1. New Approaches to Hadronic Final State Reconstruction
David W. Miller
SLAC ATLAS
July 7, 2008
SLAC Annual Program Review

2. SLAC Annual Program Review
Overview
Standard model measurements and searches for new physics will both rely heavily on the understanding of the jet energy scale, missing energy measurements and the identification of b-jets.
Through the Jet/MET/bTag Working Group (formed and led by Ariel Schwartzman) we have developed and applied new techniques for understanding and improving jets, missing transverse energy and heavy flavor in ATLAS and in collaboration with several other institutes.
Extending jet algorithm functionality and flexibility for first data
New approaches to jet reconstruction and missing energy through use of tracking information
Jet-vertex association for high luminosity jet identification, jet-energy corrections and primary vertex selection
Semileptonic b-jet energy corrections
Missing ET significance
July 7, 2008
SLAC Annual Program Review

3. Extending jet algorithms for first data
Legacy jet algorithms
Tower algorithm (TowerJets)
Built from geometrical arrangements of calorimeter cells
No noise suppression is applied
Topological clusters (TopoJets)
Built from topological clusters of calorimeter cells
Noise suppression is applied
Clusters size can grow non-linearly in presence of pile-up
New approach in ATLAS
Use the most desirable aspects of each algorithm
Towers are finite, well-defined objects
Clusters efficiently select cells above a dynamic noise threshold
Build towers using only cells used in topological clusters
Well-behaved at high-luminosity
More easily understood in first data
(DWM, SLAC)
July 7, 2008
SLAC Annual Program Review

4. Advances in jet resolution using tracking information
New approach in ATLAS: use tracks to provide additional information on jet composition
Account for jet-to-jet fluctuations in charged particle content
(Marshall, Columbia U.)
Jet structure
Jet response depends on ftrk
Improvements with ftrk correction
Jet response improved using the charged energy fraction ftrk
ftrk provides information on jet fragmentation, particle content and topology and permits jet energy corrections after all other jet-energy scale corrections are applied
July 7, 2008
SLAC Annual Program Review

5. Jets & tracks: improvements and implementation
Jet resolution
Corrections applied in bins of ftrk
Jet-by-jet fluctuations are reduced
Improvement in overall jet energy resolution
Official ATLAS software package integrated into full reconstruction
Available to all ATLAS users for first data
Documentation in ATLAS internal note
ATL-COM-PHYS
Used in Higgs Z/Wjjγγ analysis
ftrk response corrections
Improvements to jet-resolution: 10% at 40 GeV
(Marshall, Columbia U.)
July 7, 2008
SLAC Annual Program Review

6. Missing energy applications
Improvements to jet-energy has implications for missing energy measurements
Using track information to identify a systematic effect in missing energy
allows us to correct for it
(Schwartzman, SLAC)
Artificial missing energy from pT imbalance is corrected following jet energy corrections
July 7, 2008
SLAC Annual Program Review

7. Tracks-jets as complementary to calorimeter jets
New approach in ATLAS
Build “jets” from tracks in Inner Detector tracker
Inherently 3D objects (Z, η, φ) unlike calo jets
(Schwartzman, SLAC)
These track-jets are completely independent from the calorimeter
Good angular resolution, used in b-tagging trigger (see I. Aracena’s talk)
Calorimeter jets can incorporate particles from multiple interactions
Using track-jets with Z information, can separate interactions
July 7, 2008
SLAC Annual Program Review

8. Improved jet-finding efficiency with track-jets
(Schwartzman, SLAC) Δ
ttbar events η
Limited by tracker acceptance
Improved efficiency for low-pT
The use of tracks provides complementary information to calorimeters
Track-jets allow for the measurement of jet reconstruction efficiency
Energy losses in the inner detector and cryostat decrease calo efficiency but do not affect track-jets
July 7, 2008
SLAC Annual Program Review

9. Identification of fake missing energy with track-jets
Unmeasured calorimeter jets results in artificial missing energy
Using track-jets to pin-point un-reconstructed jets, events with fake missing energy can be removed
Radius (m)
Gap at η=
Beam axis (m)
(Majewski, BNL)
Missing calorimeter jet
July 7, 2008
SLAC Annual Program Review

10. Jet-vertex association (JVF)
Expect >23 multiple simultaneous proton-proton interactions (extra jets and vertices).
Tevatron techniques to account for this discount large fluctuations in the form of jets
New approach: use tracking + calorimeter + vertex information to account for jet origin
We can use the vertex detector to reconstruct these additional vertices
Associate standard calorimeter jets to primary vertices using tracks
Obtain jet-by-jet energy correction for pile-up and jet-selection criterion for calo jets
Improve jet-energy, missing ET resolution and primary vertex (PV) selection
JVF measures the fraction of charged particle transverse (track) momentum in each jet from each identified primary vertex in the event.
ATLAS detector with pileup
(23 interactions per beam xing)
July 7, 2008
SLAC Annual Program Review

11. New approach to jet identification in ATLAS
Using JVF to select hard-scatter jets in events with pile-up
Without any QCD pile-up, expect certain jet multiplicity
As instantaneous luminosity increases, additional interactions contribute jets and energy to event
Using JVF we can reliably select jets from the signal vertex
ttbar events: 2x1033 cm-2s-1
(DWM, SLAC)
Using JVF we recover the flat jet-multiplicity distribution
without raising the jet pT threshold
In official ATLAS software and available to all ATLAS users for first data
Fully documented in internal ATLAS note: ATL-COM-PHYS
July 7, 2008
SLAC Annual Program Review

12. Jet-by-jet energy corrections for pile-up contributions
Jet energy response vs. JVF: 1033 & 2x1033 cm-2s-1 (ttbar)
Jet energy response vs. pT and JVF (ttbar)
(DWM, SLAC)
Using JVF to correct jet energies in a luminosity independent manner
Derive jet-energy correction similar to that already done for ftrk to correct jets for pileup contributions on a per jet basis (new technique at hadron colliders)
July 7, 2008
SLAC Annual Program Review

13. Using jet-vertex association to select primary vertices
Muons and JVF tagged correct PV
Improving primary vertex selection in events with multiple interactions by using information from reconstructed objects in the event
jet
Jet & μ-tag
The high luminosity environment in ATLAS will present challenges to correct hard-scatter vertex identification
Extremely important for b-tagging
New approach in ATLAS
Measure efficiencies in real data with high-pT lepton tagging
Default μ
jet
jet
(DWM, SLAC)
July 7, 2008
SLAC Annual Program Review

14. Semileptonic b-jet corrections
Correcting for missing neutrino energy
Jets with constituent ν’s have energies systematically underestimated by 10%
Add neutrino energy back into jet energy
b-jet response in ttbar events
In official ATLAS jet reconstruction software
Available to all users for first data
Documented in internal ATLAS note:
ATL-COM-PHYS
(Mateos, Columbia U.)
July 7, 2008
SLAC Annual Program Review

15. Use of semileptonic b-jet correction in ATLAS
Not only will the jet energy scale of b-jets occupy an integral role in many standard model measurements in early data (such as ttbar) but the preponderance of b-jet channels in SUSY and Higgs searches renders it even more important
Data-driven methods for estimating corrections under way
Technique applied in Higgs Working Group H(120 GeV)bbbar analysis
(Mateos, Columbia U.)
July 7, 2008
SLAC Annual Program Review

16. Missing energy significance
QCD Events (fake missing ET)
W+jets
(real missing ET)
(Perez, Columbia U., Butler, SLAC)
Define a likelihood ratio given the particular event topology (jet resolutions, jet energy and missing ET) to reject events with fake missing ET
Separate true missing ET from fake missing ET
Not simply MET/σ
July 7, 2008
SLAC Annual Program Review

17. Uncovered topics in this talk
Several active areas of research and techniques not covered due to time constraint!
Quark/gluon tagging: SUSY jet+MET background reduction
b/c quark-jet separation: b-jet background reduction
Gluon splitting (gluonbbbar) removal: large source of background for SUSY b-jet+MET
Di-jet resolution measurements in first data
July 7, 2008
SLAC Annual Program Review
Page 17

18. SLAC Annual Program Review
Summary
A. Schwartzman’s Jet/MET/bTag WG and the collaborating is providing a coherent, wide range of contributions to ATLAS Jet/MET performance and physics groups
New approaches to combined detector performance using tracking and calorimetry which maximize jet-energy precision for standard model processes & new physics searches
This work has provided many important tools and techniques for understanding the first physics data from ATLAS
Roadmap for jet-energy scale being planned with this work at the forefront and will be heavily involved in the data-driven approaches to measuring the jet-energy scale and applying these new techniques
The tools and methodologies developed by Ariel’s Jet/MET/bTag group improve jets, missing ET and b-jets and are already proving useful in both standard model studies as well as physics searches.
July 7, 2008
SLAC Annual Program Review