Introduction
T2KK
FNAL
NC
In this page, I want to present how I used the beam
nue background to test the normalization fo the overall background.
Reminder:
To estimate the background in T2KK, I run over the atmospheric
SK Monte Carlo, check if an event pass the precuts (FCFV, elike,
single-ring, no decay-e).
The using its reconsctructed energy, I check the likelihood
efficiency, and add it to a histogram with the corresponding weight.
I also renormalize that background by the ratio of T2KK/SK flux.
And for that I need to have the T2KK and the SK flux in same
units, as it is shown in:
http://hep.bu.edu/~fdufour/normalization/#bg_t2kk
In order to test that what my normalization is correct I can deal with
the beam nue BG by another way.
I can deal with is the same I deal with the signal.
Reminder:
I multiply the T2KK beam nue flux by the ccqe cross-section,
then using ratio of CCqe to CC I make a CC spectrum, then I smear to
account for energy resolution, and apply likelihood efficiency. (see
report for more neat detailed explanation).
So comparing those two methods should be a good
cross-check that my normalization is right.
In order to eliminate other errors, I first looked at beam nue CCQE
events, I used the true energy, and I didn't apply the likelihood
efficiency, then I compared to two event spectrum.
Each event spectrum
The first plot is done with MC (no
precuts/no likelihood efficiency)
The second plots is done by just multipling the T2KK nue beam flux
by the nue CCQE cross-section.
Ratio
This is just the ratio of the two
plots above, clearly there is a very good agreements.
Now that I finally trust both method, I kept going and I apply
likelihood efficiency and the whole mahcinery to get real event
spectrum plots and here is what I get:
MC
method
Beam
method
Really good agreements.. I think I solved my problems in T2KK!!!!!!
For FNAL
I get similar results but the ration is on 1.2 and not 1 as you can see
there:
Event spectrum
ratio
Which is not that but, but not as clean as for T2KK.
Also when using this normalization, the total background becomes very
big.
Milind
Me
Total BG 566
2185
Beam nue 272
361
Knowing that I might be a factor of 1.2 off in my normalization, the
beam nue background could be in agreement, but the NC clearly is not.
Neutral
current BG
For Neutral current background the situation is more tricky. If I do
exactly the same thing than for beam nue, with exactly the same
normalization factor, here is what I get:
Neutral current plots:
Event spectrum
The first plot is done with MC (no
precuts/no likelihood efficiency)
The second plots is done by just multipling the FNAL numu flux
by the numu NC cross-section.
ratio
So clearly something is not working properly but it is not just a
normalization factor.
Guess: the NC cross-section is different than asking mode.gt.30 in an
ntuple...
Still working...
