘the size of Colorado’ at each stage. That would give an accurate idea of both the relative scale and the real-life volume, plus it would be graphically interesting.

  • Wired Science did a great job with similar data – zooming stacked cubes. Making it interactive would make it even cooler.

    • Oh, yes. Problem solved! Good reference. Thanks.

    • Zoomed cubes solves the problem of ever shrinking areas, however, it doesn’t allow for comparison of elements across different zooming levels. This means you cannot compare the amount of water in permafrost to the amount of water in oceans, for example.

    • Also problematic with the Wired site graphic is that it implies that only surface water is used for drinking. For those of us on private well water, who are also concerned about the impacts that natural gas drilling will have on groundwater in our region (Central NY, but also Pennsylvania, WV, and Ohio, not to mention what’s already going on in Wyoming and Colorado), this is a huge oversight. Otherwise, this could be a helpful graphic for us.

    • it’s from the United Nations Environment Program, actually (fig 4.1):


  • The title itself is somewhat misleading… The “world’s water content” should also includes a large portion of atmospheric and ground water, plus a not insignificant portion stored in biological organisms.

    I’d say rainwater is a significant source of drinking water, and should also be included… A large number of cities also derive their drinking water from de-salination processes, which means that a (very) small percentage of sea water is converted to drinkable and usable water.

    Also, the freshwater breakdown appears to ignores lakes and rivers, with 70% freshwater ice (glaciers or continental ice shelf) and 30% ground water…

    In any case, the only stats that are needed to get the point across are that 2.5% of “all” water is freshwater, 0.75% of all water is groundwater, 0.0075% of all water on earth is suitable for use, leaving 0.0006% of all the water on earth for domestic use (is that true?).

    I agree with the suggestion of using a square matrix.

  • I take most of these numbers at face value, but I’m dubious about the explanation of the 1% of the 30% of groundwater that is available for ‘direct use’. The labeling implies that 99% of groundwater is “so polluted that it is unusable.” I certainly believe that only 1% may be readily accessible and directly usable (i.e. can be drawn from a well and drunk without treatment) and that may be a good measure of ‘domestic use’ in large parts of the world. But in the U.S., for instance, almost all water that is used for domestic use is treated. Which portion of the 30% does that come from?

    I also agree that water in lakes and rivers doesn’t show up here. I assume it is groundwater, and maybe a small portion. River water is accessible and treatable but you only use a small portion of its flow. Since it is a flow, not static, I think it complicates the picture somewhat (although rivers may be a tiny portion of the whole).

    As in many cases, the context is at least as important as the numbers.

  • National Geogaphic’s latest issue is on water. They have a different take on how to break this down >> http://environment.nationalgeographic.com/environment/freshwater/freshwater-101-interactive/

    • here’s the full print from Timm Kekeritz for Virtual Water: http://virtualwater.eu/

    • Thanks, VK

      The National Geographic bit is nicely done, and tells a different story:
      – they also say 97.5% is salt water, 2.5% fresh (agree)
      – of that 2.5%, 60% is ice, 10% is surface water, like lakes and rivers, and 30% is groundwater (but some too deep to get at) (why does this graphic lump surface water–a major source of the water we use–with ice?)
      – 70% agricultural, 22% industrial, 8% “pumped into cities for drinking, washing and watering” (how does 8% turn into 0.08%? some earlier commenters noticed it, but that’s a two orders of magnitude error in the punch line!)

      National Geographic doesn’t attempt to say how much of the 40% of unfrozen fresh water is used for agriculture, industry and domestic purposes, so the 1% in the graphic comes from somewhere else.

      So partly it looks like there are two loosely connected stories here: one about where water is distributed on the planet, and the other about human use of water. Each story is clear by itself, but I don’t understand what I’m supposed to get from the connection.

  • Marc Hendel April 1, 2010 at 10:05 am

    A takeoff on the “what if the earth had 100 people?” What if the earth had only 100 gallons of water? If this is the case:

    2.5 gallons are fresh water
    1.75 gallons are in frozen form
    0.75 gallons are in water form (3 qts or 32 oz – the size of a Big Gulp at QT)
    31.68 oz are polluted
    0.32 oz are usable (~1 tsp or ~ 188 drops)
    irrigation uses 131 drops
    industry uses 41 drops
    which leaves 16 drops for domestic use

    As a side note, I am not sure why the frozen form of the fresh water is being ignored. I have read stories of companies “harvesting” water from icebergs…

    • That’s a good way of doing it. Easy to conceptualize amounts. People know what a table spoon represents… I would increase the original amount so that the amount available for domestic use winds up being measurable in tablespoons, that is much easier to understand than drops.

    • Jupiter Jones April 1, 2010 at 10:19 am

      Love it!

      I’d maybe give some reference as to how big 100 gallons is, so the viewer could better conceptualize the 16 drops in relation to it. (And for those in metric lands who might not know what a gallon is in the first place.) What is it, like, two full bathtubs or something?


  • ocean water is drinkable if processed, the entire premise is flawed. does anyone really believe that the worlds navys still carry around huge caches of freshwater for consumption? no they desalinate ocean water for usage. its fine for irrigation as well. the bulk of israels freshwater comes from multiple desalination plants as well.

  • There are all kinds of problems, but the main one is that the percentages keep changing meaning. First a proportion of the Earth’s surface area, then two mutually exclusive and collectively exhaustive proportions of the total volume of water, then three not-mutually-exclusive proportions of all fresh water, then two mutually exclusive but not-collectively-exhaustive proportions of all potable water, then one proportion of all fresh water(!).

    All leading to the conclusion that a small proportion of all fresh water is available for domestic use – so why start with all water on Earth?

    Of course, fixing the graphic wouldn’t cause the premise to make sense. Implicitly, it says that the amount of currently potable water is what matters, but both humans and mother nature create potable water from non-potable water all the time. So water supply for human domestic use is not a stock, it’s a flow.

  • Hi Nathan,

    Thank you so much for taking note of our graphic and inviting your community to offer feedback. The comments here so far are great and will be useful for us on future graphics. I would love to work with you to build a new graphic to build on the thoughts shared here. Please email me if you’d be interested in collaborating. We would update our post with the new graphic and tell the story about how you and your readers helped us in that process.

    I’m also working on a post to put on TreeHugger that will point to yours and invite additional comments on the graphic.

    Chris Tackett
    Social Media Editor

  • I think it is OK but the font/ colours are wrong.

  • Maybe someone already said this but the way to do this is as follows:

    You start with some circle (or square, etc.) whose area represents the total amount of water, and you color the fraction that is fresh. Then you draw lines guiding the eye from the small colored spot to another large circle (etc.) which now represents the fresh water, and you color the fraction that is groundwater. And then repeat.

    I’ve seen this done before, and I think it’s as good as any other method I’ve seen for helping people visualize repeatedly taking small fractions.

  • Urban Mechanic April 16, 2010 at 2:13 pm

    I think this would be a great data set to utilized small multiples to explain. Perhaps a hundred small water drops (colored according to breakdown of fresh, frozen, polluted, and etc.) with a single water drop out of the hundred blown-up and subdivided further to explain the finite break down.