Strontium 90 CONFIRMED in US
http://blog.alexanderhiggins.com/2011/04/21/radioactive-fukushima-pluton...
you can access the actual EPA results yourself at the end of the article there is an EPA query link
Search
http://blog.alexanderhiggins.com/2011/04/21/radioactive-fukushima-pluton...
you can access the actual EPA results yourself at the end of the article there is an EPA query link
The EPA did NOT detect Strontium or Plutonium
That website is very misleading in reporting these EPA data. I did the same queries ("air-filter" results, by the way) and came across the same exact numbers for Sr-89, Sr-90, Pu-238, and Pu-239, and I found one huge missing piece necessary to understand the numbers.
There are two columns needed to understand the measured activity: not just the Result Amount, but also the Combined Standard Uncertainty. The CSU must be included when interpreting results. In order for a result to be considered a detection to 95% certainty or greater, the Result Amount must be at least twice the Combined Standard Uncertainty.
Not a single result for Sr-89, Sr-90, Pu-238, or Pu-239 is above twice the CSU. The closest result I found was for Pu-239 measured in San Francisco on 3/24. The result is (6.5±5.8) * 10^(-6) pCi/m^3, but this is not statistically significant.
If there are any results above twice the CSU for these measurements, then that would be something to report.
Another weird thing about those results is that many are negative, which is physically impossible. This means that they are nondetections, and the fluctuations around zero are due to statistics.
Be careful when interpreting data straight from a database without any explanation!
Mark [BRAWM Team Member]
Strontium 89 in Milk
I understand the explanation on the air filters. What do you think about the EPA finding strontium-89 in milk, in Hawaii?
Reported as a real value, per this link:
http://www.epa.gov/japan2011/docs/rert/RadNet-Sr-Milk-Public-Release-4-2...
It'd be nice if we could examine the Hilo Air data side by side with the milk data, but I don't know how this could really be done given the limited amount of samples from EPA. (I recall not being able to find anything on the air in Hawaii to relate it to the milk.)
Yes, thank you for
Yes, thank you for mentioning that. Strontium in milk as a real value.
Proof is the pudding or milk as the case may be. If it's the milk, it's in the air.
downplay, deny, rinse, repeat.
This calls into question ALL of the data
If several sites are showing - as you allege - a false positive, it would be logical to assume that the reverse can also be true: false negatives.
No, data is fine; understanding the data requires error bars
I am not alleging that those are false positives -- I am saying that they are very likely NOT true positives, which is what others have claimed. The data were misinterpreted because the error bars were neglected. The whole point of error bars is to quantify the likelihood of false positives and false negatives, and understanding them is crucial to interpreting data.
All of these measurements (EPA, BRAWM, etc.) are fundamentally statistical in nature. There is always an uncertainty in a measurement when you are counting things -- gamma rays, alphas, etc. All results must be reported with their corresponding error bars; a measurement doesn't mean anything without error bars. The error bars give a confidence interval that the true value lies within a certain range -- plus or minus the error bars is 68%, and plus or minus twice the error bars is 95%.
For example, if I report a measurement of 7±2, the true number is very likely not 0 -- we can say that there is a 95% chance that the true number lies between 3 and 11.
However, if I report a measurement of 1±2, then the true number has a 95% chance of being between -3 and 5. This means that 0 cannot be excluded. In this case one usually reports a non-detection and a detection limit called the minimum detectable activity (MDA), which is defined to be twice the error bars. Our MDA for this example is 2*2 = 4. That is how big the measurement would need to be in order to exclude 0 at 95% confidence.
In the case of the EPA data for plutonium and strontium, the confidence intervals of the measured values do not exclude 0 at 95% confidence. This means the data are very likely to be non-detections.
The measurements still give valuable information though -- they tell us that the true levels are below the MDA to 95% certainty. Or, another way of saying this is that the likelihood of a false negative is 5% (i.e., the probability that the true value is actually above the MDA).
Mark [BRAWM Team Member]
bandstra wrote... >Another
bandstra wrote...
>Another weird thing about those results is that many are negative, which is physically impossible.
Negative results would be possible if they are publishing relative results rather than absolute results.
Yes, but relative to what?
Yes, but relative to what? These are not found in nature, so it would be relative to zero. Negative numbers in this case is a statistical rather than a physical interpretation.
IF they are relative
IF they are relative numbers, they can be relative to anything the reporters want to pick as the reference point.
I just thought to check for that field header linking to a definition. Low and behold:
"Column Name: RESULT_AMOUNT
Description: The amount, activity, or concentration measured. Amounts can be zero, non-detect (ND), positive or negative. A negative result occurs when random effects and limitations in the measurement process cause the measured value for the sample to be less than that of the laboratory blank or background, which is subtracted from the sample measurement. Although negative radioactivity is physically impossible, the inclusion of negative results allows better statistical analysis of the data."
So it appears these are relative to "the laboratory blank or background". Perhaps this is a reasonable approach (radionuclide testing is not my thing) but if the results are being manipulated in such a way they should also provide "the laboratory blank or background" data so that it can be considered. I don't see that data being included in their results.
Berkeley Pu-239 Measurements Higher In Late 1970s/Early 1980s
So, Mark. Are we suppose to totally discount the SF results?
What causes the CSU to vary so much?
What is normal background Pu-239 in the SF area?
Just FYI:
From the EPA web site, it shows levels of Pu-239 in air much
higher in the late 70's and early 80's in Berkeley relative
to the latest SF measurement.
For May 15, 1981, they show a Berkeley measurement of 21.9
ACI/M3 (ACI = Attocurie or 10E-18 curies or 10E-6 pCI). And this
was with a CSU of 1.4. So, it should meet the 95% confidence
level.
The latest San Francisco measurement was 6.5x10E-6 PCI/M3.
(they switched to PCI vs ACI before). Which I believe is equiv.
to 6.5 ACI/M3. So, the May 15, 1981 Berkeley 21.9 ACI/M3
measurement was 3.4 times the Mar 15-18, 2011 measurement.
If we could have any confidence in the latest SF measurement.
Not sure what the normal
Not sure what the normal background is, but thanks for finding those earlier measurements (others -- be sure to enter the correct units in the search or they won't come up). You are absolutely right that many of those measurements are significant, including the Berkeley one that you mentioned.
I would phrase it that a significant detection of 21.9 +/- 1.4 aCi/m^3 was made in 1981. But the recent SF measurement ruled out any activity above 11.6 aCi/m^3 (this is twice the CSU of 5.8 aCi/m^3) with 95% confidence.
You had a couple questions at the beginning of your post:
(1) The SF results are not statistically significant and should not be considered a detection of plutonium.
(2) The CSU can very depending on a bunch of things -- how long the filter was run, how long the detector measurement was performed on the filter, the efficiency of the detector, the resolution of the detector, the efficiency of the filter, and the number of background counts in the region where the plutonium signature would be. So there's no surprise that those numbers vary by a lot.
Mark [BRAWM Team Member]
Thanks, Mark!!!
As always, really appreicate your response and hard work.
Thanks Mark. I'm breathing a
Thanks Mark. I'm breathing a sigh of relief!
we might keep in mind that
we might keep in mind that the measurements, although negligible, were from march. we are now almost to the end of april, and have been finding elevated (accumulated) levels of cesium as time passes. wonder if they'll do a more recent read?
Thanks for the quick reply.
Thanks for the quick reply. That is a bit more comforting. Does the method they are using (which sounds more like chemistry than measuring emissions) have drastically different uncertainty than Alpha spec?
Here's the link to the EPA
Here's the link to the EPA site if you want to run the query yourself
http://1.usa.gov/fHtcKa
I've used bitly to shorten the URL as it's very big.
BRAWM...dchivers...others...
Can you provide some perspective on this? Is this background or from Japan? This really concerns me.
BRAWM...dchivers...others...
Can you provide some perspective on this? Is this background or from Japan? This really concerns me.
tried it for Michigan,
tried it for Michigan, Missouri, Massachusetts and Florida- no results.
Positive for plutonium in Hawaii, Washington, California.
Any idea why not for the other states?
They aren't testing for it.
They aren't testing for it.
tried to do a query, just
tried to do a query, just gives me scrambled words. What do you put in this fields to get accurate results?
Just enter The desired
Just enter The desired isotopes in the dropdown for Radionuclides and Radiation Search, then enter the starting date of 3/15/2011 and end date of 4/21/2011 and then scroll down and run the query.
I also wanted to point out
I also wanted to point out that the EPA appears to be using extraction chromatography to detect the Plutonium.
Plutonium has been detected
Plutonium has been detected too. Most of the numbers are smaller than their uncertainty values, except for SF.
I was unable to run the query from before the 15th, so we don't have any sort of baseline value, but Plutonium shouldn't be detectable at all. BRAWM, care to comment?
I did a search and found
I did a search and found that as of 1964 Plutonium concentrations averaged 1-4 pCi/m2 across the northern hemisphere from weapons testing, accidents and satellite malfunctions.
If this is already in the soil, and China is reported to be having very bad dust storms lately, potentially stirring up radioactive dust, how do we even know the levels measured ate from Japan? There is no baseline data to judge. Thanks EPA.
I queried by my state and
I queried by my state and didn't find anything unpublished. Notice how they do attribute the Plutonium to Japan. The last two columns to the right are - Project ID "Japan Deploys" and Study ID "Fukushima Nuclear Incident"
Note: since March 18. Not
Note: since March 18. Not okay.
The labels are for the
The labels are for the deployable (mobile) sensors so the results are tied to where and when the sensors were deployed.
The fixed monitors, e.g. SF, show these values too, but don't have that label.