SM: I bet the authors are glad they did not have you as a reviewer. They did a formal threshold analysis ( outlined on page 3 or p.231). The data actually does statistically support no-threshold model. Sorry mate, but just calling it 'noise' and dismissing a carefully analysed dataset is not going to cut it. The doses analysed go down to below 0.005Gy, which if 0.1Gy is 40 years background, then 0.005 is 2 years background. What would constitute a 'low dose' here?

Bugsy,

The excess cancer deaths below 0.005G is statistically zero. See table 9

Given that the excess radiation from Fukushima is in the region of 0.0001g or lower 0.00001G, where is the validity of this study?

SM:

From the WHO report referenced in Geoff's earlier post:

In Fukushima prefecture the estimated effective doses are within a dose band of 1−10mSv, except in two of the example locations where the effective doses are estimated to be within a dose band of 10–50 mSv.

-Which is approx 0.01-0.05 Gy isn't it? Which I would say is reasonably relevant.

Estimated thyroid doses are much higher in some areas.

The dose you quoted are from neighbouring prefectures and the rest of Japan.

This is of course, all aside from the mandatory evacuation and associated further disruption of the lives of many thousands of people, which is how they copped a lower dose in the first place.

And so after more than a year, they've been told that they are unlikely to die from cancer from the nuclear plant.

Great.

Now this is one plant. How many do we need worldwide?

The safety record for deaths associated is impressive, and so is the failure rate of reactors. According to this website: http://www.world-nuclear.org/info/inf06.html over 14,500 cumulative reactor years of operation, they count only three major accidents (I guess they don't count pre-1979 accidents).

This conservatively makes about 4833 reactor years of operation per major incident. And then you realise that over 440 reactors are currently working worldwide, which means a rate of about one major accident per 11 years of operation across all reactors. And that's a conservative estimate.

How many reactors are going to be needed? We currently generate about 13-14% of our power globally with nuclear, will we need to double that?

I know you are trying to say that the medical consequences of reactor failure are nearly negligible for the general population, but that doesn't really tell the whole story of what happens to an affected area post-failure does it?

Many people hold legitimate concerns over the long-term safety and feasibility of nuclear power. These risks will have to be properly weighed against mitigation of emissions.

Bugsy ... how about factoring in lives saved by nuclear power during the past 40 years. Apart from large numbers of mining deaths and diseases, if it were not for nuclear power, all the countries that have it would have had more air pollution deaths. Many of the workers at Fukushima would be dead if working installing Solar panels on the coast. Three mile island wasn't a serious accident ... costly yes, but serious no. Over a thousand people are killed in plane crashes each year. But nobody figures planes aren't safe, they just do the maths ... car or plane? plane safer, easy. People who think planes are dangerous are sent to a shrink .. But nuclear is treated differently. The evacuation more than likely killed more people than it saved. We had a large petrochemical fire in Adelaide some months back. It spewed carcinogens all over the place. Did anybody demand an evacuation? No. Did anybody even bother calculating the future deaths from the cancers? No. Did anybody bother to calculate the cancer deaths from the Chiba refinery fire in Japan after the Tsunami. Why not? Because the anti-nuclear movement has succeed in getting people to put nuclear risks in a different category. When Tokyo was firebombed during WWII, more people were killed than with the atomic bombs, they all breathed in huge amounts of carcinogens. Where are the cancer studies? Ditto the bombings of Dresden. The anti-nuclear movement typically knows absolutely nothing about cancer at all except radiation cancers. They aren't phased by 80,000 extra cancers from red and processed meat but go absolutely bananas over 20 cancers over 40 years. Why? Has any of them seen a bowel cancer patient die? Is it somehow better than dying of leukemia that it deserves no attention? It isn't. It's tragic .. and so is falling off a roof installing a Solar panel and ending up in a wheel chair. I just want risks assessed rationally and compared properly. Currently the anti-nuclear movement is another arm of big coal and a serious problem in getting action on climate change.

Bugsy,

My error, It would then appear that some people in two areas of Fukushima might be threatened with a 0.3% increase in cancer.

Note that one major incident resulted in no release of radiation, one was due to a 1 in 1000 year Tsunami that killed 20 000 people and ruined thousands of square km of farmland, and the worst was a 1950's Soviet design with no protection. The question you should ask is how was any safety system 40 years ago?

New designs are far safer, and the chances are that any "accident" like 3 mile island will not result in anything like Chernobyl or Fukushima