You can not select more than 25 topics
Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
45 lines
2.5 KiB
45 lines
2.5 KiB
\topheading{NNLO using non-local subtraction}
|
|
We have performed NLO and NNLO calculations using different methods and published the results in ref.~\ref{Campbell:2022gdq}.
|
|
NLO calculations can performed by three methods, (subtraction, zero-jettiness slicing and $q_T$-slicing.
|
|
NNLO calculations can be performed by zero-jettiness slicing and $q_T$-slicing.
|
|
|
|
\midheading{Benchmark results for NNLO}
|
|
\label{sec:NNLO}
|
|
\label{sec:benchmark}
|
|
We perform benchmark calculations with the default set of EW parameters, (see ref.~\ref{Campbell:2022gdq})
|
|
and for the LHC operating at $\sqrt s = 14$~TeV. We allow all
|
|
vector bosons to be off-shell ({\tt zerowidth} is {\tt .false.})
|
|
and include their decays ({\tt removebr} is {\tt .false.}).
|
|
For each Higgs boson process we consider the decay $H \to \tau^- \tau^+$.
|
|
For parameters that are set in the input file we use,
|
|
\begin{eqnarray}
|
|
m_H = 125~\mbox{GeV} \,, \quad
|
|
m_t = 173.3~\mbox{GeV} \,, \quad
|
|
m_b = 4.66~\mbox{GeV} \,,
|
|
\end{eqnarray}
|
|
and we use the NNLO CT14 pdf set (i.e. {\tt pdlabel} is {\tt CT14.NN}) with
|
|
$\mu_F = \mu_R = Q^2$ (i.e. we set {\tt dynamicscale} equal to
|
|
either {\tt m(34)} or {\tt m(345)} or {\tt m(3456)}, as appropriate).
|
|
Our generic set of cuts is,
|
|
\begin{eqnarray}
|
|
&& p_T(\mbox{lepton}) > 20~\mbox{GeV} \,, \quad
|
|
|\eta(\mbox{lepton})| < 2.4 \,, \quad \nonumber \\
|
|
&& p_T(\mbox{photon 1}) > 40~\mbox{GeV} \,, \quad
|
|
p_T(\mbox{photon 2}) > 25~\mbox{GeV} \,, \quad \nonumber \\
|
|
&&|\eta(\mbox{photon})| < 2.5 \,, \quad
|
|
\Delta R(\mbox{photon 1, photon 2}) > 0.4 \,, \quad \nonumber \\
|
|
&& E_T^{\mbox{miss}} > 30~\mbox{GeV} \,, \quad \Delta R(\text{photon}, \text{lepton}) > 0.3 \quad
|
|
\end{eqnarray}
|
|
For $Z$ production we also impose a minimum $Z^*$ virtuality ({\tt m34min})
|
|
of $40$~GeV.
|
|
|
|
Our benchmark results are shown in Table~\ref{NNLObenchmarks} and were performed on an Intel Xeon X5650 @ 2.67GHz
|
|
system with 16 nodes of 12 cores each. MCFM was compiled with the Intel compiler and default optimizations as well
|
|
as the default MCFM setup including all pre-defined histograms. The NNLO CPU time includes the time necessary
|
|
for the NLO calculation. The numbers for the NLO and NNLO results were obtained independently. By tweaking the
|
|
initial number of calls or the number of iterations per batch it is
|
|
certainly possible to optimize the runtimes. While the numerical precision is not yet good sufficient for most of the
|
|
fitted corrections to significantly improve the results, the fits are highly reliable and correctly estimate the
|
|
residual $\taucut$ dependence.
|
|
|
|
\input{sections/table_benchmark.tex}
|