high-Z impurity correction for Run8

[Bernhard Lauss <lauss@berkeley.edu>]

Dear Tom et al,


I have to clarify one misunderstanding, as obviously my notes
/ spreadsheets are misleading.

I did not assume that the Run8 clean-fill impurity content
is mainly nitrogen.
You probably have taken that from reading the first input
("only nitrogen as impurity")
on my spreadsheet / full calculation on page

http://weak0.physics.berkeley.edu/weakint/research/muons/private/bernhard/analysis_10May06.pdf
http://weak0.physics.berkeley.edu/weakint/research/muons/private/bernhard/highz-analysis-run8-10may06.html

This was only meant and used to estimate an error bound.


The next line says
realistic "target content"
which claims that out of the 12.592 pm of high Z impurity
yield in the Prod-50 data only 0.9 ppm are due to nitrogen
(relying on the consistent gas-chromatographic analysis on N2)
the rest of the yield is due to (most likely) oxygen.
This I used as my realistic assumption for Run8 and for the
estimate of the correction. - and to estiamte a fictious
oxygen content to be compared to the humidity monitor readings of Run9/10


In order to get another different error bound, I used the
assumption of having even argon inside, making up for
half of the unknown yield, - which is clearly
not supported by chromatography data, however I tried to
get a bound from a faster rate, too. "argon = worst1" case on the
spreadsheet) / carbon = worst2.)


So I think the impurity compositions
which we saw in Run9 and 10 are the only guidance for Run8
which we have beyond nitrogen chromatography,
and that's why I used the Run9 to estimate the effect on Run8 data,
fully agreeing with Claude's statement, that the differences
are not so large, at least between Run9 and Run8.
Run10 might have been better.

I agree with you, that the detection efficiency for oxygen
is a real problem concerning the error, that was exactly
one major point why I was pushing the water measurement
so strongly to be done during Run10.

Hope this clears up your misunderstanding caused by my
non-existing description of what is on the spreadsheet.


			Best regards

				Bernhard