Fernando Martínez-Vidal, On behalf of the BaBar Collaboration Measurements of the CKM angle  at BaBar

Fernando Martinez Vidal 1 ICHEP 2010, Paris Apex of the CKM Unitary Triangle (UT) over constrained Precise measurement of SM parameters Search for NP in discrepancies of redundant measurements Still have some work to do on  : less precisely known UT angle (most difficult to measure) CP violation measurements give angles Semileptonic decay rates (and other methods) give sides Very impressive consistency CKM angle  See also UTfit analysis at

Apex of the CKM Unitary Triangle (UT) over constrained Precise measurement of SM parameters Search for NP in discrepancies of redundant measurements Still have some work to do on  : less precisely known UT angle (most difficult to measure) CP violation measurements give angles Semileptonic decay rates (and other methods) give sides 1 Important to measure  since, together with |V ub |, selects  -  value independently of most types of NP SM candle type of measurement The constraint on  come from tree-level B→DK decays Fernando Martinez Vidal ICHEP 2010, Paris Experiments providing most of analyses today 3.5 GeV e + 8 GeV e – 657M B B pairs 3.1 GeV e + 9 GeV e – 468M B B pairs ~10-15 o This talk CKM angle 

γ from B→DK decays * * B B ICHEP 2010, Paris Measure interference between tree amplitudes b→u and b→c Use final state accessible from both D 0 and D 0 Largely unaffected by New Physics Clear theoretical interpretation of observables in terms of γ Difficult because BFs are small due to CKM suppresion ( , so not many events) and r B =|A ub |/|A cb | is small due to further CKM and color suppressions (small interference) All measurements are statistics limited Fernando Martinez Vidal 2

GLW K - K + π + π - K 0 s π 0 K 0 s ω K 0 s Φ CP-even CP-odd CP-conj K+π-K+π- K+π-π0K+π-π0 K + π - π + π - ADS K 0 s π + π - Dalitz plot Complementary methods applied on same B decay modes share the same hadronic parameters (r B,  B ) and  Strategy: many decay chains are analyzed and then combined to improve the overall sensitivity to γ In this talk we present new or recent results Dalitz plot π 0 π + π - K 0 s K + K - CP-conj K 0 s π + π - ADS K + π - CP-conj K + π - π 0 γ from B→DK decays Charged B→D (*)0 K (*) Neutral B→D (*)0 K *0 r B ~0.1r B ~0.3 K + π - π + π - * * B B 3 ICHEP 2010, Paris Atwood, Dunietz, Soni, PRL 78, 3357 (1997) Gronau, London, Wyler, PLB 253, 483 (1991); PLB 265, 172 (1991) f =

γ from B→DK decays b→cb→u b→cb→u Hadronic parameters D 0 decay strong phase variation Hadronic parameters r B =|A(b→u)|/|A(b→c)|  B (strong phase difference between A(b→u) and A(b→c)) determined experimentally from data Dalitz plot method Fernando Martinez Vidal 4 ICHEP 2010, Paris Decay amplitude No D-mixing nor CPV Bondar, unpublished. Giri, Grossman, Soffer, Zupan, PRD 68, (2003);

γ from B→DK decays b→cb→u b→cb→u Relative weak phase  changes sign Decay amplitude Fernando Martinez Vidal 5 ICHEP 2010, Paris Interference terms in decay rates proportional to Fit for x ± and y ± for each B decay mode Dalitz plot method No D-mixing nor CPV

γ from B→DK decays b→cb→u b→cb→u Decay amplitude Fernando Martinez Vidal 6 ICHEP 2010, Paris K0sρK0sρ K0sρK0sρ K0sρK0sρ K0sρK0sρ No D-mixing nor CPV Dalitz plot method Large interference in some regions of the Dalitz plot, e.g. D 0 →K S  D 0 →K *-  + vs D 0 →K *+  - DCS K * (892) + π - CF K * (892) + π - DCS K * (892) + π -

m ES = √ E *2 beam -|p B *| 2 jet-like e+e+ e-e- qq Dπ spherical e+e+ e-e- BB ΔE=E B * -E * beam The main source of background : e + e - → qq, with q=u,d,s,c Event shape variables combined into a linear (Fisher) or non linear (Neural Network) combination. Tagging information sometimes used m ES signal region sidebands Experimental techniques Fernando Martinez Vidal 7 ICHEP 2010, Paris Signal is separated from background through unbinned maximum likelihood fits to B ± → D (*) K (*) ± data using m ES,  E and Fisher (or Neural Network) discriminant Use large B ± → D (*)  ± data control samples (r B ~0.01, x10 smaller than for DK) Exclusive reconstruction of multiple B decays: B ± →DK ±, B ± →D * [D  0 ]K ±, B ± →D * [D  ]K ±, B ± →DK* ± Exploit kinematic constraints from beam energies. ΔE

γ from D Dalitz plot method Fernando Martinez Vidal 8 ICHEP 2010, Paris BaBar analysis based on complete data sample (468 M B B pairs) Reconstruct B ± →DK ±, B ± →D * [D  0 ]K ±, B ± →D * [D  ]K ±, and B ± →DK* ± final states with D→K S π + π -, K S K + K - (eight different final states for each B charge) arXiv: Sub. to Phys. Rev. Lett. B ± →DK ± Only BaBar All components Continuum + B B BG Continuum + B B + D  BG Efficiencies improved substantially (20% to 40% relative) with respect to previous BaBar measurement (383 M B B) Reprocessed data set with improved track reconstruction Improved particle identification Revised K S selection criteria: negligible background from D→  +  - h + h - and B→Da 1 (1260) PRD78, (2008)

D→K S  +  -,K S K + K - amplitude analysis Fernando Martinez Vidal 9 ICHEP 2010, Paris Extract D 0 amplitude from an independent analysis of flavor tagged D 0 mesons (D *+ →D 0 π + ) Experimental analysis using complete data sample benefits from synergy with D-mixing analysis Model determined without (reference) and with D-mixing Fit for amplitudes relative to CP eigenstates [K S  (770) and K S a 0 (980)] and assume no direct CPV See Matt Bellis talk in this session for more details ππ S-wave Kπ S-wave ππ P-wave Kπ P-wave ππ D-wave Kπ D-wave WaveParameterization K-matrix K-matrix (LASS-like) BW: ω(782), G.S ρ(770) BW: CA and DCS K*(892), CA K*(1680) BW f 2 (1270) 0 BW CA and DCS K 2 *(1430) K ± K s S-wave K + K - S-wave K + K - P-wave K + K - D-wave Wave Parameterization BW: CA and DCS a 0 (980), CA a 0 (1450) Flatte a 0 (980), BW a 0 (1450) and f0(1370) BW Φ (1020) BW f 2 (1270) 0 Main differences with Belle Good fit quality (  2 /ndof) taking into account statistical, experimental and model uncertainties arXiv: Accepted by Phys. Rev. Lett.

γ from D Dalitz plot method Fernando Martinez Vidal ICHEP 2010, Paris CP violation parameters are extracted from simultaneous unbinned maximum likelihood fit to B ± → D (*) K (*) ± data using m ES,  E, Fisher and the Dalitz plot distributions (s+,s-) B - →DK - B + →DK + B - →DK - B + →DK + B - →D*K - B + →D*K + B - →DK* - B + →DK* + Differences between B + and B - gives information on  10 arXiv: Sub. to Phys. Rev. Lett.

γ from D Dalitz plot method Use frequentist method to obtain the (common) weak phase  and the hadronic parameters r B,  B (different for each B decay channel) from 12 (x,y) observables stat syst model 3.5  significance of CPV Fernando Martinez Vidal 11 ICHEP 2010, Paris Smaller stat error due to additional data + improved reconstruction + slightly higher r B (DK) Model error benefited from overlap with D-mixing analysis (e.g. reduction of experimental uncertainties inherent to the model uncertainty) Our previous result: [Error on  scales roughly as 1/r B ] arXiv: Sub. to Phys. Rev. Lett.

γ from ADS method Fernando Martinez Vidal 12 ICHEP 2010, Paris Update with final data sample (468 M B B pairs). Previous analysis used 232 M B B pairs Reconstruct B ± →DK ±, B ± →D * [D  0 ]K ±, and B ± →D * [D  ]K ±, with D→K ± π - (“same sign”) D→K -  ± (“opposite sign”) PRD72, (2005) First sign of an ADS signal in B ± → DK ± (2.1  ) and B ± → D * K ± (2.2  ) B ± →DK ± “ Maximizes” CP asymmetry since “equalizes” the magnitudes of the interfering amplitudes arXiv: Sub. to Phys. Rev. D 2.1  “opposite sign” + +

γ from ADS method Fernando Martinez Vidal 13 ICHEP 2010, Paris Fit directly observables R ADS and B ± “same sign” yields to reconstruct the ADS asymmetry B + →DK + B - →DK - Good sensitivity to r B arXiv: Sub. to Phys. Rev. D

γ from ADS method Fernando Martinez Vidal 14 ICHEP 2010, Paris Use frequentist interpretation (similar to Dalitz plot method) to obtain weak phase  and hadronic parameters r B,  B from R ADS (*) and A ADS (*) observables r D = (5.78 ± 0.08)% δ D = (201.9 ) (HFAG, CLEOc) BaBar Dalitz plot method |A(D 0 → K - π + )| r D = |A(D 0 → K - π + )| A(D 0 → K - π + ) δ D = arg [ A(D 0 → K - π + ) ] Inputs: (HFAG) (CLEOc) arXiv: Sub. to Phys. Rev. D

γ from GLW method Fernando Martinez Vidal 15 ICHEP 2010, Paris Update with final data sample (468 M B B pairs). Previous analysis used 383 M B B pairs Reconstruct B ± →DK ± final states with D→K+K-,π + π - (CP even) and D→K S  0,K S ,K S  (CP odd) (five different final states for each B charge) Efficiencies improved substantially (40% to 60% relative) with respect to previous measurement Reprocessed data set with improved track reconstruction Improved particle identification Introduce Fisher discriminant in the fit rather than apply cut on event shape variables PRD77, (2008) B - →DK - All components Continuum + B B BG Continuum + B B + D  BG arXiv: Sub. to Phys. Rev. D

γ from GLW method Fernando Martinez Vidal 16 ICHEP 2010, Paris Extract signal yields fitting directly to observables R ± K/ , R K/ , and A CP± 3.6  significance of CPV in B ± →DK ± from A CP+ arXiv: Sub. to Phys. Rev. D

γ from GLW method Fernando Martinez Vidal 16 ICHEP 2010, Paris Extract signal yields fitting directly to observables R ± K/ , R K/ , and A CP± 3.6  significance of CPV in B ± →DK ± from A CP+ arXiv: Sub. to Phys. Rev. D Consistent with and of similar precision to Dalitz results [Dalitz results]

γ from GLW method Fernando Martinez Vidal 16 ICHEP 2010, Paris Extract signal yields fitting directly to observables R ± K/ , R K/ , and A CP± 3.6  significance of CPV in B ± →DK ± from A CP+ arXiv: Sub. to Phys. Rev. D Dalitz

γ from GLW method Fernando Martinez Vidal 16 ICHEP 2010, Paris Extract signal yields fitting directly to observables R ± K/ , R K/ , and A CP± 3.6  significance of CPV in B ± →DK ± from A CP+ arXiv: Sub. to Phys. Rev. D 4.4  significance of CPV in B ± →DK ± Dalitz+GLW

γ from GLW method Fernando Martinez Vidal 6 ICHEP 2010, Paris Weak sensitivity to r B Use frequentist interpretation (similar to Dalitz plot method) to obtain weak phase  and hadronic parameters r B,  B from R CP± and A CP± observables Fernando Martinez Vidal 17 BaBar Dalitz plot method arXiv: Sub. to Phys. Rev. D

Summary and outlook Fernando Martinez Vidal ICHEP 2010, Paris Recent progress on , the hardest UT angle to measure BaBar Dalitz plot method approach drives the measurement and benefited from synergy with D- mixing analysis (e.g. reduction of experimental uncertainties inherent to the model uncertainty) (see Matt Bellis talk later in this session) First sign of an ADS signal in B ± → DK ± and B ± → D * K ± Compelling evidence of direct CPV in B ± → D(*)K (*)± decays stat syst model 3.5  significance of CPV, B ± → D(*)K(*) ± combined, Dalitz only  significance of CPV, B ± → DK ± only, GLW only

Summary and outlook Fernando Martinez Vidal ICHEP 2010, Paris Recent progress on , the hardest UT angle to measure BaBar Dalitz plot method approach drives the measurement and benefited from synergy with D- mixing analysis (e.g. reduction of experimental uncertainties inherent to the model uncertainty) (see Matt Bellis talk later in this session) First sign of an ADS signal in B ± → DK ± and B ± → D * K ± Compelling evidence of direct CPV in B ± → D(*)K (*)± decays  significance of CPV in B ± →DK ± only, Dalitz+GLW combined

Summary and outlook Recent progress on , the hardest UT angle to measure BaBar Dalitz plot method approach drives the measurement and benefited from synergy with D- mixing analysis (e.g. reduction of experimental uncertainties inherent to the model uncertainty) (see Matt Bellis talk later in this session) First sign of an ADS signal in B ± → DK ± and B ± → D * K ± Compelling evidence of direct CPV in B ± → D(*)K (*)± decays Experiments providing most of analyses today Experiments that can also give results in near future Planned facilities 3.5 GeV e + 8 GeV e – 657M B B pairs 3.1 GeV e + 9 GeV e – 468M B B pairs ~10-15 o ~2-3 o <1 o Need to reduce the error in order to see possible deviations Close to “last word” from BaBar (~10-15 o error) Still statistics limited, even with full data sets BaBar “legacy”  average from GLW, ADS and Dalitz plot methods in progress Fernando Martinez Vidal ICHEP 2010, Paris 18

Backup

Instrumented Flux return (Muon and hadron reco.) Magnet (1.5 T) Electromagnetic Calorimeter (Detection of neutral particles) Drift Chamber (Main tracking system) Silicon Vertex Tracker (Tracking at low momentum and vertex reconstruction) Detector of Internally Reflected Cherenkov (PID, Kaon separation from pions) BaBar data taking ended on 7th April 2008 BaBar recorded 470M B B at Υ (4S) Delivered luminosity Recorded luminosity Recorded Υ (4S): 432 fb -1 Recorded Υ (3S): 30.2 fb -1 Recorded Υ (2S): 14.5 fb -1 Off peak Υ (4S) Υ (3S) Υ (2S) e-e- e+e+ 9GeV 3.1GeV BaBar detector and data sample

γ from D Dalitz plot method Signal yields as used in the final fit for CP parameters Efficiencies improved substantially (20% to 40% relative) with respect to previous BaBar measurement (383 M B B) Reprocessed data set with improved track reconstruction Improved particle identification Revised K S selection criteria: negligible background from D→  +  - h + h - and B→Da 1 (1260) B decay mode BaBar (K S  +  - ) 468 M B B BaBar (K S K + K - ) 468 M B B Belle (K S  +  - ) 657 M B B B ± →DK ± 896±35154±14757±30 B ± →D*(D  0 )K ± 255±2156±11168±15 B ± →D*(D  )K ± 193±1930±783±10 B ± →DK* ± 163±1828±6 (not updated to 657 M B B) PRD78, (2008) PRD81, (2010) arXiv: Sub. to Phys. Rev. Lett.

D→K S  +  -,K S K + K - amplitude analysis See Matt Bellis talk in this session for more details ± ±300 Extract D 0 amplitude from an independent analysis of flavor tagged D 0 mesons (D *+ →D 0 π + ) Experimental analysis using complete data sample benefits from synergy with D-mixing analysis Model determined without (reference) and with D-mixing Fit for amplitudes relative to CP eigenstates [K S  (770) and K S a 0 (980)] and assume no direct CPV arXiv: Accepted by Phys. Rev. Lett.

γ from D Dalitz plot method ICHEP 2010, Paris CP violation parameters are extracted from simultaneous unbinned maximum likelihood fit to B ± → D (*) K (*) ± data using m ES,  E, Fisher and the Dalitz plot distributions (s+,s-) B - →DK - B + →DK + Most precise measurement of (x,y) Consistent results with latest Belle results PRD81, (2010) arXiv: Sub. to Phys. Rev. Lett.

γ from D Dalitz plot method arXiv: Sub. to Phys. Rev. Lett.

γ from D Dalitz plot method arXiv: Sub. to Phys. Rev. Lett.

Other hadronic parameters relevant for  Updated search for B + →D + K 0 and B+→D + K *0 Can only proceed through annihilation or rescattering Allows an estimation of r B 0 for B 0 from r B for B + needed for B 0 →D 0 K 0 measures r B 0 sin(2  +  ) in time-dependent asymmetry B 0 →D 0 K *0 measures  in direct CPV (similar to B + ) BaBar analysis with full data sample: 468 M B B No evidence for signals arXiv: Submitted to PRD

BaBar vs WA Courtesy of V. Tisserand and the CKM fitter group

Backup

CKM angle  d s b u t c Unitarity condition from 1st and 3rd columns Wolfenstein parameterization   The electroweak coupling strength of W ± to quarks described by the CKM matrix Overconstraint apex coordinates CP violation measurements give angles Semileptonic decay rates (and other methods) give sides Precise measurement of SM parameters Search for NP in discrepancies of redundant measurements

CKM angle 

γ from B→DK decays b→cb→u b→cb→u Decay amplitude K0sρK0sρ K0sρK0sρ K0sρK0sρ K0sρK0sρ GLW CP-even CP-odd π + π - K - K + K0sπ0K0sπ0 K0sωK0sω K0sρK0sρ K0sΦK0sΦ GLW-like method No D-mixing nor CPV Strong phases varying over the Dalitz plane Large interferences in some Dalitz plot regions Gronau, London, Wyler, PLB 253, 483 (1991); PLB 265, 172 (1991)

γ from B→DK decays b→cb→u b→cb→u Decay amplitude ADS K+π-K+π- K+π-π0K+π-π0 DCS K * (892) + π - CF K * (892) + π - ADS-like method CF K * (892) + π - DCS K * (892) + π - No D-mixing nor CPV K+π-π+π-K+π-π+π- Atwood, Dunietz, Soni, PRL 78, 3357 (1997) Strong phases varying over the Dalitz plane Large interferences in some Dalitz plot regions