Hi,
below are the minutes of the analysis meeting on 03/01/2006.
Minutes:
-Adrian has determined "kinematic corrections" from ep elastic data by
assuming that the electron angles theta_e,phi_e and reaction vertex z
are correct. Subsequently, p_e, p_p, theta_p, phi_p and z_p get
corrected. As a self consistency test he shows the invariant mass
spectrum W. In the raw W spectrum, a double peak with an offset from
the proton mass is seen, after correcting the width of the peak is
reduced and located at M_p (see corrected plot, there was a sign
error at the time of the meeting). The obtained W spectrum is also
compared with the MC result (energy loss turned on, MASCARAD once
turned off and once on.). The corrected spectrum resembles more the
unradiated MC (as energy loss and radiation is not accounted for in
the reconstruction, the correction effectively also covers these
effects). Thus, for the "correction" to become more "universal",
i.e. independent of the PID, the energy loss and radiative effects
need to be accounted for in the reconstruction first, before the
remaining "net" correction are determined. Energy loss and internal
radiation are nevertheless small compared to the size of the total
correction.
Adrian will generate his "set of corrections" both for the old
(v3_4_12) and new crunch version (RECRUNCHDIR=v3_4_14/5) and make it
available.
It is still a task to parameterize the momentum loss of a particle
(p,d,pi+-,e) as a function of its initial momentum or as a
two-dimensional function of momentum and angle (from comparison
of reconstructed radiated MC incl. energy loss with the tossed
momentum). This was declared a task already in January!
-Eugene showed how momentum and angle offsets, as well as the
reconstructed beam energy of real data compared with MC data, both
radiated and unradiated (energy loss turned on). In the plots of
reconstructed beam energy, the effect of energy loss is seen in MC in
form of a deviation at the lowest proton momenta (large proton
angles). He confirmed Chris' finding in the right sector where
momentum offsets of the proton and electron have opposite
signs. Eugene generated a MC with the middle chamber in the right
sector off-set by 1mm. The result for the reconstructed beam energy
resembles much the one for the real data, though not completely
(the resulting effect for the electron is not as big as in the real
data). The situation in the left sector remains not understood. While
the reconstructed proton variables almost agree with the MC
expectations, the electron momentum is off by a considerable
amount (30-50MeV). Note that protons and electrons cover different
regions in the wire chamber. Eugene tried to introduce some other
shifts/rotations that may explain the observation in the left sector,
yet inconclusive though.
-Discussion on further strategy: As said already last week, it is not
satisfying to just make an adhoc assumption for the geometrical shift
and do a recrunch (so far the analysis of elastic data has only shown
evidence for a sensitivity to a geometry effect). The straight track
analysis with zero field also sees a shift of ~1mm, however the
precision is not so good. The zero field runs have the disadvantage
of high hit multiplicity and thus bad resolution (the evidence of a
1mm shift first appeared only as a shoulder and only restrictive cuts
improved that somewhat). The other disadvantage is that each track
crosses only one sector, i.e. three chambers, so the redundancy is
quite small.
In using cosmic rays instead, the above two disadvantages are
overcome: these tracks have multiplicity of one, and one may find
enough of them that cross the entire detectors through all six
chambers left and right. We are looking for global shifts of one (or
two) chambers relative to the other ones... a chi2 analysis
(minimization with respect to an assumed offset) of these
straight tracks are much more likely to give a unique result.
Alternatively, for each of the six chambers one can histogram the
distance of the actual hit (or segment) from the crossing point
of the fitted track at that considered chamber. Any geometrical shift
should accumulate to a sharp peak in such a histogram.
-Timelines
It is of top priority to pursue the cosmic ray analysis in a timely
manner. It was considered that this analysis should be conclusive by
mid March. The recrunch with the improved geometry will then be ready
before the end of the month. The residual kinematic offsets need to
be evaluated (and determined) and made available. We'll have a
collaboration meeting on April 7, 2006. We should be able to see how
the improvement is influencing the physics results. Upon the April
meeting, we should become very clear on the necessary actions to
finalize the results for
It is suggested to have another collaboration meeting early or mid
May before the conference season start. At this point, reliable results
are expected.
-There is some ongoing discussion on convergence and reliability of
the currently used tracking algorithm for real data that I raised
last week.
-The spin angle issue will be discussed in the next meeting.
-Inclusive p(e,e') (OF)
Tavi showed us inclusive yields and asymmetries. The inclusive yield
is obtained from summing triggers 1,2,3 and prescalefactor times
trigger 7, regardless wether trigger 2 or 3 were prescaled. Trigger 3
contributes on a very low level anyway.
Yields with full and with empty target were shown as a function of
invariant mass for all Q2 combined up to 0.35 (GeV/c)^2. The (double
polarization) asymmetry was also shown vs. W. However, Tavi's raw
asymmetries in the elastic region seemed to disagree with the elastic
asymmetries of Chris and Adrian. Also shown was a combined MC for
elastic scttering + MASCARAD plus MAID2003 for pion
production. Again, in the elastic region there was disagreement
between MC and data which needs to be resolved first. We discussed
normalization procedure: MC yields for MAID and elastic are combined
based on the relative cross sections. The normalization of MC to data
should be done based on elastic yields where the physics is well
known. Likewise, inelastic cross sections can be deduced from
normalization to the elastic yield and assuming the elastic cross
section theoretically. This inelastic normalization is modulo the
ratio of the electron detection efficiency for elastically and
inelastically scattered electrons.
Best regards,
Michael
+-------------------------------------+--------------------------+
| Office: | Home: |
|-------------------------------------|--------------------------|
| Dr. Michael Kohl | Michael Kohl |
| Laboratory for Nuclear Science | 5 Ibbetson Street |
| MIT-Bates Linear Accelerator Center | Somerville, MA 02143 |
| Middleton, MA 01949 | U.S.A. |
| U.S.A. | |
| - - - - - - - - - - - - | - - - - - - - - -|
| Email: kohlm@mit.edu | K.Michael.Kohl@gmx.de |
| Work: +1-617-253-9207 | Home: +1-617-629-3147 |
| Fax: +1-617-253-9599 | Mobile: +1-978-580-4190 |
| http://blast.lns.mit.edu | |
+-------------------------------------+--------------------------+
This archive was generated by hypermail 2.1.2 : Mon Feb 24 2014 - 14:07:33 EST