Official statements on the Fukushima nuclear accident reek of bad data. “100x safe levels”, “only 10 percent of”, “10 million times greater than”. Examples:
MSNBC – Radioactivity in water at one earthquake-crippled Japanese nuclear reactor soared to 10 million times its usual level Sunday.
Marketwatch – radioactivity in water that had pooled in one of the buildings had reached 10 million times the normal level found in the reactor core, The Wall Street Journal reported. But the company later said the reading was faulty and revised it lower to 100,000 times the normal level, the paper said.
Inquirer.net – Japanese media were quick to point out that the new rating put the latest disaster on a par with Chernobyl, although officials say only 10 percent of the amount of radiation leaked in 1986 has so far escaped from the Fukushima plant.
ABC – Officials there said radiation was 10 times normal at one point but not a threat to human health in the sprawling high-tech city of 13 million people.
NBC affiliate – A “rough analysis” of the radiation exposure in Japan and projections of excess cancers that are likely there appear to be roughly 10 to 100 times greater than with the Three Mile Island nuclear disaster.
Marketwatch – The government estimated the total amount of radiation contained in the released water at 150 billion becquerels—exceeding the legal limits by about 100 times.
Australian – A Japanese safety panel says radiation is 1000 times above normal at a nuclear plant in the country’s northeast
Guardian: The release of water that is 100 times the legal limit is an unprecedented breach of operating standards, but it is considered necessary so workers can concentrate on containing more severe leaks.
VoR – At the beginning of this week, the radiation level at the 4th generating unit was 10,000 times more than the legally permissible level.
Obviously, there is some number-fudging going on here.
I only retained one thing from college stats classes, and it was from a lecture about spotting fudged/faked numbers. In an exercise, the professor left the room and had students write two sets of single-digit numbers on the chalkboard.
One was randomly generated by a computer. The other was hand-picked by students who were told to “make it look random”.
Prof came back in, pointed at the hand-picked set, and said “that’s the fake”. Apparently there wasn’t enough repetition. It was too random. Repetition is common in real sets of numbers, so a lack of repeating numbers is a dead giveaway.
Once you understand the implications of this, all sorts of data start to look funny.
Back to Fukushima. At least some of the official releases are questionable, and likely understate the real risk.
Give the people who live in affected areas all the data. Let them decide what to do. Do not coddle, do not attempt to protect the population from information.
The risks are significant. As Greenpeace and others have noted, there is ~24x as much fuel at Fukushima than Chernobyl. And Japan is 40x more densely populated than the evacuation area in the 1986 meltdown.
It is vital that the situation be evaluated scientifically. Downplaying and underestimating the disaster has already caused significant harm and delay.