Bugsy,

http://www.ox.ac.uk/media/news_stories/2012/120612.html

This large scale study at Oxford agrees with my assessment.

Guidelines on exposure to low doses of radiation have largely been based on estimated risks from models using data from Japanese survivors of the atomic bombs, where radiation exposures were brief and very much higher. As a result, there have been some long-standing uncertainties about the extrapolation of these risks to low radiation doses.

This link to the Oxford study on leukemia

http://www.ox.ac.uk/media/news_stories/2012/120612.html

is now being used by both "sides". Noel thinks it shows "increased health risk from low dose ionising radiation" and ShadowMinister
thinks it shows precisely how tiny and insignificant low dose radiation is compared to the big guns in cancer causation ... cigarettes, red and processed meat and alcohol. Similarly, the fact that people can detect Fukushima radiation in tuna thousands of miles away says plenty about the sensitivity of measurement and absolutely nothing about the risk to the tuna. The big ocean risk from Fukushima is from debris that is now floating in the pacific. This of course comes from all the buildings not built like nuclear power stations which got wiped out with 20,000 people by the tsunami. Obviously what you should be building in quake zones on the coast is nuclear power stations because they are one of the few types of buildings designed for the job ... along with Taipei 101!

But while talking about leukemia, I notice that neither Noel nor any other anti-nuclear advocate has responded to my question about sausages. Please ... do you care about cancer or only radiation induced cancer. But perhaps its only leukemia and not bowel cancer that Noel is concerned about.

What causes most leukemia? As the Oxford article says, nobody knows. But here's something that needs more work ...

http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2842202/?tool=pubmed

A huge study showing total meat consumption is associated with a 45% increase in acute myeloid leukemia ... particularly nasty.

Is it the meat? That's not clear, there is no correlation with the major meat DNA shredders. But there are so many meat mutagens that it may be one of the lesser ones that is to blame. Still, since there is no reason at all to eat meat, it pays to be safe. With nuclear power on the other hand, there is a really really really good reason to take the tiny risks associated with radiation --- climate change and the total failure of renewable technologies to make a dent in the problem.

SM: "This large scale study at Oxford agrees with my assessment."

In what way?

The assessment about Fukushima deaths which is irrelevant to the point I am currently discussing?

Or your point on extrapolation from higher doses? Because these studies don't actually agree with your latter statement. They actually address some of the 'uncertainties' mentioned in the out-of-context quote.

And guess what? They BOTH agree that the LNT model of radiation exposure is valid.

How about that?

Now you can argue that dilution of released radiation from Fukushima will result in negligible deaths and so poses minimal risks, and I would probably agree with you. But the farmers from the region are finding it very difficult to sell their produce, which adds an economic element to the consequences of releases and accidents. You may say that the food is safe and the average Japanese consumer is silly to shun it (as the Japanese government is currently doing), but that doesn't really help maintain the markets. Consumer perception adds to the economic risks, even if the health risks are negligible.

Nuclear power is all about risk management, and the only reason you can argue that it is the 'safest' form of power is because of the risk minimisation and mitigation strategies that have been put in place wherever it has been used. The only way that people are going to be convinced to take it up in a larger scale and in areas where it has not been used before is if you can convince them that all risks have been considered and NOT idly dismissed, and that mitigation or minimisation strategies are or will be in place. The risks are real, whether they be 'relatively' small or not.

Trying to discredit what is actually reasonable research using misinterpretations and bizarre comparisons (eg. red meat consumption) is not going to help.

Bugsy, relative risks matter. As long as anti-nuclear activists keep making people more fearful of nuclear power than climate change, we are in serious trouble. I bring up sausages because it shows that the concern for safety of the anti-nuclear movement is either disingenuous or totally irrational, or both. It also shows the real disregard for climate change ... because everybody from Pachuri to Hansen has put meat reductions on the agenda.

Some comments on the Oxford paper follow, noting that I've only seen the abstract, uni press release and ScienceDaily article based on the first two (http://www.sciencedaily.com/releases/2012/06/120618150045.htm). I'd be interested in the views of anyone who has access to the paywalled version.

The most problematic aspect from my POV is with the dose assessment. "Radiation exposures were estimated for mother's residence at the child's birth from national databases, using the County District mean for gamma-rays, and a predictive map based on domestic measurements grouped by geological boundaries for radon." Using district means sounds very ropy, as the radiometric variations within districts would often be greater than the variations between them, with every possibility that human settlement patterns are biased one way or the other. Maybe this issue is addressed in the full paper (and the abstract says they've looked for confounding factors), but I'd take a fair bit of convincing.

The radon assessment would be more geographically precise, because it takes geology (the strongest control on natural radiation, at least near sea level) into account, but interestingly no statistically significant correlation was observed with radon. I'm a bit mystified as to why (apparently) geology was not also taken into account to improve the gamma dose estimation.

Finally, this is an important observation from the article: "There were no statistically significant associations between other childhood cancers and natural gamma-rays, or between any cancers and levels of radon in the natural background radiation." As Jaro touches on and Ben Goldacre explains in his wonderful book 'Bad Science', if you look for enough links between unrelated phenomena, eventually you will find a statistically significant association through random chance alone - and of course this is the one that gets the headline (but kudos here to the article for at least mentioning the null results).

Bugsy,

Evidently you have only skimmed a few of my posts as you have missed most of what I said. I said that the LNT model is intuitively correct, but no "proof" exists.

The Hiroshima study, does not actually look at low dosage exposure. The lowest dosages are in the order of 20 yrs back ground exposure in a very short time, and the "excess" deaths at these levels are all small and from forms of disease that can be caused by other factors such as a small increase in the intake of red meat. The cancers are all linked with much higher doses (100yrs background exposure).

Extrapolation from one area of data to another is scientifically tenuous, as the example I gave earlier of car crashes. A 90% fatality rate at 120kmph and a fatality rate of 60% at 80kmph does not translate into a 2% fatality rate at 5kmph for obvious reasons.

The quote I cut and pasted in my last post shows the scientific community's obvious discomfort with this extrapolation.

The Oxford study is the only study so far to show that low dosages of radiation have any measurable affect whatsoever. The studies of millions of people show a "statistically" significant increase of only one cancer type i.e. childhood leukaemia, nothing else showed any statistical increase. Even then the predictions of the effect range from a 3% increase per mSv to 20%.

This is similar to the large scale study into the linkage between mobile phones and cancer. The study of millions picked up a faint link. This was put in perspective by one of the researchers who said that a cup of coffee a day was many times more dangerous.

The safety of nuclear power is statistically proven with a wide margin. The number of deaths from all causes from nuclear generation is half or less of any other generation source including solar and wind.

The choice for the next generation is CO2 emissions or nuclear. Either AGW is a serious problem or it isn't.