Crowd-sourced radiation readings from Japan

Inspried by Shigeru Kobayashi‘s work and the desire to just make something, I created a google map visualisation of radiation readings from across Japan.

These are crowd-source readings from numerous geiger counters hooked up to the Internet. The folks at Pachube have aggregated these readings and made them available for people to play with. The readings come from sources such as local councils, motivated individuals and official readings from Japan’s Ministry of Education, Culture, Sports, Science and Technology (MEXT).

My aim with this map is to make the data easily readable and understandable, so people can very quickly get an overview of the radiation levels across Japan and are able to drill in to get further details per region.

From a user experience point-of-view, I wanted the numbers to be at a glance, avoiding the extra clicks that these mashups usually ask of the user. So you see the readings highlighted in yellow on the map. The orange circles are coloured based on the severity of the reading (the darker the orange, the higher the reading). Clicking on these circles will also bring up more details about the reading (location, timestamp, millisievert).

The toughest part of this visualisation is really understanding what the numbers mean and what impact they have on human health. The first step to this process is standardising the units of measurement, as the crowd-sourced measurements and visualisations may use a number of representations. Units here are in µSv/h (or microSieverts) and we’ve been hearing CNN and NHK World refer to the unit Milisieverts (1 miliSievert = 1000 microSieverts). I also urge other mappers out there to use the µSv/h unit, so we speak a common language.

Next we have to know what normal readings might be and what doses would result in impact on human health.

I’ve seen a few different references to ‘average background radiation’ on the web. But it’s been hard to decide which ones to use and what factors were included in these considerations of background radiation.

On the map, I’ve opted to refer to the average geiger counter reading across Japan, calculated from values provided by MEXT – 714.929 µSv / year. ( When you click on a reading to get further details, you’ll see each reading compared to the national average. I’ve also included the US natural background radiation value from Wikipedia (3000 microSieverts / year), however, the calculation of this value is unknown and may include more types of exposure than just open air geiger counter readings.

Time is also a factor as exposure is cumulative and we normally measure how much radiation we gather in our bodies based on annual exposure, rather than hourly.

Here’s a quick table to better understand the impact of millisievert exposure (gathered via Wikipedia):

  • 3 milsiv. average annual background (US)
  • 10 milsiv aircrew annual exposure
  • 50 milsiv current legal annual max
  • 100 milsiv carcenogenic annual level
  • 250 milsiv damage bone marrow, spleen
  • 1000 milsiv 5% death


Monday, March 21, 2011

There were some elevated readings in the Ibaraki region last night. One area reported levels of 2 microSieverts / hour.

Here are 2 screenshots 12 minutes apart. The first one is March 20 @ 21:14 (GMT) and March 20 @ 21:26









Here are some great links to other resources:

Pachube blogs about their crowd-sourced feed and the ‘Internet of Things in action.’

@fleepcom has some amazingly detailed graphs of data coming out of Japan, including radiation levels and earthquake

Marian Steinbach has been laboring to scrape radiation readings made public by the Japanese government website (MEXT) is tapping into multiple radiation feeds and urgent Japanese citizens to contribute their own geiger readings

– The IAEA (International Atomic Energy Agency) has an update blog on the Fukushima Nuclear Accident



  1. Whats up very nice web site!! Man .. Beautiful .. Amazing
    .. I will bookmark your site and take the feeds also?
    I’m happy to seek out numerous useful info here in the publish, we’d like develop extra
    strategies in this regard, thanks for sharing. . . . . .

  2. Can I simply just say what a relief to find a person that truly knows
    what they are discussing on the internet. Yoou definitely understand how to bring a problem to light and make
    it important. A lot more people should look at this and understand this side of the story.
    I was surprised that you aren’t more popular given that
    you surely possess the gift.

  3. You can certainly see your expertise within the article you write.
    The world hopes for more passionate writers such as you who are not afraid to say how they
    believe. All the time follow your heart.

  4. King Neece says:

    Remove radiation from any water


    Our University of Alabama patented
    solar desalination product uses no electricity, has no filters to replace, can
    be taken anywhere and extracts pure water from any contaminated water source.
    It removes radiation, fluoride, salt, pesticides, bacteria, dirt and other
    contaminants from any water. It aids people to be prepared for disasters, saves
    hundreds on bottled water and utility bills. Made tough in the U.S.A.


    Please visit us:


    These units can also be placed
    together in arrays of literally any size, as needed, to accommodate a desert
    vegetable farm or remote area water needs, etc.

  5. Anonymous says:

    Hi, I have a question about the radiation data. My home town in Mitaka Tokyo. I always see that data in Mitaka around 0.150u00b5Sv /h with yellow sign in the map nIt’s a way up high amount than other places in Tokyo. I really would like to know aboutu00a0 detail of the area. Please let me know if you know about any farther information. Thanks.

  6. Is there anybody who knows how the micro sievert data are calculated?nI do my own readings with a GS1-device and get always 5to7 times higher micro sieverts per hour readings. Any measurement results in countings, nothing more. Sieverts are calculated from the countings. I can not get official countings. Just calculated sieverts.nMy device is offically calibrated and I am confident in the readings. I would like to find out about the reason for the difference in the results. If I could get the raw data countings for example for places in Hokkaido it would be easy to find out.nNeed help.nRegards, Stefan

    • Anonymous says:

      Hi Stefan, I found some details on the GS1 and it should output microsieverts, are you saying you are only reading CPM and are manually converting to microsieverts?nnAlso in the giegermap the readings are coming from geigers that only read gamma, so your GS1 should show higher readings as it is reading all types of radiation.u00a0nnThis guy on Facebook has 2 geigers, one reading all radiation and one reading only gamma. You can see the difference in the,nHaiyan

      • I am only reading Gamma radiation. Alpha and Beta are shut of.nThe problem is I can not get information how the readings, which are published, are calculated. Readings are simple countings (impulses). Anywhere in the world. Sieverts, like on your map are calculated.nThanks so far.nStefanu00a0

        • Anonymous says:

          I believe there is a conversion factor between cpm. But I think this factor differs based on the type of isotope your counter is detecting. I found this page which might help:u00a0

    • Hendrik Warntjes says:

      Jusat saw your post. I got a GS2, and I seem to have the same problem: GS has consistently higher mcSv/h values than other device. A few indicators that it’s the translation from CPM to mcSv/hr arena) i roughly get the same CPM as the folks at Safecast (around 40-50 CPM in central Tokyo). You can check this when you download the data to the computer. However Safecast translate CPM into mcSv/h using a factor of 350 (see and they continue to do so. My GS translates using a factor of 106, i.e. while my CPM readings are identical, the mcSv/h readings are higher by a factor of 3.nb) GS do advise that around 106 is the correct factor. Safecast advise that they use a more sensitive pancake probe, i.e. they would pick up more CPM… this does not soudn right as I get similar CPM values to them with my less senstive GS2. The GS would have to report only 1/3 of counts if the pancake probe was that much more accurate.nd) in northern Germany, the GS2 measured mcSv/h are 0.08-0.012, which is consistent with official figures there.nnCan we compare our figures? I would feel slightly better if I knew that at least between our two GS we would get similar readings, i.e. it’s not miscalibration (which is unlikely given that they are supposed to be calibrated, and the test in Germany shows good figures)nnThanks!nHendrik

      • Hello,nhere one more info.u00a0 readings are based on the calculation from cpm into sieverts. There are a lot of variables involved to come to the sievert value.nThese variables are notu00a0internationallyu00a0standardized. The calibration ofu00a0devicesu00a0are also notu00a0standardized. Measurements are notu00a0comparable.nWhat drives me crazy is that the use limit values for sieverts per year, per hour, for kids, workers, etc. similar to other countries and instituions but the sievert calculation they do regardless to those respective calibrations of divices.nIf they measure and publish 3 or 4 times lower sieverts they should at least adjust their oficial limits accordingly.u00a0nMay be not intentionally, just lack of education.nI contacted someu00a0responsibleu00a0authorities but no one can answer my questions and explains the real calculation according to their devices they use.nMy readings are online every 10 min. here:u00a0, Stefan Koester-Hirose

        • Hendrik Warntjes says:

          Thanks for that. Reassuring that you are also reading higher values than official – or even other GM counters. Out of curiosity: what is your average uSv/h up there in Hokkaido? In Tokyo (with my Gamma Scout that reports higher values) it’s around 0.24 uSv/h over past few weeks. Second question, hope you dont mind, what is the conversion factor of your GS? I mean to translate from CPM to uSv/h, these devices apply a conversion factor – as teh Gamma Scout are calibrated, it should be different for each device. My factor is 106. Can I ask what you see on yours?nThanks!n

  7. On the comment No.24

    There is a map for Germany

  8. This is a fantastic source of information. Are there any plans to extend the data sources to include other countries (Canada, USA, European nations, etc)?

  9. Pingback: Life, Death and Graphic Design: The Critical Role of Information Design in Emergencies by Peggy Cady : DesignNotes by Michael Surtees

  10. Great work! I have yet to find a better looking, easier to read radiation map of japan after looking for some time. Cheers, Aron

  11. Diego Roccato says:

    Are there similar maps for Europe and the US ?

  12. Wonderful! Information is power, more power to the people. Thank you for doing this. One thing which will help is a prominent diagram to help understand radiation levels. I include such a diagram, not my work, in the ‘Website’ field in this message – please consider

    • Thanks Ben! I’m looking to visualise the impact of radiation dose at some point. I like the xkcd diagram, though I do think it leans too much toward the ‘comfort’ end of the spectrum… it’s very difficult to feel alarmed by any dose of radiation based on that chart.

  13. Please note, that official (e.g.SPEEDI) Dose Detectors are balanced to energy in order to make the sensors equally sensitive to low, middle and high energy gamma rays. The µSv calibration is done for Cesium137, but the balancing leads to trustable µSv values for other Nuclei.
    Each GM counter without energy balanced tubes may read very different µSv levels. This is especially true for Detectors like “Gammascout” and others having a window to pick up beta and alpha rays. Readings from those devices should not be mixed up with Gamma energy balanced readings in a map, as this would irritate people and misread the particular radiation situation.
    We suggest to divide maps of readings into 3 classes: Calibrated Gamma DoseDetectors , Personal counters (incl name of the unit), Surface detectors for soil and food samples (incl.alpha & beta)
    Look at this and check my comment there.
    Here in Germany, we think of staring an online course about measurement techniques and health physics for all people planning crowd radiation reading networks.

  14. The “LIVE” number just jumped from 109 to 209 locations and readings closer to the area are now visible.

  15. It’s April 12 at 10:46 Japan time, and there are no readings showing on your map anywhere near the nuclear plant or even in Tokyo. Why? According to media reports, the government just raised the incident level to 7.

  16. Great work mate! Below is an email a colleague of mine received from the US State Dept regarding levels of radiactive iodine. The Japanese govt stipulates 100 Bq/L (100Bq/kg) for water before it warns people to avoid giving the water to small children. This is 1000X higher than the US standard of 0.1 Bq/kg. For an acute dose (short time duration) the US says 100 Bq/L is OK. The key word being ‘short time’.

    Here is a link to the Gunma Prefectural Office with some levels of Iodine and Cesium given for the past three weeks or so. Lately the iodine levels have been around <1.5 Bq/L in Maebashi. Cesium levels in water have fallen below the detectable threshold of their devices.

    From: TOKYO, ACS1 []
    Sent: Thursday, March 24, 2011 2:01 PM
    To: TOKYO, ACS1
    Subject: Warden Message on Tokyo Drinking Water

    U.S. Ambassador to Japan Statement
    Warning For Parents and Caretakers About Radioactive Iodine Detected in Tokyo Drinking Water Supply
    March 24, 2011
    The Tokyo metropolitan government on Wednesday, March 23, 2011, cautioned residents that infants should drink only bottled water because radioactive iodine exceeding the limit for that age group was detected in water at a purification plant.
    The U.S. Embassy in Japan suggests U.S. citizens who live in Tokyo follow these recommendations. In addition, women who are pregnant or nursing should also follow these recommendations and drink bottled water. This guidance is consistent with the guidance that the United States Government would provide to Americans in the United States under similar circumstances.

    U.S. citizens in metropolitan Tokyo can take the following steps to safeguard the health of infants (aged 0-3 years):
    • If giving water to infants, use only bottled water.
    • Use only bottled water to mix formula, cereal or other infant foods.
    Health experts say that changing the water source for infants from tap water to bottled water should be adequate protection from exposure to radioactive iodine. No additional medication, such as potassium iodide (KI), is necessary at this time. Taking KI when it is not needed can harm a person’s health.
    At this time, no changes in drinking water are recommended for adults because the limit for adults is higher than the amount detected in the water purification plant.
    Embassy staff is consulting with health experts and radiation experts to continuously monitor these new developments. If more information becomes available, we will share it with you on the Embassy Web site.
    Q and A:
    –What are we advising U.S. citizens? Why?
    Due to an elevated level of Iodine-131 found at a local Tokyo water purification plant, we are advising American citizens in Tokyo that infants (aged 0-3), as well as women who are pregnant or nursing, should drink only bottled water.
    –What is the Japanese standard being used?
    The Japanese standard for Iodine-131 in drinking water is 100 becquerels per liter if the water is to be consumed by an infant (0-12 months) and 300 becquerels per liter if the water is to be consumed by an adult. The current reported contamination of 210 becquerels per liter is therefore about twice the permitted level for infants and about two thirds of the permitted level for adults, under Japanese regulations.
    –What is the U.S. standard?
    The U.S. Environmental Protection Agency’s published standard for Iodine-131 contamination in drinking water is 3 picocuries per liter, which is equal to about 0.1 becquerels per liter. However, the assumptions underlying the EPA standard for continuous exposure do not apply to the current situation in Japan, which is a temporary exposure resulting from an accidental release. In addition, the science of radiation protection has advanced considerably since the EPA standard was published in 1974. If one uses the latest science and makes the adjustments in the calculations underlying the EPA standard in order to make it applicable to the temporary exposure occurring in Japan, one obtains a figure practically identical to the standard that the Japanese authorities are applying.
    What is the risk of the current Iodine-131 levels to adults drinking the water?
    –Short-term exposures at the levels that have been reported do not present significantly increased risks beyond the sensitive populations already identified.

  17. Michael Reiter says:

    Hi Haiyan,

    I’m a journalist writing about your and Shigeru Kobayashis Geigermaps for a danish engineering journal. I’d love to ask you a few questions about the project. If you’re interested please get in touch!


    Michael Reiter

  18. If you are interested look at my prototype display for online weather data as well.
    Hoover over the Question mark

  19. Pingback: How the Internet of Things Helps Us Understand Radiation Levels - ReadWriteCloud

  20. Pingback: How the Internet of Things Helps Us Understand Radiation Levels - Finding Out About

  21. I guess you know it. Nevertheless there is an example for mapping in Germany
    Your flag with the data looks nice but is very much packed and overlapped.
    Small collored points with a tooltip when
    hooving over a point could be also a solution . For example like

    • Hi Stefan, thanks for your great feedback! Am on traveling this weekend and will take a look at the map again early next week. Keep the thoughts coming 🙂

  22. Great tool.
    Good job. I placed a link on my wenbsites.
    I guess You get now much more access within Japan.?
    Hope You keep it going for a long time.


  23. Something weird – I’m looking at it in Google Chrome, and when I use my mouse wheel to zoom out, the labels on the individual rad counters get bigger as the country gets smaller. Zoom in, and the opposite occurs.

  24. Pingback: Worried About Radiation? There's an App for That | Penn Olson

  25. In Japan the recomended dose rate is 0.114mikroSV/hr when according to Wikipedia the average doserate in the world is 0,23 mikroSV/hr. The later is background radiation. Question what the former is then?

    • Hi Sejoke, after some research I have come to understand that geiger counters can be calibrated to read some types of radiation and not others. So comparing your geiger counter readings to the natural background radiation values we get from Wikipedia is not useful. It’s more useful to take a normal reading with your geiger counter and then compare against that value. Thankfully the official source from japan has provided us with ‘normal band’ values for their geiger counters, so we know that the average reading across Japan is 0.081 microSv/hr. In the map, I am now comparing all geiger readings against this value to show when readings are above average and below.


  26. Pingback: Castlemaine Independent , Archive » Realtime crowd-sourced data, the internet and Japans geiger counters

  27. I need your help.
    Do you know anybody in Japan who retrieves Gamma-Scout data and places single values automaticly on the Internet.?
    I bought one and whant to upload microS/h continously automatic on my website. Regards, Stefan Koester-Hirose

  28. We are using the Ushahidi Platform to keep track of radiation readings on the West Coast. We have four stations reporting, but expect to add many more in the next few weeks. Look on our website at the health impacts of radiation fact sheet done by nuclear experts in the U.S.

  29. The real issue is whether things are getting any better. data from shows that it is as high as it was on 3/15 when reactor 3 exploded.

  30. Pingback: Haiyan Zhang’s Geiger Maps | Media Collective

  31. Pingback: Haiyan Zhang's Geiger Maps - Core77

  32. The map would be much more intuitive if the numbers were highlighted in the traditional stop light colors of green (safe), yellow (above average) and red (above safe limits)

  33. Your notice saying the dose limit is 1000µSv a year is very misleading as most people in developed countries typically are exposed to an average of 2400-3000µSv a year in the normal course of things. The average for American adults is 6000µSv a year due to higher occurance of health checks, xrays etc. Did you know this?

    • Hi there Roger, thanks for your feedback! I’ve been thinking about this issue of how to represent meaning in these readings. The figure of 1000µSv is actually from the Japanese government and it is their dose limit for public space radiation per year. Apologies. In the document where I got this number originally, the English translation merely said ‘dose limit for public’ ( At the moment I’ve removed this from the map to consider adding in this figure (as public space) and global averages, as well as linking to some sort of visualisation that shows you overall impact on health.

      If you have any thoughts on this I’d love to hear it! As these readings are for public spaces, it’s useful to consider average background radiation in public areas and assume people will accumulate on top of that food, activities and medical doses. Will do some digging on Wikipedia to find average background radiation across the globe.


  34. BRILLIANT! Great work. Thanks for making this.

  35. Shigeru Kobayashi says:


    First of all, thank you very much for your suggestion in a very short time. I totally agree on “the numbers to be at a glance.” Regarding the units, yes we, people in Japan, are familiar with mSv/h as the unit for radiation.

    I have a question. How did you determine the radius of each point? If I understand it, alpha of the fill color represents strength. I like that idea basically, removing strokes from circles might be better, since no attention will be paid to lower value points.

    Anyway, I’ll also try to suggest other visualization model in a few days. Let’s keep discussing.


Leave a Reply

Your email address will not be published. Required fields are marked *