Marina Artuso Syracuse Uni Jochen Dingfelder SLAC Bjorn

Marina Artuso (Syracuse Uni) Jochen Dingfelder (SLAC) Bjorn Lange (MIT) Antonio Limosani (Uni of Tokyo) Tetsuya Onogi (Yukawa Inst. Kyoto) WG 2 Thursday 14 December 2006 1

Outline • Inclusive Vcb • B -> Xs gamma • Inclusive Vub • Exclusive Vcb • Exclusive Vub 2

Inclusive Vcb 3

Inclusive Vcb • Latest development includes a fit for Vcb in the 1 S scheme performed by Belle (reported P. Urquijo) 2006 • Fit was also performed to data of other experiments (Babar, Belle, CDF, CLEO, DELPHI - 71 measurements) • Compare with fit done similarly in the kinetic scheme by Buchmuller and Flacher circa 2005 1 S scheme Fit (P. Urquijo) Kinetic scheme Fit (H. Flacher) 4

Inclusive Vcb The Vcb and b-quark mass values from both fits are consistent! Hopes for reducing the error • Improved measurements of B->Xs gamma • And theory errors … (M. Trott) T. Becher pointed out that in the discussion that calculations from muon decay could help this effort 5

B->Xs 6

B->Xs Much more data yet to analyse Giving us hope to lower E(cut) to 1. 6 Ge. V 7

B->Xs (P. Gambino) 8

B->Xs (P. Gambino) Total error ~7% = ± 3% (interpolation) ± 3% (parametric) ± 3% (higher orders) ± 5% (non-pert) • SM prediction now lower than experimental finding • Theoretical uncertainty at same level as exp. one • Little room, but bounds on new physics remain stringent. 9

B->Xs (T. Becher) • Becher & Neubert take the Misiak et al OPE result for Egamma > 1. 0 Ge. V and correct for Egamma > 1. 6 Ge. V using NNLO Multi-scale OPE calculation • Theoretical uncertainty at same level as exp. one • Little room, but bounds on new physics remain stringent. 10

B->Xs (M. Neubert) • There is no local OPE for the total B Xs decay rate! • Nonlocal power corrections (subleading shape functions) starting at order QCD/mb exist even for the total decay rate • Irreducible theory error, of similar magnitude as perturbative uncertainty at NNLO • Model estimates suggests effect could lower result by ~5% • Measurement of flavor-dependent asymmetry could help to corroborate our model estimates, model (VIA) estimate shows a -2% to -20% effect. 11

Inclusive Vub 12

Inclusive Vub (Experimental development) • Recently BABAR measured inclusive Vub using the prescription of Leibovich, Low, and Rothstein (LLR) that extracts Vub with reduced model dependence by combining data of the hadronic mass spectrum from B>Xulnu decays with that of the photon energy spectrum from B->Xs gamma decays 13

Inclusive Vub (Theory developments) Interpretation of Ba. Bar lepton endpoint results BLNP includes SSF effects whilst LLR & Neubert do not. 14

Inclusive Vub (Theory developments) Dressed Gluon Exponentiation (Einan Gardi & Jeppe Andersen) 15

Inclusive Vub (Theory developments) • Ongoing effort toward new sub-leading SF effects that are at order alpha_s (G. Paz ) • How to constrain the Weak Annihilation (WA) effect, besides measuring B+/B 0 rate separately? (V. Luth) • Ferrara presented a model calculation for threshold spectra and fitted their predictions to not fully corrected data of B-factories. 16

Exclusive and form factors 17

18

B Xcl 19

D** at the Te. Vatron 20

21

Higher resonances, B D*(np)l 22

More “inclusive” B D/D*/D*(np) l 23

More “inclusive” B D/D*/D*(np) l 24

Comparison Belle and BABAR and Belle results are nicely consistent “D*l puzzle” : Charged-over-neutral BF ratio does not match lifetime ratio (& isospin): 25

- A non-pert technique without using effective theory 26

27

B pl and B rl with Tagging Semileptonic tags: Hadronic tags: 41 ± 9 evts • Very clean signal High purity • Will improve with higher statistics hep-ex/0610054 28

Untagged B pl with “Loose” Reco. • Signal efficiency 4 x higher than previous untagged Ba. Bar result due to looser neutrino selection cuts • Reconstruct q 2 as (p. B - pp)2 s(q 2) ~ 0. 5 Ge. V 2 5072 signal evts • Fit signal and backgrounds in 12 bins of q 2 reduce systematic errors due to bkg BF’s and FF’s. Talk by B. Viaud CLEO has also updated their untagged B pl with full data set (14 fb-1) Talk by S. Stone 29

B pl Branching Fractions #BB 106 BF(B ) 104 CLEO ‘ 06 untagged 15. 4 1. 37 0. 15 stat 0. 12 sys BABAR ’ 06 Untagged 227 1. 44 0. 08 stat 0. 10 sys BELLE ’ 06 had tag 535 1. 49 0. 26 sta t 0. 06 sys HFAG ICHEP ’ 06 Avg 1. 37 0. 06 stat 0. 06 sys CLEO untagged Ba. Bar untagged Belle had. tag 30

B pl Form Factor and |Vub| Ba. Bar: αBK = 0. 52 ± 0. 05 ± 0. 03 LCSR LQCD 31

Comparison with Other Parameterizations Talk by P. Ball Overlay of all 4 parameterizations Best-fit FF’s normalized to BGLa 32

Quenched Unquenched Talk by J. Flynn 33

“Complex Magic” – Dispersion Relations Talk by I. Stewart 34

Other Modes , ’, w @ B factories results for hadronic tags need more statistics wl from Belle: CLEO still competitive with B factories due to higher efficiency at symmetric machine How to determine B r FF’s ? 1/ab @ B factories ~ 5000 events Full 4 D fit to decay angles not possible Need guidance from theory l , ’l from Ba. Bar: B(B+ l+ ) = (0. 84 0. 27 0. 21) x 10 -4 B(B+ l+ ) < 1. 3 x 10 -4 l , ’l from CLEO: B(B+ l+ ) = (0. 44 0. 23 0. 11 0. 00)x 10 -4 B(B+ l+ ) = (2. 66 0. 80 0. 57 0. 04)x 10 -4 35 cl G(B+ l+ )/G(B+ l+ )>2. 5 @90%

36

Form Factors BELLE: PRL 97, 061804 (2006) [hep-ex/0604049] & FOCUS Do Kℓ Unquenched LQCD Quenched LQCD Simple pole model Belle Kℓ ~2. 5 k signal events (hep-ex/0607077) LQCD: Aubin et al. , PRL 94, 011601 (2005) • CLEO-c somewhat disagree with FNAL-MILC-HPQCD prediction, Belle c 2 28/18 (K) & 9. 8/5 ( ) Belle ℓ 232 signal events q 2 (Ge. V 2) 37

CLEO-c form factors for D Pe Assume Vcs = 0. 9745 LQCD FNAL-MILC-HPQCD* D Ke . f+(0)=0. 73± 0. 07 a=0. 50± 0. 04± 0. 07 D e Assume Vcd = 0. 2238 f+(0)=0. 64± 0. 03± 0. 06 a=0. 44± 0. 07 *C. Aubin et al. , PRL 94 011601 (2005) Belle DATA FIT (tagged) CLEO-c LQCD Belle . DATA FIT (tagged) 38

Conclusion A lot of progresses has been made since last CKM workshop both in experiments and in theory, which gave much better understanding of the systematic errors. In some cases this has led to improved precisions, but even if it has not, the errors have become more reliable. 39