[G12] resolution of tagger in kinematic fitter

[Lei Guo]

So I think we have consensus in that it is undercounting if we r just using 0.001*E_beam.... Ur point of the mcc energy being different is also consistent with the fact that we had to do a run-by-run e_beam correction. Technically, we probably should use something around 0.002*e_beam in the kinematic fitting. Agree?

Sent from my iPhone

On Feb 10, 2016, at 16:35, Eugene Pasyuk <pasyuk at jlab.org<mailto:pasyuk at jlab.org>> wrote:

This would be a double counting for high energy end. The offset form the nominal bin centroid is always less than the a bin width.
What needs to be adjusted  is E_beam. We always had it deviate from what MCC reports by ~0.1-0.2%, and even up to 0.5% occasionally. But this is multiplicative factor and affects all Eg the same way.

-Eugene

________________________________
From: "Lei Guo" <leguo at fiu.edu<mailto:leguo at fiu.edu>>
To: "Eugene Pasyuk" <pasyuk at jlab.org<mailto:pasyuk at jlab.org>>
Cc: "Carlos Salgado" <salgado at jlab.org<mailto:salgado at jlab.org>>, "g12 at jlab.org<mailto:g12 at jlab.org> g12 at jlab.org<mailto:g12 at jlab.org>" <g12 at jlab.org<mailto:g12 at jlab.org>>
Sent: Wednesday, February 10, 2016 4:22:24 PM
Subject: Re: [G12] resolution of tagger in kinematic fitter
I understand that the 2007 paper is about the correction and now the correction gets applied at cooking. However, there is uncertainty to that correction. So on top of the 0.001*e_beam, there should be additional 0.001*e_gamma, for the overall resolution.

Sent from my iPhone

On Feb 10, 2016, at 15:52, Eugene Pasyuk <pasyuk at jlab.org<mailto:pasyuk at jlab.org>> wrote:


You are confusing two different things.
The tagger reconstruction returns the photon energy as a centroid of the E_bin. The bin widths is almost a constant and is ~0.001*E_beam. The beam centroids originally used are coming from the geometry of the tagger hodoscope and magnetic field.
During g10 and g11 it was discovered that bin centroids are not what we thought due to gravitational sag of the focal plane.
To determine the deviation from the expected values two methods were used. One is the calibration you refer to, the other using kinematic fit of g11 data. What you see in the tagger calibration paper is the accuracy of the centroid offset from ideal.
Once this calibration was done we implemented those corrections in the tagger reconstruction. So, in g12 cooked date it is already accounted for. So, this is not and uncertainty but correction.
What is left is the bin width. The photon energy could be anything E_bin(i)-Ebin_width/2<Eg<E_bin(i)+Ebin_width/2
The calibration gives us E_bin(i), but effects of the bin width should go in the kinematic fit. Ebin_width ~ Ebeam*0.001 and it is essentially a constant over focal plane. Carlos' note describes how to get the variance assuming uniform energy distribution within E_bin.

-Eugene

________________________________
From: "Lei Guo" <lguo at jlab.org<mailto:lguo at jlab.org>>
To: "Carlos Salgado" <salgado at jlab.org<mailto:salgado at jlab.org>>, "Eugene Pasyuk" <pasyuk at jlab.org<mailto:pasyuk at jlab.org>>
Cc: "g12 at jlab.org<mailto:g12 at jlab.org> g12 at jlab.org<mailto:g12 at jlab.org>" <g12 at jlab.org<mailto:g12 at jlab.org>>
Sent: Wednesday, February 10, 2016 3:01:22 PM
Subject: Re: [G12] resolution of tagger in kinematic fitter


The uncertainty in the energy of the photon is 0.001* E_beam (per NIM and Eugene P.)


I copy and pasted from the 2007 tagger energy calibration NIM paper ():

"In this report, we present the energy calibration of the Hall B bremsstrahlung tagging system at the Thomas Jefferson National Accelerator Facility. The calibration was performed using a magnetic pair spectrometer. The tagged photon energy spectrum was measured in coincidence with e+e- pairs as a function of the pair spectrometer magnetic field. Taking advantage of the internal linearity of the pair spectrometer, the energy of the tagging system was calibrated at the level of ?0.1%E? . The absolute energy scale was determined using the e+e-e+e- rate measurements close to the end-point of the photon spectrum. The energy variations across the full tagging range were found to be < 3MeV"




The hodos were built such that this is true (different widths)

Then consider a counter of width (energy) "a"  where the (assume) energy distribution is flat: then Variance = sigma^2 = 1/a* int^a_0 E^2 dE = a^2/3

Therefore if we distribute the uncertainty over a counter (energy bin) the variance is : sigma^2 = (0.001*E_beam)**2/3

=Carlos
_______________________________________________
G12 mailing list
G12 at jlab.org<mailto:G12 at jlab.org>
https://mailman.jlab.org/mailman/listinfo/g12<https://urldefense.proofpoint.com/v2/url?u=https-3A__mailman.jlab.org_mailman_listinfo_g12&d=AwMFaQ&c=1QsCMERiq7JOmEnKpsSyjg&r=hHzTZynKAcZP6Z9X2yfO7w&m=PiuHY8u-uXDfqVhqiOlRNHG5wet7mC3FwA6T0TgQw6o&s=t4ElkvwyKE7_HNwh-jzSttu-xdqUGQrlBrh-PUySwL0&e=>



-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://mailman.jlab.org/pipermail/g12/attachments/20160210/5b1c3854/attachment-0001.html>