Systematic uncertainty from the simulation of fragmentation

1
Systematic uncertainty from the simulation of
fragmentation
Question All energy deposits from hadronic final
state contribute to measured hadronic angle. How
well does the Monte Carlo describe the
fragmentation? (Double angle relies on good
reconstruction of ?H) How to study the
issue Define the Hadronic energy flow average
energy deposit per event in islands, calculated
in bins of polar angle This quantity is
sensitive to the simulation of the fragmentation
and also to dead material. Islands remaining
after CorAndCut energy correction and backsplash
cut. Cells corresponding to the electron were
discarded Compare hadronic energy flow in data
and MC, compare to MEPS and Ariadne details in
ZN 02-004
2

• Ariadne shows a better general agreement with
  data gt used to extract cross sections nominal
  values
• But Ariadne does not perfectly describe energy
  flow found in data gt need way of quantifying
  systematic uncertainty

Hadronic energy flow in the BCAL in selected bins
of hadronic angle. Current jet peak position
corresponds roughly to hadronic angle
Z (cm)
3
How to quantify systematic uncertainty in
fragmentation? Observation energy flow in data
is in general between MEPS and Ariadne gt MEPS
can be used to quantify uncertainty
Energy flow between proton remnant and current
jet (plotted versus polar angle/gH)
4

• But
• We know Ariadne does a better job in describing
  the data
• This process would overestimate the error (see
  accept fractions)
• The error is clearly asymmetric wrt Ariadne

Energy flow versus polar angle
5
Procedure used

• ? Mean en.flow difference between Ariadne and
  DATA - gives measure of systematic uncertainty in
  simulation of fragmentation
• ? Mean en.flow difference between Ariadne and
  MEPS - gives measure of difference between MC
  fragmentation models
• Error should be ? ? ? to reflect our knowledge

6
There is still one last bit of information
available using ? ? ? we know the error is still
overestimated because gt 80 of data points are
covered by the band (accept fraction) We can
still reduce the size of the error so that ? 68
are inside the error band. The systematic
uncertainties are, finally
?up 0.18 ? ?Ariadne - ?MEPS ?down 0.42 ?
?Ariadne - ?MEPS