Comparison of JETSET, HERWIG and ARIADNE Fragmentation Models in the WW ightarrow $(e/mu)nuqqbar$ Channels at $sqrt{s} = 189$ GeV

This work reports on studies done with a large set of Monte Carlo (1 million WW events with $\sqrt{s} =$ 188.6 GeV) that have undergone the fragmentation procedure via three different fragmentation models (JETSET, HERWIG and ARIADNE). The study examines the $WW \rightarrow (e/\mu)\nu q \overline{q}$ channels and compares the same variables between the different fragmentation models. The work is done in the context of wishing to assign a well-understood systematic error on the measurement of the W mass associated with the uncertainties due to fragmentation. Important highlights of the work are: - Understanding of kinematical differences between fragmentation models; - Demonstration that JETSET and HERWIG behave significantly differently whereas JETSET and ARIADNE appear to be indistinguishable; - Demonstration and understanding that the application of a beam energy constraint significantly softens the impact of observable differences between JETSET and HERWIG; - A number of observations are made that the differences between JETSET and HERWIG can be considered to come from different particle production rates in those models, particularly baryon rates; - Connection between gluon radiation and the biases on the W mass due to fragmentation; - Demonstration that the reconstructed hadronic system in the $e\nu q \overline{q}$ channel is the same as that in the $\mu\nu q \overline{q}$ channel; - Documentation of results quoted for Moriond 2001; - Discussion of future studies.