Transcript Slide 1

Testing the Standard Model:
A History and Rationale for
Signature-based Searches
Henry J. Frisch
Enrico Fermi Institute
OUTLINE
1. Some History (CDF-centric)
2. Rationale(s)
4. My goal- thoughtful discussion of issues
5. Exptl and Theoretical Consequences
Apologies- Something of a rant after just getting 3
papers through the CDF Godparenting process- there
is a deep diffence in opinion of how we (we_all) should
spend our time.
BUT still the best game in town…
7/7/2015
1








Brief History
Run I- eeggmet event led to search for `cousins’ in
gg+X and lg+X
Run I- Dave Toback Ph.D thesis on gg+X
Jeff Berryhill Ph.D thesis on lg+X (2.7s)
Run I- Ray Culbertson gbj+met- ???
Run II – Andrei Loginov Ph.D thesis on lg+X
Run II – gg+X- GMSB-optimized search Eunsin Lee
thesis; Dave Toback, Ray Culbertson, Sasha Pronko,
Max Goncharov
Run II- Irina Shreyber lgb+X; ttbar+g
Run II- Shin-Shan Yu, Dan Krop, Ray Culbertson,
Carla Pilcher, Scott Wilbur, HJF- gbj +X
Run II- Georgios Choudalakis, Ray Culbertson, Conor
Henderson, Bruce Knuteson- VistaSleuth:
arXiv:0712:253402 (`Industrial Strength’ sig. searches)
7/7/2015
2
SOME OF OUR COMMUNITY HAVE DISCOMFORT WITH
HAVING NO BSM MODEL AS MOTIVATION- ``cite
SOMETHING, ANYTHING”

Examples from recent b-quark+gamma +met
“Is consistent with possible BSM signature”? HUH?
”One would hope that there are some models for
new physics that.. could place limits on..” SOME?
7/7/2015
3
Rationale(s) for SM emphasis
1.
2.
Have a robust and predictive hypothesis to testthe Standard Model- testing it is classic science.
Emphasis should be on understanding and
improving the detector performance on SM
predictions- time spent elsewhere is very costly
(zero sum game for time and $ )
3.
4.
5.
7/7/2015
Exptl papers dependent on a model do not age
well- 20 years later one could use the data, but the
comparisons with models are junk, and diminish
the paper (e.g Trion-ProtoDynamics)
Particle theorists do it better- experimentalists
should concentrate on communicating results to
them and working together
Students learn the wrong lessons from poorlymotivated limit setting- complacency on $,time
4
HEP Searches*
One of the best young CDF physicists asked me
`How are discoveries made? I’ve never seen one…’
1.
2.
3.
4.
5.
6.
7.
J/Psi at BNL (LML,PJL story)
Neutral Currents at BNL (twice)
Neutral Currents at E1A Fermilab
Top at UA1
SUSY at UA1
J/Psi at Wideband Photon at Fermilab
And many others (young CDF’ers don’t know what
they’re missing!)
Point is to discover, not to set limits- so,
what are the right strategies?
* Me too- KL->mm
7/7/2015
5
Finding what you expect, and not
finding what you don’t
Rubbia and the 1984 Top Discovery
Expected and so Found
7/7/2015
Lederman and 1971 J/Psi
Unexpected and so Not Found
6
Announcement of the 1986 UA1 SUSY
Discovery (Aspen 1986)
Expected and so Found (?!)
Shot down by
Steve Ellis`cocktail’ of
SM processes
7/7/2015
7
Our Own Saga at CDF Run I
Unexpected- but fired
>30 (?) triggerselectron, met, photon,
Z, ….
First explicit signature-based search at CDF
grew out of one event-the `eeggmet’ event
7/7/2015
8
Are There Cousins to the eeggMet Event?
(with D. Toback, J. Berryhill, R. Culbertson)

Idea- likely other related signatures from
other decay chains- e.g. if this is due to C2C2,
maybe there is C2N2 production ending in 2
gammas+ 1 lepton, or N2N2 ending in 2 gammas
+met (hence `cousins’- signatures related by BR’s)

Two obvious ways to slice the eeggmet
signature2 gammas +X (2 neutral parents)
 Electron-gamma+X (1 neutral 1 charged)


Look in these signatures, as well as mmggmet+X
first explicit signature-based searches at cdf
7/7/2015
9
The eeggmet event and the ggmet signature
Observed
SM expect
D. Toback
Ph.D Thesis Univ.
of Chicago
Dec. 1997
.
PRL 81, 1791
(1998)
PRD D59,092002
(1999)
Plot of Data vs SM
expect
.
Limits on models
Which are more
useful now?
7/7/2015
Eeggmet event
10
The eeggmet event and the lgmet signature
J. Berryhill
Ph.D thesis Univ.
of Chicago
Dec, 2000
PRL 89,041802
(2002)
Phys Rev D66,
12004 (2002)
Observed
SM expect
.
7/7/2015
11
Run II eeggmet Event Followup: lg+X
In g-l+X we found a 2.7s excess over
SM. What to do?
Decided to repeat the analysis in Run II
with the SAME (published) cuts so it’s A
PRIORI. Only way to test.
Note- NOT optimized for GMSB, etc.. (!).
CDF Run I PRL: ..”an interesting
result, but … not a compelling
observation of new physics. We look
forward to more data…”
Run II eeggmet Event Followup
Andrei Loginov repeated the lgmet analysis- same cuts (no
optimization- kept it truly a priori.
Run II: 929 pb-1 at 1.96 TeV vs Run I: 86 pb-1 at 1.8 TeV
Run II-Searching for SM ttbar+ g
Irina Shreyber and
Andrei: added the
requirement of a bquark: e or m, g,
b+X
(nice channel for
BSM)
Then ask for >=3
jets+ large Ht to
look for SM
radiative top
decays: ttbar+ g
7/7/2015
14
Ttbar+Gamma: Adding a b-quark, >=3
jets, and large Ht
No suprises- nothing really
heavy, wild, or odd- to be
submitted to PRD shortly:
Consistent with ttbargamma
=0.15+-0.08 pb (tiny!)
7/7/2015
Small cross-section
15
Run I: Eeggmet event and the gbjmet
signature
Gordy Kane suggested looking for
C2N2-> stopbgN1->bcgmet

BUMP

7/7/2015
This was a prioriselected cone=1.0
for dijet mass
from UA2 W ->jj
detection
We got excited
(even Gordy)- but
data not robust to
changing cuts, and
too little statistics
to be sure…
16
Run I: gbjmet signature cont.
Different cone
sizes: a) not robust;
b) kinematic edge?
7/7/2015
D3-body vs 2-body
mass- bump and
outliers- odd? (or not)
17
Run II- the gbjmet signature


Dan Krop, Shin-Shan Yu, Carla Pilcher, Scott
Wilbur, Ray Culbertson, HJF
Look at the same 2 plots with >20 times the data
NO BUMP
As Sasha Paramonov would say,
`Alas’ (`worse than losing a girlfriend’)
Just submitted to PRD
7/7/2015
arXiv:0905.0231
18
Run II- the gbjmet signature
We observe
617 Eventsexpect (SM)
607 (too good)
Can tighten
selection
beyond Run I
to go to
smaller
crosssections
7/7/2015
19
Run I: Other Odd Events with Photons and
Leptons
Just one
example
2 legs
of the
Z (!)

7/7/2015
200
Gev
Photon
Z-boson to mu-mu balanced by a 200
GeV-Pt photon (presumably)
20
Run II: Search High Pt Z+X
Look at a central Z +X, for Pt > 0, 60, 120 GeV, and at distributions…
Need SM predictions even for something as `simple’ as this… (not easy-ask Rick
Signature-Based High Pt Z+X Searches
PTZ>0
PTZ> 60
PTZ>60
PTZ>120
Njets for PTZ>0, PTZ> 60, and PTZ>120 GeV Z’s vs
Pythia (Tune AW)- this channel is the control for
Met+Jets at the LHC (excise leptons – replace
with neutrinos).
Signature-Based High Pt Z+X+Y
Simple Counting Expt- ask for a Z + one object, or Z+ 2objects
Two Objects
One Object
Z+X+anything
Z+X+Y+anything
Signature-based W/Z+Njets Search
Crossection vs
number of jets in W
and Z events
% uncertainty vs
number of jets in
W and Z events
So, switch to a measurable that is more robust: look for new physics by precise
measurements of (W+Njets)/(Z+Njets). Systematics at few % level
(PRD68,033014;hep-ph/030388
7/7/2015
24
Return to Rationale(s)
1.
2.
Have a robust and predictive hypothesis to testthe Standard Model- testing it is classic science.
Emphasis should be on understanding and
improving the detector performance on SM
predictions- time spent elsewhere is very costly
(zero sum game for time and $)
3.
4.
5.
7/7/2015
Exptl papers dependent on a model do not age
well- 20 years later one could use the data, but the
comparisons with models are junk, and diminish
the paper (e.g Trion-ProtoDynamics)
Particle theorists do it better- experimentalists
should concentrate on communicating results to
them and working together
Students learn the wrong lessons from poorlymotivated limit setting- complacency on $,time
25
Return to Rationale(s)
Rationale 1- Testing the SM- a trigger-person’s view.
Have a perfectly good model to test!
Can hide
-1
60mb
X
4
fb
all kinds
of physics = 2.4 1013 SM
here …
events- looking
for ~10. Parts
per 1012
7/7/2015
Factor of 105
Don’t
is rejected
assume you
before data
know what it
even go to
looks like
tape
here
26
Rationale 2- Improving Detector Performance
Tuning
MC, Sim
Rick Field
UEV-
Willis
Sakumoto
pttune <20
Amazing
agreement
7/7/2015
W,Z to e,mu (Sasha)
27
Return to Rationale(s)
Rationale 2- Improving Detector Performance from
comparisons with SM control samples, single
particles, etc. (vs time spent on limit-setting)
Fake muons (Sasha)
Fake photons (Andrei)
7/7/2015
Developing Tools
28
Return to Rationale(s)
Rationale 3- Experimental papers that are built on a
specific model and parameter values- don’t age
well.
I didn’t want to cite specific examples- you’re
welcome to find your own among the SUSY
limits from the early days.
There are of course exceptions- we shot down the
Constituent Interchange Model, and very general
or well-tested models (e.g. the SM model) are of
course worth testing
NUF SAID
7/7/2015
29
Communicating results of signaturebased searches to Theorists
Proposal (R. Culbertson et al, Searches for new physics in events with a photon and
b-quark jet at CDF. Phys.Rev.D65:052006,2002. hep-ex/0106012)- Appendix A:
3 Ways:
A. Object Efficiencies (give cuts and effic. for e, mu, jets,b’s. met,….
B. Standard Model Calibration Processes (quote Wg, Zg, Wgg in lgmet,e.g..)
C. Public Monte Carlos (e.g. John Conway’s PGS)
True Acceptnce
Ratios to True (ABC)
Comparison of
full MC with
the 3 methods:
Conclusiongood enough
for most
applications,
e.g. limits…
7/7/2015
Case for gamma+b-quark+met+x (good technisig)
30
Return to Rationale(s)
Rationale 4- Particle Theorists can often test
their own models more productively than
some student, and should.
My main point is that there are exptl tasks that
need doing, including exploring other
channels and making other measurements;
running the detector, inventing new
detectors,…
Many consequences (needs discussion by
PTB)- including funding particle theory
separate from string theory, supplementing
funding particle theory from exptl funds,…
7/7/2015
31
Return to Rationale(s)
Rationale 5- Passing on the culture of time is
$
7/7/2015
32
The End-
Thank you, and apologies for some curmudgeonliness
7/7/2015
33
BACKUP
7/7/2015
34