ch3-00-monte-carlo.tex

\section{Detector Response Simulations}

An accurate simulation of the detector response is an important validation of
a collaboration's understanding of the apparatus as well as a vital tool in
the calculation of some systematic uncertainties. STAR, like many other
collider experiments, uses Monte Carlo routines to generate events with
properties similar to those measured by the experiment. These events are
collections of identified particles with well-defined kinematics. STAR
primarily relies on the PYTHIA 6 event generator \cite{Sjostrand:2006za} in
the CDF Tune A \cite{Field:2005sa} configuration to simulate proton
collisions. A subset of results from PYTHIA have been validated using the
alternative HERWIG event generator \cite{Corcella:2000bw}. The output from
PYTHIA is fed through a GEANT 3.21 \cite{GEANT321} model of the STAR detector;
the end result of this process is a set of detector signals that approximates
the detector response to a real event. Significantly, the output from GEANT
can be processed by the same reconstruction software used for real data.
Reconstruction efficiencies and other quantities useful for the estimation of
systematic uncertainties can be calculated by associating particles in the
output of the reconstruction software with particles in the ``true'' event
record generated by PYTHIA.