ILD benchmark: a study of $e^- e^+ \to \tau^- \tau^+$ at 500 GeV

TL;DR

This study evaluates the reconstruction and polarization measurement of tau pairs at the ILC-500, demonstrating effective event selection, decay mode identification, and polarization sensitivity with different detector models.

Contribution

It provides a detailed analysis of tau pair reconstruction and polarization extraction at ILC-500, comparing two detector configurations and assessing their performance.

Findings

Approximately 60% efficiency in selecting high-mass tau pairs with minimal background.

Decay mode identification accuracy between 60-90%, with high purity.

Polarization measurement precision between 0.5% and 2% across datasets.

Abstract

The process $e^- e^+ \to \tau^- \tau^+$ is of particular interest because the tau lepton polarisation can be reconstructed, allowing its chiral nature to be probed. This note reports on a study of the reconstruction of the di-tau final state at ILC-500, its selection and the reduction of backgrounds, the identification of the tau lepton's decay mode, and on the extraction of the tau leptons' polarisation. The performance of this analysis is studied in two models of the ILD detector, one larger (IDR-L) the other smaller (IDR-S), which differ in the outer radius of the TPC and of the subdetectors beyond, and in the magnetic field strength of the detector solenoid. We find that the high-mass tau-pair events in which at least one tau decays haronically can be selected with an efficiency of around 60%, with a remaining background from non-di-tau processes at the few-% level. Single-prong decay modes $\tau^\pm \to \pi^\pm \nu, \tau^\pm \to \pi^\pm \pi^0 \nu, \tau^\pm \to \pi^\pm \pi^0 \pi^0 \nu$ can be correctly identified in around 60-90% of cases, with sample purities in the range 50-90%, depending on decay mode. The sensitivity to tau polarisation was estimated in the four beam polarisation datasets envisaged for the $4\ ab^{-1}$ of data forseen for ILC-500. Statistical precisions on the polarisation in the different datasets are predicted to be between 0.5 and 2%. While some small performance differences between the two detector models are seen, they have very similar final sensitivity to the polarisation measurement.