Title: Higgs working group status report F'Cerutti and G'Unal on behalf of the Higgswg
1Higgs working group status reportF.Cerutti and
G.Unal on behalf of the Higgs-wg
- Update since ATHENS physics workshop
- SM discovery revisited Old channels and new
ideas - Higgs properties (spin, couplings,
self-coupling)te - SUSY Higgs
- Getting ready for real data analysis
- Plans for Rome Physics Workshop
- Conclusions
2Disclaimer
- Many activities and people involved over the last
year - since ATHENS 9 meetings with more than 80
presentations - Not everything can be covered in this talk
limited to few more recent items/developments - For complete list of recent notes and summary of
current activities, analysis status see Higgs-wg
web page (maintained by G.Unal) - http//atlas.web.cern.ch/Atlas/GROUPS/PHYSICS/HIGG
S/higgs.html -
3SM Higgs
- TDR discovery channels mainly inclusive
searches - look for mass peak over (unknown)
continuous background (design robust analyses) - ttH with H-gt bb
- H-gt g g
- H-gt Z Z -gt 4 leptons
- H-gt W W -gt lnln
- EW Production mechanism reduces QCD background
(tt, WWj, Zj) Vector Boson Fusion VBF - H-gt W W -gt ln ln (-gt ln jj)
- H-gt t t -gt lnn lnn and -gt lnn hn
Discovery potential Scientific Note
2003-024 Bonn, Cracow, KEK, Mainz, Montreal,
Pisa, Sheffield, Tokio, Wisconsin
4ttH, H-gtbb
- Revisited recently by J.Cammin-Bonn SN 2003-024
- Discovery potential similar to TDR loss due to
better bkg. estimate recovered by improved
Likelihood method used in mass reconstruction - Studies on background normalization,
extrapolation - Use of mass sideband (very challenging)
- Bkg. from ttjj events (reverse b tagging) mass
shape correction from MC - Other ttH channels
- H-gttt (Zivkovic,Tanaka)
- H-gtWW (Leveque,Tanaka)
- not discovery channels but useful for coupling
measurements
(30 fb-1)
5H-gt ZZ-gt 4 leptons
- One of the most investigated channel
- 4 m Cosenza, Frascati, Saclay, MPI, Dubna,
Wisconsin, Athens, Thessaloniki - 4 electrons Wisconsin, Weizmann
- 2e-2m Wisconsin
- 2e-2t Weizmann (hopeless at low mass)
- Important progress
- Analyses revised with full simulation DC1
- Use of multivariate techniques (NN and
likelihood) in parallel to TDR cuts for reducible
and irreducible bkg rejection (Cosenza-Frascati,
Wisconsin) - Study of NLO/LO k-factors vs cuts for sign
irr. Bkg (Wisconsin) - Study of bkg control samples and systematics
(Cosenza-Frascati) - Relaxed iso-cut analysis 3 iso-lep 1
Loose-iso-lep (Wisconsin) - Impact of noise and pileup on muon isolation
investigated with full simulation negligible at
low-luminosity (Dubna ATL-COM-PHYS-2004-037)
6H-gt ZZ-gt 4 m
- Mass resolution and selection efficiency studied
with DC1 data deterioration with respect to TDR
(both m-standalone and m-combined) - Some efficiency loss attributed to detector
change (services in m-spectrometer) some to
software problems - TDR performance reproduced also thanks to
multivariate techniques a part for MH180-200
GeV opening of ZZ on-shell background (probably
not correctly taken into account correctly in TDR
?)
H-gt4m x 4
E.Meoni et al. ATLAS-PHYS-COM 2004-040
R.Nicolaidou et al.- Saclay
7H-gt ZZ-gt 4e
- Analyses performed on DC1 samples e-calibration
(strong contribution of Wisconsin group) and
e-isolation studied - TDR results reproduced a part for 180 GeV ! (as
for 4 muons)
M.Riveline E.Gross Weizmann
8H-gt ZZ-gt 4e and 2e2m
- Proposed by Wisconsin looser lepton-isolation (1
loose isolation required for 1 out of 4 leptons)
slight improvement in significance 10 at the
price of having a reducible background (Zbb) at
similar level than the irreducible one
B.Mellado et al. Wisconsin ATL-COM-PHYS-2004-042
9H-gt g g
- Full simulated gg from DC1 used to check mass
resolution performance Tartarelli-Carminati
(Milan) also by Escalier (Paris) and Fang
(Wisconsin) - Mass resolution 5-10 worse than in TDR (more
material IDmore conversions) Still g-conversion
identification from MC truth instead than from ID
recon - Study of reducible background with full
simulation using DC1 QCD samples Escalier,
Loureiro(Wisconsin), Unal (LAL) - Study of NLO prediction for signal and
irreducible background B.Mellado(Wisconsin),
Unal, Escalier, Laforge - From inclusive to H n-jets analyses Mellado,
Unal, Minagawa, Kanzaki (KEK), Escalier, Laforge
10H-gt g g from DC1
- Background from fake photons reevaluated using
inclusive DC1 QCD-jj samples - Rejection similar to TDR studies (Wielers)
- q-Jets less rejected than g-Jets (harder
fragmentation function) - TDR estimate of gjet Bkg. was underestimated
factor 2-3 - Reducible B mostly q-jet for this process while
rejection parameterized on QDC-jj sample
dominated by g-jet ! - To finalize this study
- Tighter isolation cuts at low luminosity
- Reconstruction of conversions with the tracker
G.Unal
11H -gt gg NLO studies
- Irreducible background exists at NLO (caveat
LO and NLO not well defined for gg) - TDR computation LO cross-section for signal,
BornBox production of gg (x1.5 to account for
Bremsthralung contribution) - NLO computation NLO cross-section for signal,
NLO contribution for gg Box at NLO
Irreducible bkg.
Born
LO
NLO
Box
(30 fb-1)
MH120 GeV L30fb-1
Significant uncertainty on MC background
predictions Background will be known much better
with the data !
12H -gt gg From inclusive to H jets ?
- Studies based on NLO MC
- Comparison of gg1 jet with inclusive gg channel
(old idea used by CMS, no conclusion in TDR
because of lack of NLO computations)
(Mellado,Unal) - Use other variables than Mgg to discriminate
Signal and Background, for instance Pt gg
(Escalier,Laforge) - Better S/B can be achieved
- VBF H-gtgg channel investigated much better S/B
but less statistical sensitivity (50
inclusive) - Mellado et al.-Wisconsin ATL-PHYS-2003-036
- Minagawa, Kanzaki Tokyo-KEK
- VBF and inclusive channels can be combined
13H-gt W W
- Inclusive channel used in TDR counting
experiment with small shape difference Signal vs
Background - VBF production better significance -gt reach
lower Higgs masses (down to 120 GeV - Recent developments
- NLO computation for S and B (mostly WW, tt) for
inclusive TDR channel (M.Duehrssen-Freiburg) - Shape of WW background at NLO shape of
SignalBackground at LO (MH170 GeV) ! - Study of VBF analysis re-optimization with
multivariate techniques NN (improvement 30-40
on significance) and NLO bkg. Estimate (Cranmer
et al. - Wisconsin). Includes studies on bkg.
control samples - New idea Combine WW0 jet (TDR), WW2jets
(VBF) and WW1 jet (one jet forward) Quayle
et al. - Wisconsin. Statistical sensitivity looks
promising.
14VBF H-gt t t
- Key Issue here is Z-gttt mass tails need to study
on full simulation - VBF analysis of DC1 fully simulated events
started K.Cranmer-Wisconsin and
I.Rottlaender-Bonn Problems with PT-miss
calibration - Recent development (Mellado et al.Wisconsin
ATL-PHS-COM 2004-023) - H -gt tt 1 jet , jet in intermediate h region
- Expected to be less sensitive to detector
systematics/calibration - intermediate jet tagging hJgt1
- more relaxed jet veto in central region 30
GeV - Statistical sensitivity good from fast simulation
- Need higher order QCD correction for Bkg estimate
and control samples from data (study started by
Wisconsin) - Strong correlation with VBF channel combination
difficult ? Choose one of the two (But VBF
important for coupling measurement) ?
15Higgs properties
- Scientific note on H spin/parity measurement for
M(H)gt200 GeV from H-gt ZZ -gt 4-leptons SN 2003-025
(Buszello et al.) - At low-mass studies started measure CP in tth
h-gtbb (Hockeppel Bonn) or in tth h-gtgg
(McGarvie -Royal Holloway) - Measurement of ratios of couplings. Use all
channels available. New study completed by
M.Duehrssen ATL-PHYS 2003-030 - Update on H self-coupling (Dhahloff-Freiburg)
disagreement on tt bkg. with previous studies
(F.Mazzucato Geneva)
some (moderate) assumptions no significant
contributions in loop from new particles and
light quarks, only one Higgs boson
16SUSY Higgs
17CP SUSY Higgs
- New scenarios with CP violating SUSY couplings
investigated - Couplings to SM particles are changed (production
and decay) - Particularly challenging scenario suggested by
theorist (Carena et al) H1 decouples from gauge
bosons Very light H1 escapes LEP searches - Most of the parameter space covered with existing
searches - Some holes left at very low MH1 under
investigation recovered with specific topologies
like H -gt WH1 ?
Preliminary
M.Schumacher - Bonn
18Charged Higgs in the transition region
- Study of charged Higgs in the transition region
MHMtop performed K.Assamagan et al. - Preliminary study has been performed with fast
simulation and is reported in ATL-COM-PHYS-2004-00
9 - Look for gg-gtbtH with H -gt tn 2-gt3 process
with correct treatment of transition region (from
on-shell to off-shell t) HERWIG version 6.4 - Gap in transition region covered
Preliminary
19Getting ready for real data
- More details see report to Physics
commissioning meeting July 2004 (FC) - First step Detector performance and calibration
- e/m/t/g identification and measurement
- jet reconstruction (forward tagging jets
Zjet statistics will be limiting), - PT-miss calibration
- ? performance groups
- Second step estimate of background from
data/control samples - Bkg estimate easy in principle for some
inclusive channels (gg, 4-leptons) - More involved for more exclusive channels (like
VBF WW, tt) - Bkg. Normalization measure of SM processes
(ttnj, Wnj, Znj, WWnj, ggnJ with n0,1,2,)
? SM, QCD and top groups - Clear strategies on extrapolation from control
samples to signal region - Higgs could be discovered (statistically) with
10 fb-1 crucial to have detector
calibration/performance and SM background
understanding with first collected samples (we
have 1 concurrent exp.)
20(No Transcript)
21Participation to DC2
- Higgs working group has requested SB samples for
first DC2 phase and beyond work organized by
B.Mellado-Wisconsin - http//www-wisconsin.cern.ch/bmellado/MC/
- Larger samples (mainly background) will be needed
to finalize some analyses - Validation of generators/ATLFAST for signal
performed - First look at DC2 fully simulated/reconstructed
events still long way to go for FULL DC2
validation
S.Paganis - Wisconsin Very first look !
22Plans for Rome physics workshop
- Complete DC1 based re-analysis of gg and 4-lepton
channels write SN notes with updated results - Finalize systematic studies of VBF tt and WW
channels including impact of full simulation
background systematics included based on control
samples studies - Consistent update of SM discovery sensitivity
- DC2 based analyses for Rome ? Difficult to get
final results probably validation and some
particular cross-checks (NLO samples, initial
detector layout) - Measurement of some of the Higgs properties in
the low mass region (CP, spin, self-coupling) to
be revisited - Finalize updates on some MSSM channels
23Conclusions (1)
- Quite impressive amount of work performed in the
Higgs working group since ATHENS - Improvement in understanding of signal and bkg
processes (MC_at_NLO, RESBOS, K-factors) - Intensive use of DC1 data
- Validation of full simulation
- Impact of detector layout change
- More realistic performance taken into account
- Studies on control samples to estimate bkg. Size
and uncertainty started in many channels - Complementing TDR inclusive channels with more
exclusive ones (in addtion to VBF WW and tt) - H-gtgg1 jet, VBF H-gtgg
- H-gttt 1jet
- H-gtWW 0,1,2 jets
24Conclusions (2)
- Problematic issues
- Full simulation with calibrated objects needed
to finalize realistic analyses PT-miss
calibration, forward jet calibration,. - Interaction between performance/sub-detector
experts and Higgs wg very important November
Physics Week is a very good opportunity/idea - Validation of full simulation/reconstruction
still an issue - Some DC1 problems survived in DC2 production ?
- DC2 validation difficult issue many things
changed at the same time
25Other developments/ongoing works not covered here
- VBF H-gtZZ-gtqqll large mass K.Cranmer et al.
Wisconsin, Weizmann - Update of VFB H-gtWW for masses 150-300 GeV
K.Cranmer et al. Wisconsin - ttH with H-gttt L.Zivkovic et al - Weizmann,
R.Tanaka - Okayama - h0-gta1a1-gtbbtt in NMSSM S.Baffioni-LAL
- MSSM H/A -gt tt -gt ll nnnn T.Szymocha et al. -
Cracow - WH, ZH with H-gtbb J.Vossebeld et al. - Liverpool
- ttH with H-gt invisible B. Kersevan - D.Schroff
- Freiburg - H -gt WH with H-gtbb B.Mohn et al. - Bergen,
Uppsala - tt -gt WbHb X.Portell et al. - IFAE
- Studies of NLO corrections to VBF cross section
R.Mazini-Toronto - ZH with Z-gtnn and H-gtgg G.Azuelos et al. -
Montreal - MSSM h-gtmm for Mh100 GeV S.Gentile et al -
Rome/Fascati - Impact of PDF uncertainty on H production Paris,
Montpellier - Trigger efficiency for H -gt4e - V.Perez-Real-Bern
- t-ID algorithm in fast simulation L.Zivkovic
and E.Gross - Almost all the studies documented in ATLAS notes
26Spare slides
27Introduction (1)
- SM Higgs focus on low mass region- 115-250 GeV
TDR discovery channels mainly inclusive
searches - look for mass peak over (unknown)
continuous background (Design robust analyses) - ttH with H-gt bb
- H-gt g g
- H-gt Z Z -gt 4 leptons
- H-gt W W -gt lnln
- EW Production mechanism strongly reduces QCD
background (tt, WWj, Zj) Vector Boson Fusion VBF
entered in the game - H-gt W W -gt ln ln (-gt ln jj)
- H-gt t t -gt lnn lnn and -gt lnn hn
- (H-gt g g)
28Introduction (2)
- Comments on VBF
- Improvement in difficult region 120 GeV
- Important contribution to coupling measurements
more channels available at each mass (redundancy)
- Larger S/B w.r.t. inclusive channels
- Involves difficult detector aspects forward
jet calibration, central jet veto, PT-miss - Control samples to estimate the background from
the data (counting experiments) - Opened Issues on SM Discovery Potential
- non-VBF channels are from TDR Change in
detector performance/layout ? - LO Cross-section used for H production
(conservative) NLO computations/generators for
S and B ? - Only some aspects (forward jet tagging, jet
veto) studied with full simulation Need to fully
validate VBF results with up to date full
simulation
VBF Discovery potential VBF Scientific Note
2003-024 Bonn, Cracow, KEK, Mainz, Montreal,
Pisa, Sheffield, Tokio, Wisconsin
29H-gt g g from DC1
Tratarelli-Carminati
- Mass reconstruction studied on DC1 samples
special corrections/calibration - Still g-conversion identification from MC truth
instead than from ID recon. - Mass resolution 5-10 worse than in TDR
- (more material IDmore conversions)
30From inclusive gg to gg jet(s) exclusive final
states Clear trend in S/B. S/sqrt(B) precise
predictions difficult
G. Unal
For VBF H-gtgg studies from Wisconsin
ATL-PHYS-2003-036 and Minagawa VBF and inclusive
can be combined
31 t-identification
- Crucial for several channels VBF H-gttt, ttH with
H-gttt, MSSM H/A-gttt with different pt and event
topologies - t-id with full simulation (-gt detector
performance group) - D.Cavalli, M.Heldman standard tau
identification optimization with full simulation - Alternatives t-id algorithm E.Richter-Waas
(track based), J.Tanaka - Problems with t-ID in fast simulation standard
ATLFAST parameterization is not working at low-Pt
and has a strong process-dependence (L.Zivkovic
and E.Gross) - Attempt to perform an event by event t-ID in
ATLFAST (Weizmann) using fast-shower variables
(Mainz) - Not yet satisfactory results obtained with both
approaches work in progress - All channels containing taus studied with fast
simulation ONLY must be validated with full
simulation
32Example tt 1 jet B.Mellado et al. Wisconsin
(Atlfast)
5 sigma for MH120 GeV for 30 fb-1
33MSSM Higgs
- New scan for typical scenarios performed by
M.Schumacher - Atlas note on the way
tt reach for (30 fb-1)
(300 fb-1)
34Trigger issues
- Study of impact of new HLT menu on Higgs channels
performed in collaboration with HLT-TDAQ group - http//cerutti.home.cern.ch/cerutti/new_trigger_me
nu.html - Inclusive approach is in general good no need of
more exclusive triggers - Most problematic channels
- bbh/A with h/A-gtbb (MSSM)
- VBF H-gtbb (SM couplings)
- VBF H-gt Invisible
- Possible solutions for 4-b channels under study
J.Kanzaki and I.Vivarelli b-tag at LVL-2 with
looser PT-jet - Trigger efficiency for H -gt 4e channels studied -
V.Perez-Real-Bern no significant loss between
LVL1 and offline recon.