monitoring muons; physics 2: top,di-boson,susy,dark matter melissa franklin doe review 2009

Monitoring Muons; Physics 2:top,di-boson,SUSY,Dark Matter

Melissa FranklinDOE Review 2009

8/21/2009

M.Franklin - DOE Site Visit 2

Muon Data Quality Monitoring: Harvard contributions

• Moed, Kagan,Guimaraes,Belloni, Martinez, Prasad started the muon monitoring project at ATLAS from Harvard.

• Moed and Kagan wrote the code in the DQMF framework to monitor all muon chambers(MDT,RPC,TGC) at Level1 and Level 2 based on Gnam histograms, make decisions as to which chambers are not working properly etc. and write displays to show what is working or what is not.

• Continually updated DQMF framework with new releases• And established an input/output connection to the database

• Belloni, Martinez, Prasad, Guimaraes became muon detector, readout and offline monitoring experts. They also became the experts who taught others to run shifts.

• Key participants in the muon run analysis task force

8/21/2009


New/old monitoring effortMost online monitoring of raw data using DQMF

•Continuously sampling data from the RODS (Level 1)•Gnam application also running on Calibration stream (Level 2) but we are not looking yet

MUON

8/21/2009


Monitoring tasks accomplished• Experts in Muon Online Data Quality Monitoring since 2007 -- Maintained at least 1 expert at CERN since

project began (Shulamit Moed and/or Michael Kagan) -- Experts in all data quality tools available for MDT shifter-- Responsible for MDT DQMF online monitoring sw infrastructure-- We define and develop the DQMF structure for MDTs (~7000 histograms monitored)-- We develop additional automated tools and algorithms to study MDT online data with DQMF-- Use these tools to determining the quality of the data, and identify/study problems in real time -- Responsible for defining how MDT data should be monitored online, how the tools can best asses data quality, and how to interface the tools best with the MDT shifter ﾊ -- Helped develop DQMF online monitoring infrastructure for RPC and TGC -- Currently help all Muon sub systems maintain and develop tools for DQMF online monitoring-- Responsible for training MDT shifters in Data Quality monitoring -- Hold the ﾒMuon data quality tools ﾓ session during Muon shifter training sessions Status and Developments ﾊ -- Spent a great deal of time studying the data and understanding the problems, and tuning our monitoring structure, algorithms and error thresholds -- When we began, system was in chaotic state, difficult to understand data, and difficult to define tests to properly identify problems (many chambers would fail our DQMF tests)-- As of August 2009, Only 25-35 out of ~1100 chamber fail our DQMF tests. We are investigating each of these problems, many of which we already understand -- Developed graphical interface for MDT system online monitoring with DQMF. -- Greatly increases the ability of the shifter to navigate large MDT system, study data and identify problems -- Developed tools for automated archiving of data quality results into Databases -- Develop and maintain MDT online monitoring documentation / twiki-- Constantly in Control room monitoring MDTs during data taking, both while on and off shift Harvard will maintain its role as Muon Online data quality experts at CERN as we approach first beam of the LHC -- Michael Kagan will be present at CERN as expert-- Two 3rd year students (Laura Jeanty and Giovanni Zevi della Porta) also at CERN, becoming Muon monitoring experts.Expanding our efforts to offline monitoring -- Not all problems identified online can be solved online-- Using our expertise in MDT online data and taking a larger role in understanding problems in offline data-- Following problems from online to offline and solving them with complete offline data sets-- Beginning to understand how to correlate MDT data problems with other ATLAS sub systems-- One example: spikes recently found in the TDC spectrum of MDT.

8/21/2009


Level 1 monitoring

8/21/2009


GnaMon: A Gnam Histogram Interpreter

• MDT specific stand-alone tool (not in tdaq)– Can be used online and offline: Runs on Gnam root files– Performs statistical analysis without need of reference histograms– Highlights suspicious tube clusters like multilayers, tube layers,

mezz.

Dropped chambers

Run: 125032TGC triggering bottom half

Hit occupancy overview for the MDT’s from GNAMON

8/21/2009


Test integration case: Spikes in ONLINE monitoring TDC spectra

QuickTime™ and a decompressor

are needed to see this picture.



Chi-square fit fail --->> DQMF warning to shifters

Improve fitting algorithm to identify spikes

8/21/2009


Move to offline to investigate



Spike events have tens of thousands of MDT hits!

Often the “charge” associated with hits in spikes is zero

8/21/2009


Is there a pattern in crates, grounds, ?

8/21/2009


Summary of monitoring plans• Laura Jeanty, Michael Kagan, Giovanni Zevi della

Porta, Shulamit Moed(ex-officio), Melissa Franklin, new post-doc

• New group taking advantage of the wide knowledge of Harvard muon group who will work on finding and fixing problems found by monitoring while improving monitoring

• Use online, calibration stream, offline, knowledge of DAQ, DQMF, GNAM , pulser runs etc to solve problems

• Post-doc resident at CERN arriving fall

Physics 2 OverviewWhat we thought in the spring!

Use top physics to look beyond the SM

Guimaraes, Franklin, Morii, Mills, Belloni, Prasad(5), Smith(5) +

Prasad thesis- top cross-section 50/pb

Smith thesis - ttbar resonanceNew schedule --> Smith thesis

WZ cross-section 50/pb

8/21/2009


pp X ttbar + x• Many models for new physics (extra dimensions, little Higgs, SUSY….) predict heavy resonances, some

of which (notably the Randall-Sundrum gluon) couple preferentially to top• Tevatron limits are model dependent but exclude resonances below ~1 TeV• Can significantly extend this reach at 14 (and even 10) TeV• As mass of resonance increases, decay products of top become increasingly collimated experimental

challenges– overlapping jets– nonisolated leptons from W decay

pT > 20 GeV

To the left: number of reconstructed jetscone algorithm, cone = 0.43 jet bin

many signal events, particularly for 2 TeV resonancefew events from ttbar (dominant background)but: W+jets may become important

Identifying a jet from two partons (such as from the two quarks of a W decay)

8/21/2009


Identifying Merged Jets• Jet Eccentricity

– geometrical measure of elongation of jet energy deposit in the calorimeter

– 0 = circular, 1 = highly elliptical/elongated• Uncorrelated with jet mass, another good

discriminant• ATL-com-phys-2009-338

8/21/2009


WZ feasibility studies• W+Z l+l-l+: motivation

– “bread and butter” Standard Model measurement– Key background for same-sign lepton and trilepton searches– Develop understanding of multilepton data, technology for estimating

backgrounds• Back-of-the-envelope feasibility check: cross section measurement with

100 pb-1

– 10 TeV• S/B ~ 5 ~36/7

– 7 TeV• S/B ~ 5~20/4• Promising enough to take the next step• More detailed assessment of expected precision

8/21/2009


Top with low energy and small sample

E (TeV)sel ee sel esel 4 4 11 55 8 22 106 13 36 167 20 53 258 27 81 3510 50 134 62

Generated di-lepton top (e/) signal samples at 4, 5, 6, 7, 8, 10 TeV

Selection eff. For dilepton events flat across energy range

events in 50/pb Selected

8/21/2009


Top cross-section

• Lepton+jets• Prasad thesis

# of events with hi pt muon and jets in 50/pb

Plot made before latest CERN announcement

8/21/2009


New physics with di-leptons in the longer term

SUSY, Dark Matter etc• Laura Jeanty(3), Giovanni Zevi della Porta(3),

Mills, Franklin …

8/21/2009


q

q~

q~

q’

q q’ o~

o~

~

~

l~ l

lW

gg

g g~

g~

Same Sign Dilepton Signature for SUSY

• Putting together possible gluino decays – (also possible with squark production)

• One illustrative decay process:

leptons

jets

missing energy

8/21/2009


Same Sign Dilepton Signature: Motivation

• How often does a gluino gluino pair produce same sign dileptons?

• Total branching fraction to same sign leptons ~ 4%– 12% if include top or bottom leptonic decays– for comparison, 40% of gluino gluino decays have no leptons

• The signature has multiple handles - leptons, jets, missing energy

• Does not rely on missing energy and jets alone, unlike other SUSY searches

€

gg → ˜ g ̃ g → jet jet jet jet l±l± / E T

100%

t~

t b o~

~ lW

g~

22%

~60%: q = t

97% Other contributing decays from gluino decaying to sbottom

gluino decays to isolated lepton: 26%

gluino gluino pair decays to dileptons: 7.5%

gluino gluino pair decays to same sign dileptons: ~4%

€

gq → ˜ g ̃ q → jet jet jet l±l± / E T

€

gg → ˜ q ̃ q → jet jet l±l± / E T

8/21/2009


Can we trigger on it?

Low pt leptons>hard to trigger

8/21/2009


SUSY backgrounds: our specialty

8/21/2009


Discover dark matter at the LHC

Recent results in astrophysics suggest a possible dark matter

candidate which could be produced at the collider

8/21/2009


Possible production mechanisms @ LHC

8/21/2009


Clear signature - lepton jets

Analysis summary

Study di-leptons WZ, ttbar, SUSY initially; ttbar resonance, dark matter, more SUSY over time

Study Z cross-section, Z pt, Z + jets angular distributions initially:

ZHiggs, Z’,Zgamma over time

monitoring muons; physics 2: top,di-boson,susy,dark matter melissa franklin doe review 2009

Documents