ALICE HLT Jet Studies
Content:
7. Results
7.1 Results from standalone tests of the different jet-finders
7.1.1 Cone Algorithms
7.1.2 kT Algorithms
7.1.3 Charged Jet
Algorithm
7.1.4 Hijing Background Studies
7.1.5 Trigger Algorithm studies
7.2 Combined results
7.2.1 Trigger Efficiciencies
and Background Rates
7.3.1 Parts of the Review Talk given at Cern
7.3.2 Review Page at Cern
A. Status/Updates...
As usual, preliminary and
incomplete!
01/15/02: Added CDR review talk - see section 7.3.1
Some slides shown at the Panel Review.
01/09/02: Added CDR section - see section 7.3
Results and parts of the corresponding comments that went into the CDR.
11/15/02: New results on HIJING background - see section 7.1.4
First estimate of trigger background rate as function of nParticles,PtThreshold
and Cone size.
11/14/02: New results on HIJING background - see section 7.1.4
Preliminary results on number of charged particles/cone in hijing events
(trigger background rate).
11/13/02: Updates results from the cone jet finder - see section 7.1.1
Has now study of trigger efficiency with the proposed trigger algorithm
for different trigger algorithm parameters.
11/07/02: New section on HIJING background studies - see section 7.1.4
11/05/02: First results from the cone jet finder - see section 7.1.1
10/31/02: First results from the charged jet finder - see section 7.1.3
10/30/02: Let's start again - divided web page into different
ones...
Lot's of work in the meantime, but webpage not updated...
09/22/02: First results from HIJING / updated PYTHIA results
- new section 7
Still w/o merging/joining and cone finder only... Added figures
on protojet properties for PYTHIA. First results on central HIJING events,
same figures as for PYTHIA events. Figures are now links to eps versions.
09/19/02: First description of models - new section 4, details still missing...
09/18/02: Include comments from Constantin - new section 1, lots of changes in the
other sections...
09/05/02: First version of cone finder implemented
- first results on Pythia simulation in
section 7 :-)
So far w/o merging/joining, still no description of algorithm
in section 5, no description of Pythia parameters...
Updated comments on TPC &
HLT simulations...
Complete ALICE simulations are coming - maybe these are the
ones to use (makes proposal in section
6.3 unnecessary)
08/29/02: First version of this page:
So far only section 6
(description of ALICE simulation enviroment), I'm currently working
on implementing/testing the different jet finders (short description
in section 5). For the moment I'm working only
with PYTHIA simulations, as soon as I get reasonable results, I show them
in section 7 and complete the model section
on PYTHIA (4.1) and the description of the
algorithms...
1. Abstract/Physics
motivation
The goal of jet spectroscopy in heavy ion collisions is the
measurement of the radiative energy loss of the leading partons traversing
QCD matter e.g. jet attenuation. The standard approach would be to
compare high Et inclusive jet production cross section in heavy ion collisions
(Pb+Pb) and elementary hadron collisions (p+p), so that one gets the
jet suppression factor by applying QCD scaling.
The design of the ALICE DAQ is such, that one can record about
10 central Pb+Pb events of the TPC. The TPC delivers up to 200 good
events (pileup free + DDL bandwidth saturates) per second at design luminosity.
Mangano et al. predicts from QCD a jet production rate of
around one jet with Et in about 2000 such events where Et>100 GeV.
Jet attenuation could lower the rate by a factor 2-3, which means,
that we can roughly expect one >100GeV jet in 4000 to 6000 such events.
The standard DAQ will therefore roughly record a few 1000
of such jets per ALICE year. An online jet trigger at a rate of 200Hz
could improve the statistics by a factor of ~20.
2. Goals of
this study
Preliminary studies of jet-production in heavy ion collisions
at sqrt(s_NN)=5.5TeV for the ALICE-High Level Trigger are investigated
in order to:
- study different jet finders (jet definitions),
- test if they are appliable as trigger algorithm
The overal goal is to found a profund base for more careful
studies of HLT jet triggers.
3. Jet definition
- Since the scattered partons carry nonzero colour charge
and the observed hadrons do not, there can be no unique assosciation
of a jet of hadrons with a single initial quark or gluon.
"Jets are ambiguous"
or
"The jet definition depends on the used algorithm"
- In hadron-hadron collisions producing a high-pT jet,
only a small part of the final state hadrons is associated with
the jet. The other particles present in the event can be thought
of as associated with "beam jets", e.g. soft interactions of the remaining
partons in the incident hadron which are in first order uncorrelated with
the hard process (the jet). And this is even worse for heavy ion collisions,
since there a more "beam jets"...
8. Summary
From the present level of jet physics simulation, to be able to present an optimized
HLT cone jet finder trigger for inclusive jets of E_T > 100 GeV that delivers
75-80 % efficiency with a background suppression selectivity of 1-2 %.
The expected CPU time budget is surely below 10% of the effort
required for cluster finding and tracking.
Thorsten Kollegger, kollegge AT (no spam) ikf DOT uni-frankfurt DOT de
Constantin Loizides, loizides AT (no spam) ikf DOT uni-frankfurt DOT de
IKF - University of Frankfurt
Last updated: 01/17/2003