Calorimeter Gaps Approach Soft QCD Meeting 14 th Oct 2 0 10 Tim Martin Paul Newman University of Birmingham 1 Soft Diffraction 2 Diffractive Events Colour singlet exchange Can be ID: 788961
Download The PPT/PDF document "Single Diffractive Cross Section" is the property of its rightful owner. Permission is granted to download and print the materials on this web site for personal, non-commercial use only, and to display it on your personal computer provided you do not modify the materials and that you retain all copyright notices contained in the materials. By downloading content from our website, you accept the terms of this agreement.
Slide1
Single Diffractive Cross Section
Calorimeter Gaps Approach
Soft QCD Meeting
14th Oct 2010
Tim Martin, Paul NewmanUniversity of Birmingham
1
Slide2Soft Diffraction
2
Diffractive Events
Colour singlet
exchange. Can be Single or Double
proton dissociation. Diffractive mass can be anything from
p+π0 up large systems with hundreds of GeV
invariant mass. Soft PT spectrum. Large forward energy flow.
Less activity in the inner detector.
Pythia
Single Diff.
Transverse
Momentum
Transfer
Slide3Diffractive Mass Distributions
3
Pythia8
Measure dσ
/dξ over the ξ range accessible at ATLAS
Constrain
the
relative weights of
SD
,
DD
and
ND
At
7 TeV
we should be getting
really big
diffractive systems!
Slide4Truth Level Energy
Flow
4
Normalised to a single event in ξ.
Differential
in
η
.Expect
rapidity gap devoid of all soft activity
.
Single Diffraction
Linear
relation between
η
of leading particle and log(
ξ
), slope = 1 for
ln
(
ξ
).
Part
hadronisation
, mostly kinematics.
η
max
=
η
of
Leading Truth Particle in X System
Are there any MC diffractive models which do
not
use the lund string hadronisation model?
Slide55
Trigger Efficiency
Trigger events with the
MBTS. Trigger is fully within calorimeter acceptance so use loose trigger requirement.
L1_MBTS_2.
`Here be dragons’
Unexploreable
phase space with Calo + MBTS analysis.
Correct for trigger inefficiency in data using MC here
Slide66
Experimental Re-
definiton
In
Double Diffraction, if the smaller system (MY
) is too small to get into the ATLAS calorimeters, then the event is
experimentally indistinguishable (exc. ZDC) from a single diffractive event. This occurs
around ξY of 10-6
Reclassify as Single Diffractive.
Name the new sample
Single Diffractive Like
New Definition:
Double Diffractive Like
Systematic Check
: Location of cut
Slide77
New Cross Sections
Single
and Double Diffractive Like take into account the migration of events from the Double Diffractive to Single Diffractive
samples. Visible refers to the
subset of the samples which can be triggered with the MBTS.
Slide88
Samples & Calo Cells
Current data sample of
~750k, Pythia8, Phojet, Pythia6; ND, SD, DD
MC Empty
MinBias and RNDM stream for 1
st 7 TeV run:
Run 152166~
600k
L1_MBTS_2
&
BCID
=
1
Events
7719
mb
.
Noise in
TILE
show non-Gaussian profile (double Gaussian).
TILE is not required for hermetic coverage of ATLAS and is
currently excluded
from the analysis.
mc09_7TeV.105003.pythia_sdiff.recon.ESD.e514_s764_s767_r1307/
mc09_7TeV.105004.pythia_ddiff.recon.ESD.e514_s764_s767_r1307
mc09_7TeV.105001.pythia_minbias.recon.ESD.e517_s764_s767_r1307
data10_7TeV.00152166.physics_MinBias.recon.ESD.r1239/
data10_7TeV.00152166.physics_RNDM.recon.ESD.r1239/
mc09_7TeV.106096.PhojetNdiff.recon.ESD.e514_s764_s767_r1307
mc09_7TeV.106097.PhojetSdiff.recon.ESD.e514_s764_s767_r1307
mc09_7TeV.106098.PhojetDdiff.recon.ESD.e514_s764_s767_r1307mc09_7TeV.108316.Pythia8_minbias_ND.recon.ESD.e533_s764_s767_r1307mc09_7TeV.108317.Pythia8_minbias_SD.recon.ESD.e533_s764_s767_r1307
mc09_7TeV.108318.Pythia8_minbias_DD.recon.ESD.e533_s764_s767_r1307
Oldrich
&
Pavel
EM Barrel
Slide99
EM Barrel
EM
Endcap
Inner
EM
Endcap
Outer
FCAL
HEC
Tile
Calo
Cell
Significance
Data10 MBTS_2
Data10 RD0_EMPTY
MC Py8 Inelastic
MC Empty
Very Bad
Slide1010
Cell Requirement
Using an
Energy VetoCurrently require 1 cell above 5.5 E/σ
.Can perhaps improve our sensitivity by combining cells?
No good for all cells in a slice.
E.G Barrel slice:
39,925 cells
RMS Noise
of all cells combined:
8.6 GeV!
Swamps any minimum bias energy deposition.
However, combination of cells with E/
σ
> 3 looks more promising.
Order 50
cells
per
η
slice, RMS noise of ~
100
MeV
–
manageable
.
Under current investigation, may improve systematics.
Slide11Gap Finding
11
Largest
η Gap
E.G.
Single
Diffractive
Event
M
X
M
Y
η
•Trigger on
L1_
MBTS_2.
•
Negligible
pileup
, no
prescales, low beam backgrounds
.
•Divide the calorimeter into
10x
rings spanning 1 unit of η
.
•Ring
is
full if 1 or more calo cells satisfy E/σ
Cell
>
Threshold
•Look
for
largest continuous gap of empty rings
.
•Plot
Start of Gap
[|η|]
vs. Size of Gap [Δη].
•The
start of the gap is defined as the
|
η| of it’s farthest edge from η=0
•Provides
separation
of diffractive and
non-diffractive.
MBTS
Being Quantified
Systematic Check
Slide1212
Overall Trigger Eff.
Trigger efficiency calculated from
MC on a bin-by-bin basis.Fractional breakdown of event types in each bin displayed below.
Overall
Trigger
Efficiency
MC
Bin-by-bin
Breakdown of
Event types
Key
Totally Full
Detector Bin
Totally Empty
Detector Bin
Slide1313
Tr. Eff Correction
Trigger
efficiency corrections are applied bin-by-bin to the data histogram.Bins with a MC trigger efficiency of < 50 % are not used in the fit.These are the `here be dragons bins’
Before Correction
After Correction
Low Efficiency
Slide1414
Formation of MC Templates
ND
SD
DD
Map
2D distribution onto 1D
plot for the purpose of fitting.
Pre-trigger MC distributions
.
Normalise
each
MC shape to a single event.
Bins with overall trigger efficiency < 50% are plotted last –
beyond the black line.
These bins are not included in the fit.
Slide1515
Check of Fit
Form pseudo-data by applying bin-by-bin Poisson smear
to MC prediction.Fit relative SD, DD and ND templates to pseudo-data.Fit constrained to sum to 1
One Trial
Residuals for one trial
1000
Trials
Slide1616
Fit to Data – Pythia8
Fit to
data normalization in fitted bins.Log likelihood fit.MC shapes do not fully describe the data. Poor χ
2
Pythia8:
ND:
73.6%
SD Like: 23.7%
DD Like: 2.7%
Fit:
ND:
74.2%
SD Like: 18.1%
DD Like: 7.74%
Slide1717
Detector Acceptance
To convert to a
cross-section, need to correct back to the particle level.Run gap finding algorithm over truth level data.
Move from Calo Cells
Truth ParticlesMove from E/
σ cut
PT cut
Build
bin-by-bin correction histogram to apply to data
.
c.f
the trigger efficiency
correction.
Correction factors manageable
in
Single Diffractive Like
(
edge gap
) region.
Large correction factors
in
Double Diffractive Like
region (
central gaps
)
Slide1818
Single-Diffractive
Comparison of algorithms at:
Truth (PT > 200
MeV
)
Reco
(E/σ > 5.5)
Hatched boxes denote located gaps.
Correct effects such as:
Low P
T
particles not reaching
callo
.
Multiple scattering
.
Slide1919
Cross Section
This slide 0
MeV PT cut gap definition
Data is... Corrected for Trigger Efficiency &
for Detector Acceptance.
Scaled to Lumi of run
.Plotted systematics, `best and worse case scenarios’
E/
σ
> 5 noise
AND
+
10% Cell energy scale
AND
111%
Lumi
.
E/
σ
>
6
noise AND -10% Cell energy scale AND 89% Lumi
.
Last two bins have Trigger Efficiency < 50%
Suggestion from R. Barlow! Will check adding in quad as well.
Systematic Band
(systematics invert)
Default & fitted
fractional visible
Xsec
MC lines
Statistical error
negligible
Slide2020
Cross Section
This slide 100
MeV PT cut gap definition
Interesting sensitivity to choice of PT
cut defining cross section.
Most effect at small gaps, explained as increasing pollution from non-diffractive events.
Flick back and forth...
Slide2121
Cross Section
This slide 200
MeV PT cut gap definition
Flick back and forth...
Slide2222
Cross Section
This slide 500
MeV PT cut gap definition
Flick back and forth...
Work ongoing to match a sensible P
T
cut with the E/σ cut.No particles in the gap (P
T
> 0 cut) looks sensible.
Can also publish differential in P
T
Slide2323
Cross Section
Combination into a single result.
Pythia8 taken as trigger and acceptance correction MC
Maximum deviation per data point for Pythia6 and Phojet taken as systematic.Added in quad.
, last two (low trigg
. Eff.) bins not plotted.
0
MeV
P
T
Definition
100
MeV
P
T
Definition
Slide24Conclusion
24
Single diffractive like cross section d
σ/dη = 1 mb η-1 Measurement of the single diffractive like cross section at 7 TeV possible with very little minimum bias data.
Many
cross checks, systematic checks, systematic quantification
is underway.
Still a bit more work to do with Calo Cell selection &
P
T
dependence on acceptance
correction
.
We are working to
compare with
Tevatron
results at 2 TeV
.
We are writing, detailed internal
note for the analysis being produced at
the moment.
Details theory, all steps & cross checks.