Radiation Dosage Chart

Tuesday, March 29th, 2011

Sorry this is so late. I just came back from holiday to an inbox full of requests to render up a Radiation Dosage chart.

Personally, I think XKCD has done a great job already. But I am here to serve.

Hopefully our chart will also help counter some of the confusion circulating about radiation and peril. Here also is a great piece by Anil Dash on understanding radiation exposure.

The data is here:

If you like this image, we’re selling a hi-res, instantly downloadable PDF.
It’s $2.50 / £1.50 / €1.70 with all money going to Japan Crisis Relief.

If you’re interested in more charitable artwork, please try DesignersForJapan – some great work there.

Buy in our Store or use this direct PayPal purchase link:
(You’ll be instantly emailed the PDF). Thank you!

Books and Store

Our Beautiful Books - Information is Beautiful Information is Beautiful Store

Show Comments ( )

  • Ethan Jewett

    I’m not sure how this helps.

    You have here a logarithmic scale that is in no way obvious as a logarithmic scale. You have two huge indicators (color and width) and one smaller indicator (size of the typeface used for the numbers) pushing viewers to treat it as a linear scale and read a 100,000 mSv dose as 100% more than a perfectly safe 1 mSv dose. It is really 100,000% more.

    You can only figure this out by reading the numbers, and really only by reading the tiny numbers on the left. But the whole point of a visualization is it makes the numbers visual, and this one doesn’t do that accurately.

    • Dr. ed Martin

      YEs, keep the “old” chart because the new one is wrong.

  • Jameson Toole

    A beautiful visualization to be sure, but I am worried its a bit misleading. The smallest dose you show, .1 micro-sieverts is a full 9 orders of magnitude smaller than the largest dose, (100 sieverts). Displaying the type of linear growth without being able to compare the smallest dose to the largest could mislead people into thinking that the doses we routinely experience are not so far from the elevated levels that cause serious illness.

  • typo police

    “lifetime risk cancer”?

  • Egg Syntax

    Blargh! David, I generally love your work, but I have a big problem with this. The same people who are already confused about the dangers of radiation are going to have no idea how a log scale works. Furthermore, you only clarify that it’s logarithmic in the fine print. The people who most need to be better informed will look at this and think that the immediate-coma-and-death dosage is about twice what you get from a CT scan. I urge you to take this down and rework it with a linear scale (as challenging as that is given the large range of dosages represented) or else somehow make it much, much clearer what a logarithmic scale means, and that this chart uses one.

  • John

    Its pretty, but the triangle’s scale seems a little misleading.

    For example, the width of the bottom of the triangle at 10 usv width is about half the width of the triangle at 1000 usv, and about 20% of the width of the triangle at the 100,000 msv, which is what, a hundred million times bigger than 10 usv?

    instead of being super tall, this graph should be super SUPER super wide to show the extremity of that 100k msv dose. without labels, it make it look like the bottom dose is only 5x worse than background radiation…

  • teko

    Well done, much easier to see the scale than XKCD’s

  • Chris Pudney

    Along with your and Randall Munroe’s radiation charts I’ve also come across this Nuclear Radiation Chart, which also visualizes dose rate info. I’m compiling a list of these radiation charts, so please let me know if you find others.

  • Tomas

    Hi David,

    Definitely one of the best so far, but i see a big problem which is it’s still very hard to compare values. e.g. Hourly dose, One day dosage, one year. One day at two sites 50 km from Fukushima is equivilant to ~650 times your average daily backround dosage. A month at the same two sites equals “increased lifetime risk is evident”.

    Makes it easier to compare impacts.

  • Dan

    How fast would I have to eat those 100,000 bananas to die of radiation poisoning?

  • Romanian reader

    You might be interested in this Mediafax is a Romanian news agency and it seems that they skipped the part with naming their sources for the visualization, even after the original source is named in the comments. Maybe they thought that would be all.

    On a different note, thank you for your work and for your site, you are doing a great job!

  • sarah aubry

    Gorgeous. You are inspiring me to become an information designer.

  • Daniel Eriksson

    The log-scale is a bit confusing. Anyway, it’s a clean and powerful visualization.

  • Jesse

    I just wanted to point out that the URL on the graphic is slightly incorrect. returns a 404, however returns the proper spreadsheet.

    Also maybe I missed it, but what’s the source for the info in the table?

    sweet graphic!

  • Jesús R.

    If 500 mSv implies “lifetime increase in cancer risk: 1 in 250″, I think it would be 1/250 x 100 = 0,4% increase in cancer risk. Am I wrong? Considering that average cancer risk is about 20%, a 0.4% increase doesn’t sound much alarming… With 1,000 mSv it would be 0,8%, so I guess I’m doing something wrong, but I don’t know what… Could somebody enlighten me? Thanks

    • Geoff Russell

      Average lifetime cancer risk is about 40% (not 20%), although it certainly varies a lot between places. If you have 250 people then about 0.4×250=100 will get cancer during their lifetime. But good treatment may see them die of something else. If those 250 people are given a 500 mSv radiation dose then about 101 will get cancer. So no, getting 500 mSv shouldn’t be alarming. If you smoke, drink or eat red and processed meat, they are much bigger risk factors than doses of that order. There are places on the planet where people get 100-200 mSv every year. There are no obvious high cancer rates in such areas … although low populations make the solid epidemiology difficult … e.g., Ramsar in Iran.

  • Alex

    This is a very clean visualization, unlike Munroe’s. On the other hand, it’s the handling of scale that made Munroe’s visualization so effective and exceptional. I had to figure out, through the numbers, that the scale on your chart was probably logarithmic.

    Your visualization is pretty, I grant that. One the other hand, it’s not a visualization. There’s nothing in it that couldn’t be shown through numbers alone. There’s no “aha!” moment. Even worse, if there’s any insight it’s probably based on seeing the scale as linear, and then it’s faulty.

    When I watched Munroe’s chart I was blown away, because it really set lower radiation doses in context. Yes, it’s easy to plot them on a logarithmic scale, but while many people “get” logarithmic scales mathematically they do not understand them well enough to be able to correctly compare sizes.

  • Nick

    Pardon me for joining the choir, but I also agree that this chart presents a log scale in a way that appears to be a linear one. It makes the range look relatively tiny.

  • Penny J

    Nice use of colours moving from safe to fatal. Please would you change the final 1000 SV into 1000 Sv, at the bottom of the chart and note the correct use of spaces between the number and the unit – the latter also applies to 50 km and 10 min. Also, at the top of the chart, it should be micro-sievert not capitalized; and maybe you should explain somewhere that the rem is an old unit (no longer used in developed countries).

    Jesus R has a good point – the usual attributable lifetime risk is 5 % per sievert which would make 1 in 20 for 1000 mSv not 1 in 125. I would willing buy your poster if you make all these changes!

  • Paul Merrill

    For all the space faring enthusiasts it would be interesting to add a line to include an orbit around the earth, a trip to the moon and a trip to mars. I recon a return trip to mars may be a scary number.

  • James C

    It may be useful to also insert additional lifetime risk of cancer for 1 in 100,000 and 1 in 1,000,000 as these are common policy settings for defining acceptable levels of risk.

  • Shaik

    What might be more illustrative is to have the whole chart be in microsieverts. The range of the numbers is much more apparent when you’re comparing a 10(daily dose) to a 10,000,000(fatal dose)

  • Hans from Germany

    Hello David,

    I carefully train my students to use CORRECT measurement units.
    And so they know, that a very important distinction is drawn between
    – dose (accumulated radiation) and
    – dose rate (dose/time = accumulated radiation divided by time of exposure).
    Checking newspapers and articles in the days of the Fukushima disaster
    they very often found a complete mixup of these two measurement units.

    And when I presented your “Radiation Dosage Chart” they at once noticed,
    that the second line of your chart ” 0.4 Natural radiation in the human body”
    describes neither a “dose” nor a “dose rate” … and they were a bit “angry”.
    I promised to contact you and to ask you to correct this.

    Furthermore I agree to the following two comments:

    “I also agree that this chart presents a log scale in a way that appears to be
    a linear one. It makes the range look relatively tiny.”

    “What might be more illustrative is to have the whole chart be in microsieverts.
    The range of the numbers is much more apparent when you’re comparing a
    10(daily dose) to a 10,000,000(fatal dose).”

    In spite of all that your chart gives a nice overview over radiation injuries !

    Kind regards
    Hans from Germany

  • Billy

    They removed the cancer therapy dosage 20,000 sv. from the graphic.
    Probably threatened by the cancer mafia.

    • miriam

      Hi Billy,

      We temporarily deleted it as we’re trying to refine and revise the graph so that it conveys the difference between an acute dose and a longer dose over time – which we feel the image doesn’t do right now. We’ve got a new version in the pipe and will upload soon!

  • Stefan Lasiewski

    You mention “background dose received by an average person on an average day”, but don’t explain what background dose for an average person means.

    It would be useful to include a reference natural sources of radiation like Radon, which is the leading source of radiation exposure for most people in North America according to the “How are people exposed to radiation?” at (Also see the pie chart),

    And even though these radiation sources are small, it still has a significant impact according to the EPA:

    Radon is estimated to cause about 21,000 lung cancer deaths per year

  • Mary

    May I suggest that you reverse the x and y axes?

  • Aaron Macpherson

    Why was the following entry removed from the 1.0 version of the chart?
    20,000 mSv “highly targeted dose used in cancer radiotherapy”

    • miriam

      Hiya Aaron,

      Thanks for your comment. We’ve temporarily deleted it as we’re trying to refine and revise the graph so that it conveys the difference between an acute dose and a longer dose over time – which we feel the image doesn’t do right now. We’ve got a new version in the pipe and will upload soon!

      Researcher, Information is Beautiful

      • why

        Your explanation does not convince me! If you really think that the graph isn’t quite right because it doesn’t convey the difference between acute and longterm doses why didn’t you remove ALL acute doses? You just deleted THIS one cancer dose! Why?

        Better to put the original visualization back and write a comment referring to the imperfections of the graph! And after that you could make two distinct graphs to separate the two kinds of doses.

        And btw. there are a lot of proofs that the cancer treatment kills the people and not the cancer. By editing the graph like you did you look like someone who wants to hide this truth!

        Another btw: Where did you get your data from?
        Waiting for your new chart….

  • cs30109

    Actually, you get no dose from a banana. Potassium is under homeostatic control in the body so if you take in additional potassium from a banana, you just end up secreting the difference. There’s no net gain in potassium concentration in the body.

  • rpath

    On your full chart, the 5.0 exposure from dental x-rays is ranked lower than the 3.5 from a day in Tokyo — is the an error? They appear to be reversed, or the 3.5 number is incorrect.

    • miriam

      It’s a typo. Thanks for pointing it out. Will correct!

      Researcher, IiB

  • Kit Halsted

    Why does it go 0.1, 0.25, 1.0, 5.0, 3.5, 10…? Makes me question whether background radiation in Tokyo (listed as 3.5) is really a third of background everywhere else (listed as 10), which, of course, makes me question the whole chart…

    • miriam

      Hi Kit,

      Yep, the 5.0, 3.5 thing is a typo, now corrected, cheers. The Tokyo figure should have been (and now is), 7.68 µSv/day (rounded to 7.5).

      The Tokyo reading was taken from here:
      You can see the many readings within Tokyo also vary quite a lot between around 0.1 µSv/hour, or 2.4 µSv/day, and 0.6 µSv/hour, or 14.4 µSv/day. The 7.68 µSv/day (0.32 µSv/hour) figure taken from Tokyo Bay was selected because it falls roughly in the middle of the range.

      Figures for natural background radiation (which accounts for the vast majority of most people’s radiation exposure) vary *massively* by geographical area. This UN report, gives an annual dose of ’1-10 mSv, depending on circumstances at particular locations, with sizeable population also at 10-20 mSv.’ [Table 4, page 8 of In some areas of Iran, it can get up to 260 mSv/year, with no apparent ill effects on the local population noted so far. [See - though you'll need a login.]

      The Tokyo daily reading given is therefore well within the normal worldwide range for natural background radiation, but bear in mind the wide variation in both.

  • question

    Where would smoking a cigarette be on the scale

  • Tom

    Information is missing:

    You remove the 20K cancer treatment dose from your graph.

  • specialsymbol

    Now, this chart has obviously been done as a reaction to the Fukushima accident.
    However, there are four things I’d like to mention in this case:

    1. Most of the radioactive stuff in Fukushima has been blown (and washed) out to the sea. So to speak it’s no surprise that the land measurements are comparably safe. As unsurprising as we incorporate a quite high dose via food, since this hasn’t been the first time (it’s not only seafood: look at bavarian regulations on shot boars, still about 1/3rd of them have to be disposed of). I wonder what they measure in Oregon and Washington state right now.

    2. While it is true that there are many misunderstandings about radiation and its effects just look at the 100µSv value (1hr 3km away from the plant). Doesn’t look bad, it’s the same as a medical treatment. But no sane person would want to receive a chest x-ray every hour. Radiation effects are all about time.

    3. Todays background radiation is quite high (yet not as high as in the sixties..). For example low-background steel has to be recovered from sunken (war-)ships to build sensitive radiation measuring equipment. Todays steels are (or better become during the production process) too contaminated.

    4. The trustworthiness of values measured (or better, released) by governments and companies is questionable to say the least – e.g. TepCo published relatively low values for quite some time. Just to admit later that they maxed out their meters, radiation could’ve been higher but they couldn’t take any better measurements. In Germany after Chernobyl physicists were forbidden to take measurements on their own, universities had to collect Geiger counters from the staff. This was to prevent upsetting the populace and not to question nuclear energy.

    Having studied physics (and nuclear medicine) I know it’s hard to grasp the mechanics behind radiation and it’s effects on the human body. But trust me: the less there is, the better. Radioactivity is all about chance. You can take a lot without being affected, yet even the slightest dose could do harm. No country needs to rely on nuclear energy, no matter how often they repeat that claim (it’s a common practice for politicians, repeating “facts” until everyone takes them for the truth instead of proving them – especially if there is no proof, or even proof for the contrary). There are only two reasons for nuclear energy, one is to obtain the ability to build nuclear weapons. The other is to distribute loads of money from the many to the few (and disguise it as a “public service”).

    • specialsymbol

      Ok, just to add to my #2: I appreciate that you implemented a timeframe, but I only managed to see it now.. guess it’s a little bit late here.

  • Geoff Russell

    Great chart. I don’t really buy the criticisms of the log scale. Technical people will see this instantly. Joe Average doesn’t know what a log scale is and even pointing this out won’t help. But what you’ve done by NOT scaling the graphic to fit in a single A4 page is to require people to scroll and this gives a sense of scale that is reasonable. So I reckon you’ve done a great job.

    I have one suggestion. 1.5 packs a day may deliver 36 mSv, but delivers a much much higher risk of cancer than 36mSv delivers. You really need two lines, the one you already have and a second saying 1.5 packs per day delivers a lung cancer risk equal to X mSv. I’m not sure what X will be, but somebody will know!


  • Johnnieace45

    Check out my youtube video johnnieace45, titled
    ” radiation level after nuclear stress test”. It pegged my dosimiter to max level of 350 mSv even when 15 feet from the unit.
    Has been 6 hrs and many glasses of water, I am still lit up at 340 mSv.
    Didn’t realize they were going to use such a high dose. – never again!