D Mixing Plots for Early Summer `99

[Wrong Sign D Page] [CLEO] [HN Home Page]

D Mixing Plots for Early Summer `99

The plots below are based on a fraction of the total CLEO-II.V data that has been processed; 5.6 1/fb of integrated luminosity. The complete dataset (9.1 1/fb) will be processed and analyzed later this summer.

Q and Mass Plots for D*+ -> D0 pi+, D0 ->K-pi+, with D*+ momentum greater than 2.2 GeV/c, for 5.6 1/fb of integrated luminosity. The number of events is about 10100. The excellent resolution in Q (about 200 keV) is primarily due to 3 layer, double sided, silcon vertex detector. The resolution in M is about 6.0 MeV, which reflects improvements in CLEO track fitting due to the Kalman-type algorithm, as well as the reduction in multiple scattering from use of a Helium-Propane mixture in the drift chamber. The events in the Q (M) projection are required to be within 2.4 standard deviations of the nominal M (Q) All cuts employed analysing the Wrong-sign sample (see below) have been applied.
Q and M Plots for D*+ -> D0 pi+, D0 ->K+pi- with D*+ momentum greater than 2.2 GeV/c, for 5.6 1/fb of integrated luminosity. The histogram shows the Wrong-sign data. Backgrounds are suppressed primarily with tight kinematic cuts and secondarily with loose dE/dx cuts. The components of the background and Rws are determined by a four parameter 2D binned likelihood fit (Q vs M). The normalization of background components with distinct shapes are determined separately. We find

Rws=(0.45+-0.09(stat.)+-0.06(sys.))%

The shift in central value from APS was determined to be 1/2 systematic and 1/2 statistical.
Proper Time of the D0 from a reconstructed `primary' vertex to the `secondary' vertex where the D0>K-pi+ decay takes place, for events that pass 2.4 sigma cuts in both K-pi+ mass and Q. Additional cuts are applied (85% efficient for signal) to ensure a well reconstructed flight distance. Typical flight distance is 200 microns. Only the vertical beam profile of CESR is small (20 microns FWHM) relative to 200 microns so the vertical dimension alone determines the proper-time. The gaussian core resolution is about 1/2 of a lifetime; there are tails, which can be seen on a logarithmic plot
Description of our parameterization of the time evolution of D0 ->K-pi+.
Transparency of our parameterization of the time evolution of D0 ->K-pi+.
Proper Time of the wrong-sign sample. Top: Overlay is Right-sign data normalized to number of events in Wrong-sign plot. Note that 40% of the events in Wrong-sign sample are background. Bottom: Overlay is Monte Carlo prediction for Wrong-sign proper time with Rws=Rdcsd. The same plot with an additional overlay on the bottom plot where Rws=Rmix=x'^2/2 (all mixing is through x') is shown in blue. This corresponds to x'=.095.

We are studying the origin of the difference between wrong-sign data and Monte Carlo prediction.
Current CLEO II.V Mixing Limits are identical to those presented at APS. These are a conservative estimate given our current understanding of the data. The red region is the 90% allowed region from the D0->K+pi- analysis. The difference in the lifetime of CP+ (D0 ->K+K-,pi+pi-) and CP average (D0 ->K-pi+) samples allows one to probe y directly. The 90% allowed limits are shown assuming the difference in FSI between
D0 ->K-pi+ and D0B ->K-pi+ can be neglected.
Our sensitivly estimates once the full 9.1 1/fb is analyzed are below. 1 standard deviation will be

0.020 for x'
0.014 for y'
2 x 10-4 for Rmix

Please see description above for definition of x' and y'

Here are some additional plots added for Marina Artuso

worldm.ps is the world's previous allowed regions in the R_mixing vs. R_DSCD plane, for y'=0 (cos(phi)=0). Yellow is E791 K-Lepton-nu 90%, cyan is Aleph 95%, magenta is CLEO-II Kpi 1 sigma; black ellipse is E691, where horizontal is all Kpi, and vertical combines Kpi and K3pi, 90%; the red is E791 1 sigma, where again vertical is combo of Kpi and K3pi, and horizontal is all Kpi. Here is an overlay showing the current CLEO II.V 90% limit using 5.6 1/fb.

tauvrat.ps shows how the mean wrong-sign lifetime varies as a function of the ratio of the mixing rate to the total rate. When there is no interference (that is, cos(phi)=0, all mixing proceeds through the real part of the virtual transitions), the mean wrong-sign lifetime goes from 1 tau_d to 3 tau_d as mixing goes from none to all of the wrong sign rate. When interference is involved, and the mixing is small, the lifetime goes up faster near the left side of the plot, when y'>0,x'=0 (cos(phi)=1), (all mixing through real transitions, dominated by the CP + states), relative to y'=0 (cos(phi)=0), and vice versa for y'<0,x'=0 (cos(phi)=-1) (real transitions, CP - states dominant). At the right hand side of the plot, the tendencies of + and - are reversed, in essence because the influence of interference is larger in the normalizing denominator than in the numerator that has the t integral.

[Wrong Sign D Page] [CLEO] [HN Home Page]

hnn@charm.physics.ucsb.edu
asner@mail.lns.cornell.edu

Last updated 6/21/99