Basic Talk Info

Download Report

Transcript Basic Talk Info 

Top Quark Mass
Measurements
at Hadron
Colliders
G. WATTS (UW/SEATTLE, CPPM)
For the DZERO, CDF, CMS,
and ATLAS collaborations
July 15, 2014
The Top Quark
2
Just like other Fermions
Except:
π‘šπ‘‘ ~40 × π‘šπ‘
The next
heaviest
quark!
The Mass gives the top
quark a special role in the
Standard Model
β€’ Only fermion which has a significant coupling to the Higgs
β€’ Plays key roll in many important physics processes
β€’ Flavor physics, Electro-weak processes
β€’ It plays a special roll in a number of Beyond the Standard Model
theories as well
The Top Mass
We have known π‘šπ‘‘ almost since it was discovered.
By far the most precisely measured quark mass!
While it behaves like any other quark in the Standard
Model, its mass gives it a unique role.
β€’ Only version for which the coupling to
the Higgs is important
β€’ Stability of the SM Higgs potential
at high scales
A consistency check for the
Standard Model!
β€’ Shows up in a number of production
loops
β€’ 𝑔𝑔 β†’ 𝐻 at the LHC contains a top
loop
β€’ Heavy Flavor physics (e.g. 𝐡𝑠 β†’ πœ‡πœ‡)
production
3
4
Current World Average: 173.3 GeV.
Known to better than 0.5 %!!
Higgs mass is known to better than 0.3%
Top is easier to discover:
πœŽπ‘‘π‘‘ = 6.8 𝑝𝑏 at 𝑠 = 1.98 TeV
𝜎𝐻 = 0.6 𝑝𝑏 at 𝑠 = 1.98 TeV
Top is harder to reconstruct:
No clean easy to see peak like 𝐻 β†’ 𝛾𝛾!
All final states involve jets
Each measurement
deserves at least a
seminar
I have chosen a
few extra results
LHC
Tevatron
5
Decays
6
𝑑𝑑 β†’ π‘Š + π‘π‘Š βˆ’ 𝑏
Classified by the Ws’ decay
Dilepton events
Clean, but low statistics
~4%
Lepton + Jet events
Good compromise
Reasonable background
~30%
All Hadronic events
Huge multi-jet background
~44%
Top mass has been measured
in all decay channels.
The Tevatron & The LHC
The Tevatron is coming out with its final results
β€’ 10 𝑓𝑏 βˆ’1 of data at 𝑠 = 1.96 TeV
β€’ Well understood detector
β€’ Sophisticated analysis techniques
The LHC is just coming online in the π‘šπ‘‘ world
β€’
𝑠 = 7 TeV results well developed
β€’ 8 TeV results just appearing
β€’ Statistics are much better due to the much higher πœŽπ‘‘π‘‘
The much larger statistics will eventually open the
door to new π‘šπ‘‘ measurement techniques.
7
Extracting π‘šπ‘‘ from Data
Detector gives you 4-vectors. Use Griffiths!
β€’
β€’
β€’
β€’
β€’
β€’
Does not always give you 4-vectors (neutrinos!)
Detector/Object resolutions (e.g. Jet Energy Scale)
Background contamination
Incorrect reconstruction (e.g. bad jet assignment)
Top mass width
Etc.
Two common methods to address this:
Matrix Element
Uses all the information
Computationally very expensive
Template Method
Flexible, subsets the information used
β€œFairly easy” to implement
What do we measure? The Pole mass? The MC mass?
8
The Jet Energy Scale
Common curse for all methods
Lepton+Jets
β€’ Experiments normally measure in
independent control sample.
β€’ Resolution not good enough for a stateof-the-art top mass measurement.
In situ Jet Energy Scale measurement
π‘Š β†’ π‘ž π‘žβ€²
Two poorly
measured
objects
One very well
measured
object
Many techniques will
constrain π‘šπ‘žπ‘žβ€² to be π‘šπ‘Š as
part of the global fitting
process.
Global fit over the full sample
β€’ Scale all jets by a constant
factor to achieve constraint
Flavor Jet Energy Scale
9
The Matrix Element
Approach
10
A reverse Monte Carlo
MC
Generates
100K events
Distributions of
kinematic
variables for all
objects
β€œMap of
kinematic
phase space”
Turn that around
Given a single event in data, how dense a part
of kinematic phase space is it in?
Repeat for all major backgrounds and signal: 𝑃 π‘šπ‘‘π‘œπ‘ , π‘ƒπ‘π‘˜π‘”
𝑃
ME – Multiple Steps
ALPGEN +
Pythia
𝑃 π‘šπ‘‘π‘œπ‘ =
Detector
Simulation
1
𝑑𝑑
πœŽπ‘œπ‘π‘ 
π‘šπ‘‘π‘œπ‘
Reconstruct
ion
11
4 vectors of
reconstructed
objects
24
𝑀𝑖
𝑖=1
Normalization
Sum over all possible jet
assignments
β€’ Which jet is the first tops?
β€’ Which jets belong to the W?
A weight reflecting
the probability of
those jet assignments
β€’ 𝑏-tagging
probabilities
ME – Multiple Steps
ALPGEN +
Pythia
𝑃 π‘šπ‘‘π‘œπ‘ =
Detector
Simulation
1
𝑑𝑑
πœŽπ‘œπ‘π‘ 
π‘šπ‘‘π‘œπ‘
Reconstruct
ion
12
4 vectors of
reconstructed
objects
24
𝑀𝑖
𝑦
𝑦
π‘‘πœŒπ‘‘π‘š12 𝑑𝑀12 π‘‘π‘š22 𝑑𝑀22 π‘‘πœŒβ„“ π‘‘π‘ž1π‘₯ π‘‘π‘ž1 π‘‘π‘ž2π‘₯ π‘‘π‘ž2
𝑖=1
10 dimensional integral over phase space
β€’ Mass of the tops, W’s
β€’ Directions of the b-quarks
β€’ Lepton and neutrino direction
Note no mention of data 4-vectors yet!
ME – Multiple Steps
ALPGEN +
Pythia
𝑃 π‘šπ‘‘π‘œπ‘ =
Detector
Simulation
Reconstruct
ion
13
4 vectors of
reconstructed
objects
24
1
𝑑𝑑
πœŽπ‘œπ‘π‘ 
π‘šπ‘‘π‘œπ‘
𝔐𝑑𝑑
𝑀𝑖
𝑦
𝑦
π‘‘πœŒπ‘‘π‘š12 𝑑𝑀12 π‘‘π‘š22 𝑑𝑀22 π‘‘πœŒβ„“ π‘‘π‘ž1π‘₯ π‘‘π‘ž1 π‘‘π‘ž2π‘₯ π‘‘π‘ž2
𝑖=1
2
π‘π‘Žπ‘Ÿπ‘‘π‘œπ‘› π‘“π‘™π‘Žπ‘£π‘œπ‘Ÿπ‘ ,𝜈
Sum over incoming parton flavors
All neutrino solutions
The Leading Order Matrix
Element
β€’ Given all the phase space
parameters
β€’ Weight for the kinematics
values
β€’ Uses all available information
β€’ At leading order
ME – Multiple Steps
ALPGEN +
Pythia
𝑃 π‘šπ‘‘π‘œπ‘ =
Detector
Simulation
4 vectors of
reconstructed
objects
Reconstruct
ion
24
1
𝑑𝑑
πœŽπ‘œπ‘π‘ 
π‘šπ‘‘π‘œπ‘
𝔐𝑑𝑑
π‘π‘Žπ‘Ÿπ‘‘π‘œπ‘› π‘“π‘™π‘Žπ‘£π‘œπ‘Ÿπ‘ ,𝜈
14
𝑀𝑖
𝑖=1
2
𝑦
𝑦
π‘‘πœŒπ‘‘π‘š12 𝑑𝑀12 π‘‘π‘š22 𝑑𝑀22 π‘‘πœŒβ„“ π‘‘π‘ž1π‘₯ π‘‘π‘ž1 π‘‘π‘ž2π‘₯ π‘‘π‘ž2
PDF’s
𝑓 β€² π‘ž1 𝑓 β€² π‘ž2
2
𝛼 𝛽
πœ‚π›Όπ›½ π‘ž1 π‘ž2
Transverse
momenta of
incoming partons
Ξ¦6
βˆ’ π‘šπ‘ž21 π‘šπ‘ž22
Phase Space Factor
ME – Multiple Steps
ALPGEN +
Pythia
𝑃 π‘šπ‘‘π‘œπ‘ =
Detector
Simulation
4 vectors of
reconstructed
objects
24
1
𝑑𝑑
πœŽπ‘œπ‘π‘ 
π‘šπ‘‘π‘œπ‘
𝔐𝑑𝑑
π‘π‘Žπ‘Ÿπ‘‘π‘œπ‘› π‘“π‘™π‘Žπ‘£π‘œπ‘Ÿπ‘ ,𝜈
Reconstruct
ion
15
𝑀𝑖
𝑦
𝑦
π‘‘πœŒπ‘‘π‘š12 𝑑𝑀12 π‘‘π‘š22 𝑑𝑀22 π‘‘πœŒβ„“ π‘‘π‘ž1π‘₯ π‘‘π‘ž1 π‘‘π‘ž2π‘₯ π‘‘π‘ž2
𝑖=1
2
𝑓 β€² π‘ž1 𝑓 β€² π‘ž2
2
𝛼 𝛽
πœ‚π›Όπ›½ π‘ž1 π‘ž2
Ξ¦6 π‘Š(π‘₯, 𝑦; 𝜌, πœŒβ„“ , … )
βˆ’ π‘šπ‘ž21 π‘šπ‘ž22
Transfer Functions
β€’ Given a generated jet with 𝑝𝑇 , πœ‚ what is the probability DZERO
will reconstruct values x and y?
β€’ Detector and reconstruction resolution
DZERO π‘šπ‘‘ using the ME
Method
16
In used at DZERO since Run I
β€’ Use different top mass in
the Matrix Elements
β€’ Vary the Jet Energy
Scale in the transfer
functions
174.98 ± 0.58 π‘ π‘‘π‘Žπ‘‘ ± 0.63(𝑠𝑦𝑠) GeV
Total error is equivalent to
March world average!
3 years of work (old result):
176.01 ± 1.01 π‘ π‘‘π‘Žπ‘‘ ± 1.29(𝑠𝑦𝑠) GeV
3.6 𝑓𝑏 βˆ’1
What Did 3 years get?
β€’ Speed (CPU) to allow better MC stats
β€’ X100 increase means MC stats error
drops from ~0.25 GeV to ~0.05 GeV.
β€’ New Jet Energy Scale Calibrations
β€’ ISR modeling
β€’ Constrain by studies in Drell-Yan data
The πœ™πœ‚βˆ— variable is
sensitive to Z boson
recoil (𝑝𝑑 ).
Gives an
experimental bound
to ISR mis-modeling
Systematic error on π‘šπ‘‘ reduced from ~0.25 to 0.06 GeV
β€’ General 𝑑𝑑 modeling improvements
17
Template Method
18
Using a distribution sensitive to π‘šπ‘‘ :
Make it for each sample
Simulated sample at π‘šπ‘‘ = 167.5 GeV
Simulated sample at π‘šπ‘‘ = 172.5 GeV
Simulated sample at π‘šπ‘‘ = 177.5 GeV
Use a likelihood to
estimate template
compatibility
π‘šπ‘‘
Can do in two dimension
β€’ Jet energy scale
β€’ Top mass
Top Mass In Dilepton
Events
19
4% of all decays, split into
𝑒𝑒, π‘’πœ‡, and πœ‡πœ‡.
Very little SM background!
CDF’s basic selection: Observe 520 events, expect 78% purity
ATLAS’ basic selection: Observe 2913, expect 96% purity
Really excellent top lab
Except…
For 2 𝜈!!!
There are no 4-vectors
for the two!!
Template Method
Need distributions that
are strongly correlated
with the top mass
Template method to
figure out the top mass
ATLAS
The average π‘šπ‘™π‘ in the event
Two permutations (take smallest)
Avoid the missing 𝐸𝑇 resolution
Good
separation
power
20
CDF Template Variables
21
Fully reconstruct the top mass
Problem: detector measures missing 𝐸𝑇 = π‘πœˆ1 + π‘πœˆ2
𝑇
There are not enough constraints to solve for solution!
The πœ™ weighting method
πœ™1
Grid in the azimuthal angles
πœ™2
β€’ Fit for the top mass at each
grid location.
𝑓𝑖𝑑
β€’ Resulting π‘šπ‘‘ is the
template variable.
β€’ Weight by fit πœ’ 2 .
The fit πœ’ 2 includes terms for:
β€’ All the measurements (2 leptons, two jets, missing 𝐸𝑇 )
β€’ Top mass and the (constrained) W mass
Statistics Isn’t The Problem… 22
Broad peak, but decent
separation power.
Leading systematic:
Jet Energy Scale!
This measurement is statistics limited.
Can something be done?
Statistics Isn’t The Problem… 23
Broad peak, but decent
separation power.
Leading systematic:
Jet Energy Scale!
This measurement is statistics limited.
Can something be done?
CDF creates a second template variable:
π‘€π‘‘π‘Žπ‘™π‘‘ =
𝑙1 βˆ™π‘1 ×𝑙2 βˆ™π‘2
𝐸𝑏1 𝐸𝑏2
+ 120 GeV
β€’
β€’
β€’
Depends on 4-vector of leptons
Direction of jets
No Jet Energy Scale, no Missing 𝐸𝑇
And combines the two, optimizing for minimal error
𝑓𝑖𝑑
𝑀𝑑
𝑓𝑖𝑑
= 𝑀 βˆ™ π‘šπ‘‘
+ 1 βˆ’ 𝑀 βˆ™ π‘šπ‘‘π‘Žπ‘™π‘‘
Dilepton Top Mass Results
Standard Template Method
Jet Energy Scale isn’t fit: not enough constraints
Statistics already making a big difference here
24
Top Mass in All Hadronic
Decays (CDF & CMS)
44% of all decays. Largest
single decay class.
Overwhelmed by SM QCD
background!
6 Jets
After CMS requires 6 jets
4 jets with 𝑝𝑇 > 60 GeV
5th with 𝑝𝑇 > 50 GeV
6th with 𝑝𝑇 > 40 GeV
Estimated signal purity is 3%
Signal Efficiency is 3.5%!
25
Improving the Purity
Unique Handles:
26
2 𝑏-jets
Look for 𝑏-tagged jets
2 π‘Š β†’ π‘ž π‘žβ€²
Perform kinematic fit:
β€’ Know π‘šπ‘Š
β€’ The two π‘šπ‘‘ are the same
2 𝑑 β†’ π‘Šπ‘
(CDF)
Mass of the
pairs of light
quark jets
π‘€π‘Š is the well measured
value of 80.4 GeV
Mass of the
pairs of light
quark jets
𝑓𝑖𝑑
π‘šπ‘‘π‘Ÿπ‘’π‘ , 𝑝𝑇,𝑖 free
parameters
Every jet permutation is tried
Minimum πœ’ 2 is kept
Improving the Purity
1. Requiring the fit to converge
2. Very basic cuts on the πœ’ 2
Raise CMS’s purity to 39%
Additional kinematic selection
CMS: Δ𝑅𝑏𝑏 > 1.5
CDF: Neural Network
Raise CMS’s purity to 54%
CDF has a purity of 57%
27
Extracting the Mass
π‘šπ‘‘π‘Ÿπ‘’π‘π‘œ
π‘šπ‘‘
The Template Method
Fit for both Jet Energy Scale and π‘šπ‘‘
28
Lepton + Jets From CMS
Full 𝑠 = 8 TeV result: 19.7 𝑓𝑏 βˆ’1
Initial selection is > 100K
events and 94% pure
A simple kinematic fit
to clean up incorrect
jet assignments
29
Analysis is very similar
to the All-Jets analysis
from CMS
QCD background is negligible!
𝑓𝑖𝑑
β€’ Each possible jet assignment gives π‘šπ‘‘
β€’ Each is weighted by the fit probability
Largest systematic error is
the flavor dependent Jet
Energy Scale (0.41 GeV)
Conclusions



30
Field is still rapidly evolving
quark mass

World average submitted in
March

Top and anti-top have
consistent masses

CDF dilptons and all-hadronic


DZERO matrix element
πœŽπ‘‘π‘‘ measurements that can
clarify which mass we measure.

CMS all-hadronic and
lepton+jets

Becoming like the W mass…
What is next?

Tevatron will finish putting out
β€œfinal” mass measurements

LHC’s statistics and purity mean
it should quickly surpass the
Tevatron.

LHC Run 2 projections
Other measurements with the 𝑑
If you
believe
BICEP2!
Awaiting the next world
Combination…
31
Current World Combination
CMS Combination
Tevatron Combination
±0.95
±0.76
±0.64
32
Systematic Errors
ATLAS Lepton+Jets Template
33
34
ATLAS dilepton 7 TeV
CDF dilepton
35
CDF all jets
CMS all jets
36
CMS All Jets 7 TeV
37
DZERO Lepton+Jets ME
Tevatron Combination