Information, Misinformation, Disinformation (or, “These aren’t the droids you’re looking for.”) Part 1

droidpic016

Whose job is it to make this stuff easy to understand?

“YOU CAN’T ALWAYS FIND WHAT YOU’RE LOOKING FOR…..”
[Skip to Part 2] At Safecast we assumed from the start that our data should be accurate, easy to understand, informative, well-visualized, and easily accessible. In many respects this simply reflects “best practices” in information design, as well as a vision of social responsibility in which openness and transparency are paramount virtues. So when we make decisions about how to present our data, we adhere to principles of intuitiveness, depth, context, and dare we say it, beauty in design. We want to make it as easy as possible (AEAP) for people to find what they’re looking for, and to find out what it means. That’s why we’re continually miffed when official sources of information seem to be operating under an entirely different set of assumptions.

To be honest, the seriousness of government missteps and opacity during the early weeks of the disaster led us to accommodate ourselves to vastly lowered expectations in terms of the quality and accessibility of information we’d probably see from official sources. Even though it’s their job, and they are legally required to provide many kinds of information, many of us prepared ourselves for endless tooth-pulling and fact-checking about radiation information. So first, I’d like to give a sort of brief status update:
1) The government has made a lot of information available, more than we expected (because we expected nothing).
2) It still needs to be scrutinized, fact-checked, and independently confirmed.
3) There are still some areas where a lot of us have been pulling teeth for months and still haven’t been able to get the information we’re looking for.

So let’s just agree to live with #2 and #3 for the moment. It means constant effort on our part, but enough of us are constitutionally well-equipped for this kind of research-based tug-of-war that it’s not really that onerous at this point. We get good at it, we build trust, and people who were once opponents sometimes become allies, because frankly, they need our help.

But #1 is where we find ourselves really scratching our heads. There is all sorts of official information available, and a lot of it is proving reliable, but it’s rarely as easily accessible or informative as it should be. In fact, locating and using the data is usually as difficult as possible (ADAP) considering how easy it is now to find good information and web designers, and how inexpensive it has become. It should be easy to do a good job, if the people in charge really care about doing a good job.

The Canberra Fastscan series of whole body counters is being used by many hospitals in Fukushima to measure the internal contamination of residents.

REPORTING THE RESULTS OF WHOLE-BODY TESTS:
One brief example would be how results of internal contamination monitoring done with whole-body counters (WBC) are being reported. These tests are being done in municipalities all over Fukushima prefecture, sometimes under the direction of the prefecture itself, more often by individual hospitals on behalf of their municipalities. Fukushima Prefecture has put up a web page to communicate the results of testing done through Oct. 2012, as well as a more detailed breakdown. They’ve gathered quite a lot of numbers for us:

The WBC report page provided by Fukushima Prefecture summarizes the key data in this table. It shows how something can be entirely accurate and extremely uninformative at the same time.[

Fukushima Prefecture WBC results page (Japanese)
Fukushima Prefecture WBC results breakdown (pdf; Japanese)

They report that of 90,050 people tested from June through October 2012, 90,024 of them are determined to be exposed to an effective dose of 1 millisievert or less from internal contamination. The breakdown tells us how many people of each age category were tested in each municipality. Well, thank you very much Fukushima Prefecture. But it would have been very easy to give much more informative data. In fact, we wonder why they haven’t you been following the lead of Minamisoma, Hirata, and other towns in the prefecture which give us data that looks like this:

The towns of Hirata, Minamisoma, and several others have taken the initiative to provide WBC results in a much more detailed way, which also happens to be visually readable. This graph from Hirata summarizes and presents more useful data than Fukushima Prefecture’s table above.

Minamisoma City WBC results (Japanese)

They provide a full breakdown by contamination levels in Bq/kg, by age group, by gender, and showing change over time. There’s information about food and water sources, how much time people spend outdoors, all with clear and consistent graphs. The official Fukushima Prefecture report doesn’t even tell us how many of the “less than 1 millisievert” people actually had no contamination detected at all (ND). What we need to know is what the actual internal contamination levels were in Bq/body and Bq/kg, because while even a person with a high count in Bq/body still might not reach a 1mSv dose, we obviously want to reduce these levels quickly. We need detailed information in order to understand the effectiveness of food protection and decontamination measures, to be able to monitor government policy in these areas well, and to be informed enough to take action if necessary. The Fukushima Pref. reporting has all the hallmarks of a manager somewhere having decided that the most important thing is just to provide evidence that the tests were done, and that they hit their targets. The actual information content was absolutely secondary. Dr. Masaharu Tsubokura and Dr. Ryu Hayano advise Minamisoma City General Hospital and Hirata Central Hospital, as well as other municipalities, on their WBC programs. They spent months working with hospital staff testing various kinds of equipment and working out calibration issues to find the most reliable procedures and methods, and they gave great thought to how to report the results informatively, pushing for openness and candor. Dr. Tsubokura writes very informative and readable regular blogs, and sends email newsletters to help people interpret the results and put them in context: Dr. Tsubokura’s blog (Japanese)

This is very important, because the results require some interpretation. For instance, people need to know that a WBC test is like a snapshot, which captures the current state, but really can’t say much about how contaminated a person was several months previously. The tests also cannot tell us much about external exposures. Dr. Tsubokura has done an excellent job of explaining the limitations of the tests, the importance of using well monitored food sources, and why it’s best to measure entire families together, among other things. Fukushima hasn’t taken any of this into consideration in its WBC reporting. The kind of full reporting Minamisoma has promoted is gradually being adopted by other municipalities, but not by the prefecture, and certainly not by the central government. The latter don’t feel any obligation to make it easy. I think they expect us to be grateful that there’s any information available online at all. Some see conspiracy to deceive at work here, some see knee-jerk obfuscation, but it’s just as likely that no-one responsible for putting this information out was specifically tasked with making graphs, including a Bq/kg breakdown, etc.. If they weren’t specifically told to do it, they didn’t. We see evidence of this kind of thinking with distressing frequency.

This murasoi fish, a type of rockfish, was caught near the breakwater alongside the Fukushima Dai-ichi NPP in January, 2013. It measured an eye-popping 254,000 Bq/kg. (Tepco photo)

REPORTING THE RESULTS OF SEAFOOD TESTS:
Another brief example involves contamination testing results for fish and seafood. Since last year, both the Ministry of Health, Labor, and Welfare (MHLW) and the Ministry of Agriculture, Forestry and Fisheries (MAFF), through its Japan Fisheries Agency (JFA) branch, have been testing fish for radioactive contamination and publicizing the results. Fish as well as other food items are tested before being put on the market, and types that test above the gov’t limit of 100Bq/kg are not allowed to go on sale. From the beginning many people have been skeptical that food was being tested reliably enough to really keep unsafe items out of stores, and have criticized the government for not communicating the results clearly. The testing results from MAFF, in this case covering from Oct. 1 to Dec. 6, 2012, look like this:

Both the Ministry of Health, Labor, and Welfare (MHLW) and the Ministry of Agriculture, Forestry and Fisheries (MAFF) have been providing detailed statistics on test results for food contamination. This is the first page of a 34-page spreadsheet from MAFF. Trying to glean trends and changes over time from this spreadsheet is a test of diligence.

Recent Japan Fisheries Agency test results (pdf; Japanese)

There’s a lot of data here, dozens of pages worth. Hundreds of individual test entires, with results, testing parameters, dates, the location the fish were caught, etc.. The data can be downloaded in both pdf and excel formats, and all the older data is available. A few months ago, a number of us were looking though this data to try to determine if contamination levels in fish were decreasing or not, because that’s what most people really wanted to know. It was extremely laborious to count items that were above the approval limit, to note the highest levels found for each month, etc.. And then in October of this year, a US-based researcher named Ken Buesseler at the Woods Hole Oceanographic Institute released a study that used the same MAFF database, and sliced and diced it to show what kinds of fish were decreasing in contamination and where, and what kinds were not:

Buesseler’s graphs, which are based on essentially the same MAFF data, tell the entire story in a glance.

Buesseler Fukushima fish study summary

Buesseler Fukushima fish study abstract and link to whole paper (paywall)

Thank you, Dr. Buesseler! This information was already in the MAFF data, but someone with a bit of expert knowledge about fish types and habitats had to take the time to put it in context, visualize it for the rest of us, and publicize it. By grouping the results according to categories which reflect their ecological niches, and graphing the results simply, we can see that bottom-feeders (in particular those caught off Fukushima), as well as fresh water fish, remain very contaminated; pelagic species (which live near the surface) are in general much less contaminated; and epipelagic (habitat from surface to 200 meters below) and neuston (mainly coastal, surface dwelling species) have basically been under the government’s 100Bq/kg limit since around Sept. 2011. It’s particularly important to know about the bottom-feeders, because the fact that their contamination levels have not been decreasing much indicates how contaminated the seafloor off Fukushima Prefecture is (further research that Buessler and his colleagues participated in shows that the main mechanism keeping the seafloor and bottom feeders contaminated is plankton poop which drifts down and is eaten by the fish. See: Woods Hole symposium presentations). Why didn’t MAFF lay it out this clearly? Why, as in the case of the WBC data, did it have to wait until someone outside the government, who didn’t have responsibility for it, stepped up, spent the time, and gave it the required thought?

If you’re beginning to see a pattern here of the government either being less than totally helpful or actually making things ADAP, there’s more….

to Part 2

7ads6x98y