[dmoy]

I don't know of any authoritative resources on radiation->cancer risk, can anyone provide more information? The article starts out with a cool premise but I have a hard time swallowing the linear rem->%chance relationship described here

"If 25 rem gives you a 1% chance of getting cancer, then a dose of 2,500 rem (25 rem times 100) implies that you will get cancer (a 100% chance)"

[archgoon]

Unfortunately, we don't have the data. Yep, it's non linear, but all of our data is basically on two ends of the spectrum. Low amounts of radiation on one end, accumulated from natural sources or working at nuclear plants, and high amounts of radiation on the other, accumulated from survivors of Hiroshima, Cherynobyl, and some nuclear accidents.

So, given those data points, with clusters on two ends of this spectrum you're stuck with interpolation. Ideally, we'd like a good, predictive, biological based model that explicitly showed how you go from biological damage to cancer or death. We don't have that (though there are people researching it), so we go with statistical techniques. This results in basically 3 different proposed models.

http://en.wikipedia.org/wiki/Linear_no-threshold_model

The Linear No Threshold Model:

Draw a line between the two regions. Bam! Done.

This is clearly wrong for very large doses, but is used mostly to try and estimate the effect of a small change in radiation exposure to predict increase in cancer incidence. This is very important for public policy. Of course, this makes all these models politically contested.

http://en.wikipedia.org/wiki/Threshold_model

Threshold Model

This model predicts that small amounts of radiation has zero effect on cancer incidence. According to this model, going to higher altitude, or taking a plane flight, won't increase your risk of cancer.

http://en.wikipedia.org/wiki/Radiation_hormesis

Radiation Hormesis

A little radiation is good for you! The biological argument is that your body gets used to dealing with damage from radiation. Don't tax it too hard and you'll be stronger. So take a plane trip, and enjoy the X-ray scans!

[justatdotin]

probably worth noting that it's LNT, which implies No Safe Dose, that forms the basis for international radiation protection standards

[davvid]

A little radiation is good for you!

The science says otherwise.

http://www.rrjournal.org/doi/pdf/10.1667/RR2629.1

http://www.washingtonsblog.com/2011/04/low-level-doses-of-ra...

[fixedit]

Hate to keep taking issue with your comments, but they are pattently wrong. Using a study of atomic bomb studies and an anti-nuke web site do not constitute science.

Studies Ive read, completely contradict your statements (http://ehp03.niehs.nih.gov/article/fetchArticle.action?artic...) for one. There are others, and I am sure that I will not be convincing you, but I didn't want your voice to be the only one on the issue here.

[tsotha]

The LNT model is clearly wrong, but people like numbers even if they're wrong, which is why you see it used in planning and media reports. There are probably fewer cancer cases than LNT would predict, but beyond that we simply don't know.

[lisper]

Wikipedia is your friend:

http://en.wikipedia.org/wiki/Roentgen_equivalent_man

It's not authoritative, but if you really want to dig in to the issue the references at the end of the article are a good place to start.

[davidcuddeback]

That's about where the author lost me. Up to that point, he seemed to be pulling numbers from research, but then he makes a transition to pulling numbers and formulas out of his ass. For example, I'm not convinced that "if spread out over 1,000 people, so that everyone received 2.5 rem on average" is a valid deduction.

He then goes on to explain how "to know how many excess cancers there will be." I listened to an interview with a radiation expert [1] a couple weeks ago, and he made it very clear that we don't know how to make those calculations that the author of this article is claiming to know how to make. If I were to try to regurgitate the points from the interview, I'd risk spreading a misunderstanding of the issues. So instead, if you're interested in what the experts think on this subject, you can refer to the original source [1].

[1] http://www.pointofinquiry.org/nuclear_risk_and_reason_david_...

[yk]

The spreading out effect is a direct consequence of the linear no threshold model. If you have a total dose D spread over N people, then everyone receives a dose D/N. With some constant k converting individual dose to cancer rates, you end up with an expected number of k D/N * N cases of cancer. ( So it is almost certainly wrong, however it is not understood how wrong it is.)