PROPOSALS FOR DISCUSSION TOPICS (C.Lourenco)

Regarding the specific physics topic of quarkonium production and polarization, we already had some
very nice discussions among us during the workshop in Palaiseau, just after ICHEP.  And we agreed at
that moment to "keep in contact" and share our questions and answers to ensure that our experiments
will provide high-quality measurements (and hopefully consistent with each other).  A few emails have
been exchanged and some nice discussions took place among some of us in Eilat, during the Hard Probes
2010 conference.  The meeting in Dec. 3rd offers an ideal setting to have more discussions and learn from
each other what the prospects are, which problems we have found, how we solved them, etc.  There are a
few "trivial details" which we can quickly review.  For instance, when we give differential cross sections as
a function of pT, how do we calculate the <pT> in each bin?  Before or after efficiency corrections?  And do
we subtract the contribution from non-J/psi events in the J/psi peak in the calculation?  And do we use the
reconstructed pT or do we apply some method to recover the "generated" pT?  If so, do we use an "ansatz"
method or something else?  Then, there are less trivial details like how we "normalize by the width of the
rapidity bin" to calculate a "double differential" cross section, dsigma/dpTdy, which can more easily be
compared to other experiments or theory.  We cannot simply divide by the bin width, given the limited
acceptance in the bin.  And how do we take into account the polarization of the J/psis from B decays in the
extraction of the "J/psi B fraction"?  Do we all use EvtGen?  I am sure that we can collect a nice list of all
the details that are important to make a solid comparison of the cross sections obtained by each LHC
experiment, and with theory.  If you tell us about your items, we could prepare a list of how those "details"
are dealt with in each experiment's analyses.  We could then go through the list and see if the differences
are important or not.

An even more interesting point we should discuss at the meeting is what we have already learned from the
existing results, of relevance to the other experiments.  For instance, ALICE does not yet have enough data
in the barrel (electrons) to measure a pT distribution of the J/psi and verify that the function used to calculate
acceptances is suitable.  They can use a CDF-inspired function (if the energy dependence is small) or the
function they measure in the forward muon arm (if the rapidity dependence is small).  By looking at the data
in several rapidity bins, CMS and ATLAS can see that the more "mid-rapidity" the events are, the harder is
their pT distribution, and ALICE can benefit from this information.  Also the comparison between LHC data
and CDF data for the J/psi B fraction is quite "illuminating": no strong rapidity dependence has been seen so
far and the increase from 2 to 7 TeV is rather small.  It would be nice to prepare new "compilations", as those
which Hermine did for the workshop in Palaiseau, with the latest results from each experiment, profiting from
increased data samples.  What other quarkonium-related results should we compare among us?

The more difficult questions we are facing in our work are, as far as I know, "how do we take into account the
polarization inputs" and "how do we compare the measurements with theory".  Concerning the polarization
issue, we should review in the meeting what we have learned since ICHEP.  CMS explored four "extreme"
scenarios: full transverse and full longitudinal polarizations in two different frames.  ATLAS stayed in one
frame but added extreme values for the azimuthal anisotropy, a very nice idea.  Then some of us gave the
results in different columns of a table, others presented the polarization "uncertainty" as an error band.  We
should see if we have now better procedures, aimed at combining the results of the different experiments
in a consistent way.

The data-theory comparison is the most difficult problem to solve.  The "best model in the market", NRQCD,
provides direct J/psi production cross sections, versus pT, and includes a certain polarization trend, which
becomes increasingly transverse as pT increases.  The data, so far, corresponds to prompt production,
including feed-down contributions from psi(2S) and chi_c1 and chi_c2, and the corresponding states for the
Upsilon family.  We do not know the NRQCD prediction for the pT dependent feed-down contributions and
we do not know the polarizations of the 2S and 1P states, an important ingredient to calculate their respective
acceptances.  My own talk at the workshop in Palaiseau addressed these problems (feed-down and
polarization unknowns, and how they affect the comparisons between data and theory).  In short, we are in a
situation where we can probably get an excellent agreement between our data and "NRQCD" as long as we
select a suitable set of assumptions regarding feed-down yields and their polarizations (versus pT). 

What can be done to improve this situation?  Restricting the data-theory comparisons to Pythia is not really a
suitable solution.  Nobody believes that the Pythia settings / code are particularly good at describing quarkonia. 
If they are, it would be mostly by chance than because of some enlightening idea of our Pythia friends.  In this
topic we really need more interactions with the theorists making calculations.  By the way, Jean-Philippe might
have news on calculations of the psi(2S), etc, in the CSM.  CASCADE seems to be a good complement to
PYTHIA.  I would like to learn more about it.  Maybe we could ask Hannes to give a short talk at the meeting.
Also the CEM, complemented with FONLL calculations for the open beauty part, seems to be doing a pretty
good job at describing the J/psi B fraction vs. pT and rapidity (Ramona Vogt and Randy Nelson, at HP2010).