1: \documentclass[12pt]{article}
2:
3: \usepackage{cite}
4: \usepackage{epsfig}
5: \usepackage{graphics}
6: \usepackage{inputenc}
7: \inputencoding{latin1}
8:
9: \textheight 610pt
10: \textwidth 420pt
11: \oddsidemargin 0cm
12: \parindent 0mm
13: \parskip 2mm
14:
15: \def\kt{$K_\perp$}
16: \def\ycut{$y_{\rm cut}$}
17: \def\lum{$\mathcal{L}$}
18:
19: \begin{document}
20:
21: \sloppy
22:
23: \begin{titlepage}
24:
25: {\par\raggedleft
26: \texttt{February} \texttt{2007}\\
27: CERN-PH-TH/2007-019\par}
28: \bigskip{}
29:
30: \bigskip{}
31: {\par\centering \textbf{\large Reconstructing Sparticle Mass Spectra using
32: Hadronic Decays}
33: \large \par}
34: \bigskip{}
35:
36: {\par\centering J. M. Butterworth$^{1}$, John Ellis$^{2}$ and A. R.
37: Raklev$^{2,3}$\\
38: \par}
39: \bigskip{}
40:
41: {\par\centering
42: {\small $^1$ Department of Physics \& Astronomy, University College
43: London}\\
44: {\small $^2$ Theory Division, Physics department, CERN}\\
45: {\small $^3$ Department of Physics \& Technology, University of Bergen}\\
46: \par}
47:
48: \bigskip{}
49:
50:
51: \begin{abstract}
52: \noindent
53: Most sparticle decay cascades envisaged at the Large Hadron Collider
54: (LHC) involve hadronic decays of intermediate particles. We use
55: state-of-the art techniques based on the \kt\ jet algorithm to
56: reconstruct the resulting hadronic final states for simulated LHC
57: events in a number of benchmark supersymmetric scenarios. In
58: particular, we show that a general method of selecting preferentially
59: boosted massive particles such as $W^\pm, Z^0$ or Higgs bosons
60: decaying to jets, using sub-jets found by the \kt\ algorithm,
61: suppresses QCD backgrounds and thereby enhances the observability of
62: signals that would otherwise be indistinct. Consequently,
63: measurements of the supersymmetric mass spectrum at the per-cent level
64: can be obtained from cascades including the hadronic decays of such
65: massive intermediate bosons.
66: \end{abstract}
67:
68: %\pacs{12.38.Aw}
69: %{\it PACS:13.87.Ce, 13.87.Fh, 13.65.+i } \\
70: {\it Keywords: Monte Carlo; Supersymmetry; Jets; LHC}
71:
72: \end{titlepage}
73:
74: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
75: \section{Introduction}
76: \label{sec:intro}
77: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
78:
79: In a number of commonly considered supersymmetric (SUSY) models,
80: strongly-interacting sparticles will be abundantly pair-produced at
81: the Large Hadron Collider (LHC), and the resulting events containing
82: jets and missing energy will stand out above the Standard Model
83: background. The next challenge will be to make sense of the events,
84: which will typically involve cascade decays through several
85: intermediate sparticles. It will be particularly desirable to obtain
86: information on the SUSY masses and branching ratios. It has been
87: demonstrated that cascades involving lepton emission often provide
88: relatively clean signals, which can be used to reconstruct several
89: sparticle masses in some favourable benchmark scenarios, see
90: e.g.~\cite{Hinchliffe:1996iu,edges}. However, these decay channels
91: often suffer from branching ratios that are low compared to hadronic
92: decay modes, reducing the available sample size and therefore
93: restricting access to high-mass sparticles. Moreover, the charged
94: leptons are often associated with neutrinos, for example in chargino
95: ($\chi^\pm$) decays. The presence of the neutrinos would confuse the
96: interpretation of the missing-energy signal provided by escaping
97: neutralinos in scenarios with $R$-parity conservation, as we assume
98: here, and make the reconstruction of chargino masses difficult. Some
99: progress on techniques for measuring chargino masses has been reported
100: for scenarios with one leptonic and one hadronic decay
101: chain~\cite{Nojiri:2003tv}, but these require models with a favourable
102: combination of the corresponding branching ratios. The ability to
103: reconstruct purely hadronic cascade decays would facilitate the
104: discovery and measurement of charginos in more general cases, as well
105: as enable heavier sparticle masses to be reconstructed.
106:
107: In this paper, we take a new approach to the reconstruction of
108: fully-hadronic sparticle events, starting from the missing-energy
109: signal provided by the neutralino in $R$-conserving models, which
110: suppresses Standard Model (SM) backgrounds. We use jet analysis
111: techniques based on the \kt\ algorithm to identify the hadronic decays
112: of massive bosons such as the $W^\pm$, $Z^0$ or Higgs boson ($h$). In
113: general, the majority of sparticle cascade decays yield fully-hadronic
114: final states containing such massive SM particles decaying to $q -
115: {\bar q}$ jet pairs~\cite{DeRoeck:2005bw}. A certain fraction of these
116: bosons, dependent on the sparticle masses in the scenario considered,
117: are highly boosted and the jets from the decays are closely aligned,
118: indeed overlapping, in pseudo-rapidity ($\eta$) - azimuthal angle
119: ($\phi$) space. Such a situation presents both a challenge and an
120: opportunity. The challenge is that the choice of jet finder becomes
121: crucial: the treatment of jet overlaps can be sensitive to rather soft
122: and subtle QCD effects, and yet somehow one needs to retain the
123: information that the jet arises from the two-body decay of a massive
124: particle. The opportunity is that, since the jets are close together,
125: there is no combinatorial background; no need, for example, to combine
126: all pairs of jets to see if they reconstruct to the $W$ mass.
127:
128: We address the reconstruction issue using the sub-jet technique that
129: was proposed previously as a way to identify high-energy $WW$
130: events~\cite{Butterworth:2002tt}. This technique and its extension to
131: hadronic sparticle decays are described in
132: Section~\ref{sec:subjet}. Then, in Section~\ref{sec:benchmarks} some
133: specific benchmark SUSY scenarios and the specific decay chains of
134: interest are described: these involve typical decays of intermediate
135: heavy charginos or neutralinos such as $\tilde\chi^\pm_1\to
136: W^\pm\tilde\chi^0_1$ and $\tilde\chi^0_2\to Z^0, h\tilde\chi^0_1$. The
137: simulations and analysis method are described in
138: Section~\ref{sec:analysis}, and the subsequent Section contains the
139: results and conclusions.
140:
141: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
142: \section{Jet Finding and Sub-jets}
143: \label{sec:subjet}
144: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
145:
146: The identification of jets which originate from a decaying massive
147: particle begins by using a jet algorithm to define the jet. In this
148: analysis the \kt\ algorithm~\cite{Catani:1993hr} is used, in the
149: inclusive mode with the $E$ reconstruction scheme. For each particle
150: $k$ and pair of particles $(k,l)$, the algorithm calculates the
151: quantities
152: \begin{eqnarray}
153: d_{kB} & = & p_{Tk}^2, \nonumber \\
154: d_{lB} & = & p_{Tl}^2, \nonumber \\
155: d_{kl} & = & \min(p^2_{Tk},p^2_{Tl})R^2_{kl}/R^2,
156: \end{eqnarray}
157: where $p_{Tk}$ is the transverse momentum of particle $k$ with respect
158: to the beam axis and
159: \begin{equation}
160: R^2_{kl}=(\eta_k-\eta_l)^2+(\phi_k-\phi_l)^2.
161: \end{equation}
162: The parameter $R$ is a number of order one. If it is set below unity,
163: it is less likely that a given particle will be merged with a jet,
164: which in turn leads to narrower jets. Thus $R$ plays a somewhat
165: similar role to the adjustable cone radius in cone algorithms. In this
166: analysis, initial studies indicated that $R = 0.7$ provided a good
167: compromise between efficiency and the mass resolution, and this value
168: is used throughout. Returning to the algorithm: all the $d$ values are
169: then ordered. If $d_{kB}$ or $d_{lB}$ is the smallest, then particle
170: $k$ or $l$ is labelled a jet and removed from the list. If $d_{kl}$ is
171: the smallest, particles $k$ and $l$ are merged by adding their
172: four-momenta. The list is recalculated and the process is repeated
173: until the list is empty. Thus the algorithm clusters all particles
174: into jets, and a cut on transverse momentum can then be used to select
175: the hardest jets in the event. The algorithm is infrared safe, and has
176: the additional benefit that each particle is uniquely assigned to a
177: single jet. Recently a fast implementation of the algorithm has been
178: developed~\cite{fastjet} which makes it practical for use even in the
179: very high multiplicity events expected at the LHC.
180:
181: In selecting a candidate for a hadronic decay of the $W^\pm, Z^0$ or
182: $h$, first cuts on the $p_T$ and the pseudo-rapidity ($\eta$) of the
183: jet are applied so as to ensure that they are energetic enough to
184: contain a boosted heavy-particle decay and are in a region of good
185: detector acceptance. A cut is then applied on the mass of the jet
186: (calculated from the four-vectors of the constituents) to ensure that
187: it is in a window around the nominal mass of the desired particle.
188:
189: The next step is to force the jet to decompose into two sub-jets. The
190: main extra piece of information gained from the sub-jet decomposition
191: is the $y$ cut at which the sub-jets are defined: $y \equiv
192: d_{kl}/(p_T^{\rm jet})^2$, where $p_T^{\rm jet}$ is the transverse
193: momentum of the candidate jet containing the sub-jets $k$ and $l$. In
194: the case of a genuine $W^\pm, Z^0$ or $h$ decay, the expectation is
195: that the scale at which the jet is resolved into sub-jets
196: (i.e., $yp_{T}^{2}$) will be ${\cal {O}}(M^{2})$, where $M$ is the
197: $W^\pm, Z^0$ or $h$ mass. For QCD jets initiated by a single quark or
198: gluon, the scale of the splitting is expected to be substantially
199: below $p_T^2$, i.e., $y \ll 1$, since in the region around the jet
200: strongly-ordered DGLAP-like~\cite{DGLAP} QCD evolution dominates.
201:
202: This distinction provides new information in addition to the jet mass
203: itself, as is illustrated in Figures~\ref{fig:scale}a
204: and~\ref{fig:scale}b, where the correlation between the jet mass and
205: the splitting scale is shown for $W^\pm$ jets and QCD jets
206: respectively. The events shown are $W^\pm$+jet events and SUSY events
207: generated using {\sc Pythia 6.408}~\cite{pythia}, and $W^\pm$+3jet
208: events generated using ALPGEN~\cite{alpgen} for the matrix element,
209: HERWIG~6.510~\cite{herwig} for parton showering and {\sc
210: Jimmy}~\cite{jimmy} for the underlying event. The parameters for the
211: underlying event and the parton showers were those of the ATLAS tune
212: of {\sc Pythia} and the CDF tune A of {\sc HERWIG} and {\sc Jimmy},
213: taken from~\cite{tunes}~\footnote{More details on the SUSY event
214: generation are given in Section~\ref{sec:analysis}.}. These models
215: have been shown to give a good description of a wide variety of
216: data. In particular, the modelling of the internal jet structure by
217: leading-logarithmic parton showers is known to be good for jets
218: produced in $p\bar{p}$ collisions~\cite{tevjets}, $ep$ collisions and
219: photo-production~\cite{herajets}, and in $e^+e^-$ annihilation events
220: and $\gamma\gamma$ collisions~\cite{lepjets}.
221:
222: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
223: \begin{figure}
224: \begin{center}
225: \epsfxsize 15cm
226: \epsfbox{scale.eps}
227: \caption{Scatter plot of the jet mass against the jet splitting scale
228: $yp_T^2$ for (a) jets from $W^\pm$ decays (determined by a match better than
229: 0.1 units in $\eta-\phi$) and (b) QCD jets in $W^\pm$+jet and SUSY
230: events. The distributions of the splitting scale are shown in (c) and
231: the $y$ distributions in (d), for the same types of jets, after a cut
232: on the jet mass at $75 < M < 90$~GeV. In all these plots, the
233: requirement $p_T > 250$~GeV is applied to all jets. The histograms in
234: the lower two plots are normalised to unity.}
235: \label{fig:scale}
236: \end{center}
237: \end{figure}
238: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
239:
240: Although there are some differences between the results of the {\sc
241: Pythia} and HERWIG/ALPGEN simulations, the conclusions are
242: similar. Fig.~\ref{fig:scale}a confirms that the splitting scale in
243: $W^\pm$ decays is relatively large, typically $\ge 20$~GeV, whereas
244: Fig.~\ref{fig:scale}b shows that the splitting scale for QCD jets with
245: masses $\sim 80$~GeV is typically $\le 20$~GeV. The distributions of
246: of $p_{T}\sqrt{y}$ and of $y$ are shown in Figures~\ref{fig:scale}c
247: and~\ref{fig:scale}d for $W^\pm$ jets and QCD jets. The distributions
248: are qualitatively similar, whether they are generated in SUSY events
249: or in conventional $W^\pm$+jets events, and whether (in the latter
250: case) they are generated using {\sc Pythia} or HERWIG. The scale of
251: the splitting is seen to be peaked close to the $W^\pm$ mass for
252: genuine $W^\pm$ decays, whichever the environment in which they are
253: produced, and to be softer for QCD jets which just happen to
254: reconstruct to the $W^\pm$ mass, again in both environments.
255:
256: This distinction was noted and successfully used to identify $W^\pm$
257: decays in simulations of $W^+W^-$ scattering at LHC energies
258: in~\cite{Butterworth:2002tt}. Here, this information is used in the
259: analysis described in Section~\ref{sec:analysis} to refine the
260: selection of $W^\pm, Z^0$ and $h$ candidates in sparticle decay
261: cascades. Although no detector simulation is employed in this
262: analysis, the cut on the scale has been shown previously to be robust
263: against the effects of the underlying
264: events~\cite{Butterworth:2002tt}, as well as effects due to the
265: calorimeter granularity and resolution~\cite{theses}.
266:
267: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
268: \section{SUSY Decay Chains and Benchmark Scenarios}
269: \label{sec:benchmarks}
270: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
271:
272: The pair-production of heavy sparticles in generic SUSY models yields
273: on-shell electroweak gauge bosons or Higgs bosons via the decays of
274: heavy gauginos that appear as intermediate steps in cascades, e.g.,
275: ${\tilde q}_L \to \tilde\chi^0_2 q$ or ${\tilde
276: q}_L\to\tilde\chi^\pm_1 q'$, followed by $\tilde\chi^0_2 \to Z^0, h +
277: \tilde\chi^0_1$ or $\tilde\chi^\pm \to W^\pm +
278: \tilde\chi^0_1$ decays. Exceptions are cases with gaugino masses that are
279: nearly degenerate, in which there may be large branching ratios for
280: decays via off-shell heavy bosons. Since the largest branching ratios
281: for $W^\pm, Z^0$ and $h$ decays are those into hadronic ${\bar q} q$
282: final states, purely hadronic final states dominate in cascade decays
283: via on-shell bosons, and these are also potentially important in
284: off-shell decays.
285:
286: For the decay chain
287: \begin{equation}
288: \tilde q_L \rightarrow \tilde\chi_1^\pm q' \rightarrow \tilde\chi_1^0
289: W^\pm q'
290: \label{eq:chargino}
291: \end{equation}
292: shown in Fig.~\ref{fig:qW}, one can demonstrate that the invariant
293: mass distribution of the quark-$W$ system has a minimum and maximum
294: given by
295: \begin{equation}
296: (m_{qW}^{\rm{max}\slash\rm{min}})^2=
297: m_W^2+\frac{(m_{\tilde q_L}^2-m_{\tilde\chi_1^\pm}^2)}{m_{\tilde\chi_1^\pm}}
298: (E_W\pm|\vec{p}_W|),
299: \label{eq:endqW}
300: \end{equation}
301: where
302: \begin{equation}
303: |\vec{p}_W|^2=
304: \frac{(m_{\tilde\chi_1^\pm}^2-m_{\tilde\chi_1^0}^2-m_W^2)^2-4m_{\tilde\chi_1^0}^2m_W^2}{4m_{\tilde\chi_1^\pm}^2}
305: \label{eq:pW}
306: \end{equation}
307: is the $W$ momentum in the chargino rest frame. If measurable, these
308: endpoints give a model-independent relationship between the three SUSY
309: masses, modulo the existence of the decay chain. If both endpoints can
310: be determined experimentally, the squark mass can be eliminated,
311: giving the chargino mass in terms of the lightest
312: neutralino. Supplementary model-dependent assumptions, such as the
313: relationship $m_{\tilde\chi_1^\pm} \approx 2m_{\tilde\chi_1^0}$ that
314: holds approximately in the constrained MSSM (CMSSM), can then be used
315: to determine individual masses of all three sparticles in specific
316: theoretical frameworks.
317:
318: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
319: \begin{figure}
320: \begin{center}
321: \psfig{file=qW.eps,width=7.0cm}
322: \caption{The SUSY decay chain of Eq.~(\ref{eq:chargino}).}
323: \label{fig:qW}
324: \end{center}
325: \end{figure}
326: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
327:
328: The decay chains
329: \begin{equation}
330: \tilde q_L \rightarrow \tilde\chi_2^0 q \rightarrow \tilde\chi_1^0 h q
331: \label{eq:chi02h}
332: \end{equation}
333: and
334: \begin{equation}
335: \tilde q_L \rightarrow \tilde\chi_2^0 q \rightarrow \tilde\chi_1^0 Z^0 q,
336: \label{eq:chi02Z}
337: \end{equation}
338: involving the second lightest neutralino, have endpoint formulae with
339: the same structure, given by the substitutions
340: \begin{equation}
341: W\rightarrow h, \quad \tilde\chi_1^\pm\rightarrow\tilde\chi_2^0
342: \end{equation}
343: and
344: \begin{equation}
345: W\rightarrow Z, \quad \tilde\chi_1^\pm\rightarrow\tilde\chi_2^0,
346: \end{equation}
347: respectively, in Eqs.~(\ref{eq:endqW}) and (\ref{eq:pW}).
348:
349: In~\cite{DeRoeck:2005bw} benchmark models with such cascade decays
350: were explored in a more generic setting than that allowed in the
351: CMSSM. Relaxing the GUT-scale universality of the Higgs scalar masses
352: or introducing a gravitino dark matter candidate admits values of the
353: (supposedly universal) scalar squark and slepton mass $m_0$ outside
354: the narrow range allowed by the cold dark matter
355: density~\cite{Bennett:2003bz,Spergel:2003cb} within the restrictive
356: CMSSM framework. This, in turn, allows for a wide range of values for
357: the branching ratios of the lightest chargino and the next-to-lightest
358: neutralino into $W^\pm, Z^0$ and $h$. In our study, we use the
359: benchmark points $\alpha$, $\beta$, $\gamma$ and $\delta$
360: from~\cite{DeRoeck:2005bw} to illustrate these possibilities. The
361: points $\alpha$, $\beta$, $\gamma$ are scenarios with relatively light
362: SUSY masses, that should be easy to probe at the LHC, while $\delta$
363: represents a more challenging scenario with lower cross sections. Common to
364: all four points is the large branching ratio for squark to
365: chargino/neutralino decays, $\rm{BR}(\tilde q_L\to\tilde\chi_1^\pm
366: q)\approx 60$\% and $\rm{BR}(\tilde q_L\to\tilde\chi_2^0 q)\approx
367: 30$\%. The corresponding chargino and neutralino bosonic branching
368: ratios are shown in Table~\ref{tab:BRs}.
369:
370: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
371: \begin{table}\begin{center}
372: \begin{tabular}{cccc} \hline
373: Point/BR & $\tilde\chi_2^0\rightarrow\tilde\chi_1^0 Z$
374: & $\tilde\chi_2^0\rightarrow\tilde\chi_1^0 h$
375: & $\tilde\chi_1^\pm\rightarrow\tilde\chi_1^0 W^\pm$ \\ \hline
376: $\alpha$ & 98.6 & 0.0 & 99.6 \\
377: $\beta$ & 7.5 & 64.5 & 79.0\\
378: $\gamma$ & 0.0 & 0.0 & 99.9 \\
379: $\delta$ & 5.4 & 92.0 & 97.5 \\ \hline
380: \end{tabular}
381: \caption{Branching ratios for $\tilde\chi_2^0$ and $\tilde\chi_1^\pm$ in the
382: selected SUSY benchmark models. The decays are calculated using {\sc
383: SDECAY~1.1a}~\cite{Muhlleitner:2003vg}.}
384: \label{tab:BRs}
385: \end{center}
386: \end{table}
387: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
388:
389: We show the predicted shape of the invariant mass distributions for
390: the kinematically-allowed quark-boson combinations in the decay chains
391: (\ref{eq:chargino}), (\ref{eq:chi02h}) and (\ref{eq:chi02Z}), for all
392: the considered benchmark points, in Fig.~\ref{fig:imtruth}. These
393: distributions consider only the kinematics of the decay chain and
394: assume zero width for all particles. The distributions have a
395: characteristic trapezoidal shape with only small variations in the
396: location of the upper edge with the changing boson mass. This
397: similarity is due to the neutralino/chargino mass degeneracy typical
398: in CMSSM models, and demonstrates that the upper endpoint given by
399: Eq.~(\ref{eq:endqW}) is to a large extent insensitive to the boson
400: mass. Locating the upper endpoints for two different decay chains
401: involving the chargino and neutralino, respectively, would be an
402: excellent test of this mass degeneracy. If they are similar, this would point towards SUSY
403: scenarios where both the second-lightest neutralino and the lightest
404: chargino are `wino-like'.
405:
406: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
407: \begin{figure}
408: \begin{center}
409: \epsfxsize 15cm
410: \epsfbox{imtruth.eps}
411: \caption{The invariant mass distributions of $qW^\pm$, $qZ$ and $qh$
412: combinations for the chosen SUSY benchmark points.}
413: \label{fig:imtruth}
414: \end{center}
415: \end{figure}
416: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
417:
418: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
419: \section{Simulation and Analysis}
420: \label{sec:analysis}
421: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
422:
423: \subsection{Simulation}
424:
425: In order to simulate sparticle pair-production events at the LHC, we
426: use {\sc Pythia~6.408}~\cite{pythia} with {\sc CTEQ 5L}
427: PDFs~\cite{Lai:1999wy} interfaced to {\sc HZTool}~\cite{hztool}, with
428: some minor changes to allow for simulations of SUSY scenarios. Decay
429: widths and branching ratios for the SUSY particles are calculated with
430: {\sc SDECAY~1.1a}~\cite{Muhlleitner:2003vg}. For the $\alpha$, $\beta$
431: and $\gamma$ benchmark points we simulate a number of SUSY events
432: equivalent to $30$~$\rm{fb}^{-1}$, giving results that should be
433: representative of the early reach of the LHC experiments at low
434: luminosity. For the $\delta$ benchmark point, with its considerably
435: higher masses and lower cross sections, we simulate a number of events
436: equivalent to $300$~$\rm{fb}^{-1}$, yielding results that should
437: indicate the ultimate reach of the LHC experiments with their design
438: luminosity.
439:
440: We have also generated SM backgrounds with {\sc Pythia} in five $p_T$
441: bins from $p_T=50$~GeV to $7$~TeV. These samples rely on the parton
442: shower to simulate extra jets. This should be a reasonable
443: approximation in the important kinematic regions for some processes, such as $t \bar{t}$,
444: where the scale of the hard interaction is $>350$~GeV and we rely on
445: parton showers to simulate jets at around 150-200~GeV. However, this
446: is not adequate in all cases, and so in addition we have investigated
447: other possibly important sources of background, by using ALPGEN/HERWIG
448: to generate final states containing dibosons plus one or two jets, or
449: a single boson plus two or three jets~\footnote{Appropriate cuts,
450: considering the final-state selections to be used in
451: Section~\ref{sec:cuts}, have been applied to the partons to reduce the
452: amount of event generation required, and the ALPGEN parton-shower
453: matching scheme was used where appropriate.}. In these cases the jet
454: multiplicities are not well modelled by parton showers, but the
455: internal jet structure should still be well described. These
456: backgrounds turn out not to be very important for the $\alpha, \beta$
457: and $\gamma$ scenarios, but are found to be significant in the
458: $\delta$ scenario due to the small SUSY production cross section. For
459: the multiple jet processes there is a small amount of double counting
460: with the {\sc Pythia} samples, which we neglect here, giving a
461: conservative estimate of the backgrounds in this sense. The
462: generated numbers of events for the various processes, per $p_T$ bin
463: where used, are shown in Table~\ref{tab:efficiency} along with the
464: corresponding integrated luminosities.
465:
466: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
467: \begin{table}\begin{center}
468: \begin{tabular}{lrrrr} \hline
469: Sample & $N_{\rm{generated}}$ & \lum\ [fb$^{-1}$] & $N_{\rm{pass}}(\alpha-\gamma)$ & $N_{\rm{pass}}(\delta)$ \\ \hline
470: $t\bar t$ & & & 256.7 & 1287.0 \\
471: \ 50-150 & 26,500,000 & 93.0 & & \\
472: \ 150-250 & 10,000,000 & 95.6 & & \\
473: \ 250-400 & 3,500,000 & 120.0 & & \\
474: \ 400-600 & 500,000 & 129.6 & & \\
475: \ 600- & 500,000 & 902.4 & & \\
476: $Wj$ & & & 5.2 & 34.5 \\
477: \ 50-150 & 1,100,000 & 0.1 & & \\
478: \ 150-250 & 1,100,000 & 2.9 & & \\
479: \ 250-400 & 1,100,000 & 20.2 & & \\
480: \ 400-600 & 1,100,000 & 154.3 & & \\
481: \ 600- & 600,000 & 507.2 & & \\
482: $Zj$ & & & 3.2 & 3.0 \\
483: \ 50-150 & 100,000 & 0.0 & & \\
484: \ 150-250 & 100,000 & 0.6 & & \\
485: \ 250-400 & 100,000 & 4.3 & & \\
486: \ 400-600 & 100,000 & 32.7 & & \\
487: \ 600- & 100,000 & 199.7 & & \\
488: $Wjj$ & 157,800 & 114.5 & 49.2 & 450.5 \\
489: $Zjj$ & 112,000 & 99.9 & 43.9 & 417.7 \\
490: $Wjjj$ & 50,300 & 227.9 & 127.8 & 1109.4 \\
491: $Zjjj$ & 27,300 & 156.6 & 194.4 & 1782.9 \\
492: $WW/WZ/ZZ$& & & 9.6 & 95.3 \\
493: \ 50-150 & 100,000 & 1.8 & & \\
494: \ 150-250 & 100,000 & 29.2 & & \\
495: \ 250-400 & 100,000 & 158.2 & & \\
496: \ 400-600 & 100,000 & 945.2 & & \\
497: \ 600- & 10,000 & 437.0 & & \\
498: $WWj$ & 201,200 & 100.7 & 9.8 & 98.3 \\
499: $WZj$ & 162,400 & 90.2 & 0.0 & 0.0 \\
500: $ZZj$ & 69,500 & 426.5 & 2.3 & 17.6 \\
501: $WWjj$ & 107,300 & 98.7 & 23.4 & 215.8 \\
502: $WZjj$ & 179,000 & 248.4 & 55.2 & 455.5 \\
503: $ZZjj$ & 18,900 & 167.0 & 5.9 & 59.3 \\ \hline
504: \end{tabular}
505: \caption{The numbers of generated events, separated by $p_T$ bin
506: where used, the corresponding integrated luminosities
507: and the numbers of events passing the cuts for the $qW$ distribution, as
508: described in Section~\ref{sec:cuts}. Not shown are $2\to 2$ QCD
509: events, which are found not to contribute.}
510: \label{tab:efficiency}
511: \end{center}
512: \end{table}
513: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
514:
515: No attempt to simulate detector effects has been made, but we have
516: included semi-realistic geometrical requirements for jets and leptons,
517: restricting ourselves to leptons with $|\eta|<2.5$ and jets with
518: $|\eta|<4.0$~\footnote{Details of the jet reconstruction were given in
519: Section~\ref{sec:subjet}.}. We calculate the missing transverse energy
520: from the vector sum of the momenta of all the visible particles within
521: $|\eta|<5.0$, excluding neutrinos and neutralinos. For our
522: investigation of Higgs decays into $b\bar b$ pairs we use a simple
523: statistical model for identifying two collimated $b$-jets, by tagging
524: any jet matched to the direction of two $b$ quarks with $p_T > 15$~GeV
525: at the parton level~\footnote{For a match we require a distance to the jet of
526: less than 0.4 units in the $(\eta, \phi)$ plane for each $b$ quark.} as
527: resulting from two collimated $b$-jets with an efficiency of $40$\%,
528: and using a mis-tagging rate of $1$\% for all other jets. The numbers
529: used are a naive estimate from single $b$-jet tagging at the LHC. More
530: exact predictions for the performance of the LHC detectors on such
531: jets will require the full simulation of a detector, which is beyond
532: the scope of this study.
533:
534: \subsection{Signal Isolation}
535: \label{sec:cuts}
536:
537: In order to isolate events with the decay chain (\ref{eq:chargino}),
538: we use the following cuts:
539: \begin{itemize}
540: \item Require missing energy $\not\!\!E_T>300$~GeV;
541: \item Require at least one $W^\pm$ candidate jet with
542: \begin{itemize}
543: \item transverse momentum $p_T>200$~GeV,
544: \item jet mass around the $W$ mass: $75<m_W<105$~GeV,
545: \item separation scale $1.5<\log{(p_T\sqrt{y})}<1.9$;
546: \end{itemize}
547: \item Veto events with a top candidate, i.e.\,a jet-$W$ combination with invariant
548: mass in the range $150-220$~GeV.
549: \item Require two additional jets with $p_T>200,150$~GeV;
550: \item Veto events containing leptons ($e$ or $\mu$) with
551: $p_T>10$~GeV.
552: \end{itemize}
553: The asymmetric cut on jet mass is due to the tendency of the jet
554: algorithm to overestimate the jet mass and energy by including
555: contributions from the underlying event and parton shower. Jets which
556: pass the jet mass cut for $W$ candidates are re-calibrated to the
557: known $W$ mass by rescaling the four-vector. To find the invariant
558: $qW^\pm$ mass in events that pass all the cuts we combine the $W^\pm$
559: candidate with any jet that passes the $p_T>200$~GeV requirement. This
560: creates some combinatorial background from signal events where we have
561: picked the wrong jet.
562:
563: The remaining non-SUSY background is mainly a mix between $t\bar t$
564: events and vector boson (single and pair) production in association
565: with multiple jets. We show the surviving number of events for the
566: given cuts and integrated luminosities in
567: Table~\ref{tab:efficiency}. While this background is relatively
568: unimportant for the benchmarks with larger cross sections, it is
569: highly significant for the $\delta$ benchmark. Thus we have imposed
570: one further cut for this benchmark:
571: \begin{itemize}
572: \item Require that the angle in the transverse plane between the
573: missing momentum and the $W$ candidate is larger than $\pi/8$.
574: \end{itemize}
575: The reason for this cut is that a significant fraction of the
576: surviving $t\bar t$ events feature the $W\to\tau\nu_\tau$ decay of a
577: highly boosted $W$, where the $\tau$ subsequently decays
578: hadronically. This gives large amounts of missing energy from
579: neutrinos, and the possibility of misidentifying the $\tau$ jet, or a
580: collimation of the $\tau$ jet and the $b$ jet from the same top as
581: the $W$ candidate. The result is a strong correlation between the
582: missing momentum and the direction of the $W$ candidate for the $t\bar
583: t$ background.
584:
585: For the other two decay chains, (\ref{eq:chi02h}) and
586: (\ref{eq:chi02Z}), we use the same signal extraction procedure,
587: replacing the cut values for the jet mass cut and the separation scale
588: cut with appropriate values. For $Z$ candidates we require
589: $90<m_Z<115$~GeV and $1.6<\log{(p_T\sqrt{y})}<2.0$, while for the
590: Higgs boson we require $110<m_h<140$~GeV and
591: $1.8<\log{(p_T\sqrt{y})}<2.1$. In the Higgs case, we further require
592: $b$-tagging as described in the previous Section, while we do not
593: implement the lepton veto. For our mass reconstruction in
594: Section~\ref{sec:mass} we also look at the relatively clean signal of
595: leptonic $Z$ decays. We keep the cuts on missing energy and additional
596: jets and require two opposite-sign, same-flavour leptons with
597: $p_T>10$~GeV and an invariant mass in the range $85-95$~GeV.
598:
599: \subsection{Invariant Mass Distributions}
600:
601: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
602: \begin{figure}
603: \begin{center}
604: \epsfxsize 13.5cm
605: \epsfbox{m_alpha.eps}
606: \caption{Invariant mass distributions for SUSY benchmark scenario
607: $\alpha$: combinations of jets and $W^\pm$ candidates (a) with and (b)
608: without the cut on the separation scale, (c) the $\ell q$ invariant
609: mass distribution resulting from $W^\pm \to \ell^\pm \nu$ decays,
610: combinations of jets and $Z^0$ bosons decaying (d) hadronically and
611: (e) leptonically, and (f) combinations of jets and $h$ bosons. Signal
612: - blue, solid lines; SUSY background - red, dashed lines; SM
613: background - green, dotted lines.}
614: \label{fig:m_alpha}
615: \end{center}
616: \end{figure}
617: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
618:
619: The resulting invariant $qW^\pm$ mass distribution for the $\alpha$
620: benchmark point is shown in Fig.~\ref{fig:m_alpha}a, for our
621: simulation of 30~fb$^{-1}$. There is a clear signal peak from events
622: that contain the decay chain (\ref{eq:chargino}), above a fairly
623: smooth SUSY background and a small SM background. Their sum yields the
624: black points, with the statistical errors also shown. The SUSY
625: background has a small peak under the signal peak. This is due to a
626: large number of events with misidentification of $Z$s from neutralino
627: decays as $W$s, which can be understood from the neutralino branching
628: ratio in Table~\ref{tab:BRs}. We see immediately the remnant of an
629: edge effect at $m_{qW}\sim 500$~GeV, as expected from the true
630: invariant mass distribution shown in Fig.~\ref{fig:imtruth}. However, the
631: lower edge of the trapezoidal distribution in Fig.~\ref{fig:imtruth}
632: at $m_{qW}\sim 300$~GeV is not visible, largely as a result of the
633: hard cuts used on jet momenta. For comparison, we also show in
634: Fig.~\ref{fig:m_alpha}b the corresponding distribution for the
635: $qW^\pm$ mass distribution if the cut on the separation scale is {\it
636: not} imposed: the edge structure is less significant, though still
637: clearly present, with the SUSY background being larger relative to the
638: signal.
639:
640: The $\ell q$ invariant mass distribution resulting from $W^\pm \to
641: \ell^\pm \nu$ decays is shown in Fig.~\ref{fig:m_alpha}c: here we have
642: required a lepton ($e$ or $\mu$) with $p_T>40$~GeV instead of the
643: $W$-jet cuts. Due to the smearing from the escaping neutrino, there is
644: no interesting edge structure in the signal, and the lack of any
645: efficient cuts other than the $p_T$ of the lepton means that the
646: background dominates. While there may be some information to be gained
647: from the peak positions of the distribution, the leptonic decay is
648: clearly more difficult to use than the hadronic decay utilising the
649: \kt\ jet algorithm.
650:
651: We also show the corresponding invariant mass distribution for $qZ^0$
652: combinations followed by $Z^0$ decays into hadrons in
653: Fig.~\ref{fig:m_alpha}d, and for leptonic $Z^0$ decays in
654: Fig.~\ref{fig:m_alpha}e. For the hadronic decay the small mass
655: difference between the $W$ and $Z$ means that the signal events (blue)
656: containing the decay chain (\ref{eq:chi02Z}) are swamped by a large
657: SUSY background consisting of events with a misidentified $W$, despite
658: the higher cut values for jet mass and separation scale.
659:
660: In the case of leptonic $Z$ decays the background can be reduced much
661: more effectively, by selecting same-flavour, opposite-sign lepton
662: pairs. As with the jets, this pair is rescaled to the known $Z$
663: mass. The resulting distribution exhibits the expected signal peak for
664: $qZ$ combinations (blue), smeared by combinatorial effects from
665: picking the incorrect jet, over a smaller SUSY background.
666:
667: Finally, in Fig.~\ref{fig:m_alpha}f we show the invariant $qh$ mass
668: distribution, where there is, as expected, no signal because of the
669: zero branching ratio of the $\alpha$ benchmark.
670:
671: The corresponding distributions for benchmark $\beta$ are shown in
672: Fig.~\ref{fig:m_beta}, again for a simulation of 30~fb$^{-1}$. We see
673: in panel (a) the expected distinctive edge structure (blue, solid)
674: rising above the SUSY background: the SM background is again small. As
675: in the case of benchmark $\alpha$, we see in panel (b) that the
676: signal-to-background ratio is worse if the cut on the separation scale
677: is {\it not} imposed. We also observe that the SUSY background has
678: {\it no} visible peak in the signal peak region, due to the small
679: branching ratio of the neutralino to $Z$ for the $\beta$
680: benchmark. Panel (c) of Fig.~\ref{fig:m_beta} shows that, as in the
681: case of benchmark $\alpha$, it would be very difficult to extract
682: information from a leptonic $W$-decay signal. Nor, according to panels
683: (d) and (e), does it seem possible in the case of benchmark $\beta$ to
684: extract a $Z$-decay signal, at least at the considered integrated
685: luminosity. This might have been anticipated because of the much smaller
686: $\tilde\chi_2^0\rightarrow\tilde\chi_1^0 Z$ branching ratio in this
687: case compared to the $\alpha$ benchmark.
688:
689: For the $qh$ distribution in panel (f) the situation is far better. We
690: see a clear edge in the distribution in the expected region of
691: $m_{qh}\sim 650$~GeV. The statistics are low, and are naturally
692: dependent on the $b$-tagging efficiency. In addition, the observation
693: of the edge relies on the use of the sub-jet separation scale
694: cut. Given the good signal-to-background ratio apparent after this
695: cut, we expect that a lower $b$-tagging efficiency could be
696: compensated by higher statistics.
697:
698: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
699: \begin{figure}
700: \begin{center}
701: \epsfxsize 14cm
702: \epsfbox{m_beta.eps}
703: \caption{Various invariant mass distributions obtained in a simulation
704: of events for SUSY benchmark scenario $\beta$. See
705: Fig.~\ref{fig:m_alpha} and text for details.
706: }
707: \label{fig:m_beta}
708: \end{center}
709: \end{figure}
710: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
711:
712: In the case of benchmark $\gamma$, shown in Fig.~\ref{fig:m_gamma},
713: also for a simulation of 30~fb$^{-1}$, we expect only one observable
714: distribution. We see again the familiar features of a strong hadronic
715: $W$-decay signal with the cut on the separation scale in panel (a) and
716: a weaker signal-to-background ratio without this cut in panel
717: (b). Again there is a peak in the SUSY background under the signal
718: peak. For $\gamma$ the on-shell decay
719: $\tilde\chi_2^0\rightarrow\tilde\chi_1^0 Z$ is not allowed
720: kinematically, but proceeds off-shell and the decay products are again
721: misidentified as $W$s. As expected, there are no detectable signals in
722: leptonic $W$ decay, in hadronic or leptonic $Z$ decay, or in hadronic
723: $h$ decay.
724:
725: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
726: \begin{figure}
727: \begin{center}
728: \epsfxsize 14cm
729: \epsfbox{m_gamma.eps}
730: \caption{Various invariant mass distributions obtained in a simulation
731: of events for SUSY benchmark scenario $\gamma$. See
732: Fig.~\ref{fig:m_alpha} and text for details.
733: }
734: \label{fig:m_gamma}
735: \end{center}
736: \end{figure}
737: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
738:
739: We turn finally to the case of benchmark $\delta$, shown in
740: Fig.~\ref{fig:m_delta}. We recall that in this case the simulation
741: corresponds to an integrated luminosity of 300~fb$^{-1}$, in view of
742: the higher masses of the sparticles and hence the lower cross
743: sections. We see in panel (a) that the hadronic $W$-decay signal is
744: less clear in this case, and that for the first time the SM background
745: dominates over that due to SUSY. There are virtually no signal events
746: in panels (c, d) and (e), corresponding to leptonic $W$ decays,
747: hadronic and leptonic $Z$ decays, respectively. However, there is a
748: possible $h$ signal in panel (f). The limited amount of generated
749: events for the SM backgrounds results in relatively large weights for
750: the background at this integrated luminosity, obfuscating the edge
751: structure. We again emphasise the necessity of the sub-jet scale cut
752: and the dependence on the $b$-tagging efficiency assumed.
753:
754: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
755: \begin{figure}
756: \begin{center}
757: \epsfxsize 14cm
758: \epsfbox{m_delta.eps}
759: \caption{Various invariant mass distributions obtained in a simulation
760: of events for SUSY benchmark scenario $\delta$. See
761: Fig.~\ref{fig:m_alpha} and text for details.
762: }
763: \label{fig:m_delta}
764: \end{center}
765: \end{figure}
766: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
767:
768: \subsection{Sideband Subtraction}
769:
770: In order to be able to measure the positions of the expected edges of
771: the invariant mass distributions for signal events, and hence
772: constrain the sparticle masses, we would like to subtract the SM and
773: SUSY backgrounds, without model-dependent assumptions on their
774: shape. We do this by performing a sideband subtraction, where we
775: imitate the background that does not feature a correctly identified
776: boson by collecting a sample of events from the generated ``data'',
777: that features boson candidates with masses away from the resonance
778: peak of the boson mass in question. Using events in two bands
779: (region~II and III) on either side of the signal isolation interval
780: (region~I) for the jet mass distribution, each with half the width of
781: the signal band, we recalibrate the boson mass to the nominal peak
782: value as described above, and perform most other cuts as for the
783: signal. The exception is the sub-jet separation scale cut, which is
784: highly correlated with the jet mass cut, and is thus ignored for the
785: sideband sample. We show the jet mass distribution and the signal and
786: sideband regions for $W$ candidates at the $\alpha$ benchmark point in
787: Fig.~\ref{fig:sideband} (left). Only the missing-energy cut given in
788: Section~\ref{sec:cuts} has been applied to the events.
789:
790: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
791: \begin{figure}
792: \begin{center}
793: \epsfxsize 15cm
794: \epsfbox{sideband.eps}
795: \caption{Left: Jet mass distribution for $W$ candidates at the $\alpha$
796: benchmark point following a cut on missing energy:
797: $\not\!\!E_T>300$~GeV. The SUSY signal (blue, solid), SUSY background
798: (red, dashed) and SM background (green, dotted) contributions are also
799: shown separately. Right: Invariant mass distribution of $qW$
800: combinations for the $\alpha$ benchmark point in the signal region
801: (region I, points with error bars), in the sideband regions (region II
802: dashed, region III dotted) and for the sum of sideband events (dashed
803: dotted). Also shown is the fit to the sideband distribution (solid
804: line), rescaled to the signal distribution.}
805: \label{fig:sideband}
806: \end{center}
807: \end{figure}
808: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
809:
810: The two resulting distributions are added and fitted with a
811: third-degree polynomial, giving the shape of the background. This is
812: shown, again using the $\alpha$ benchmark point as an example, in
813: Fig.~\ref{fig:sideband} (right). The background is rescaled to the
814: full distribution from the signal band, shown with error bars in
815: Figs.~\ref{fig:m_alpha}-\ref{fig:m_delta}, using bins at higher
816: invariant masses than the observed edges. The rescaled background is
817: then subtracted from the full distributions, the results of which are
818: shown for the $qW$ invariant mass distributions in
819: Fig.~\ref{fig:m}. While this procedure primarily models background
820: from fake $W$s, we find that it does a good job of describing all the
821: background near the upper edges, perhaps with the exception of the
822: $\delta$ benchmark point which is dominated by SM events with real
823: $W$s. For the other three benchmarks, with much less SM background,
824: the jet rescaling to the $W$ mass for the sideband samples gives a
825: distribution that is similar enough to model well also this background.
826:
827: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
828: \begin{figure}
829: \begin{center}
830: \epsfxsize 15cm
831: \epsfbox{m.eps}
832: \caption{Invariant mass distribution of $qW$ combinations at the four
833: benchmark points, $\alpha$ (top left), $\beta$ (top right), $\gamma$
834: (bottom left) and $\delta$ (bottom right), after sideband
835: subtraction. Also shown are fits to the upper edges of the
836: distributions, and for $\gamma$ also the clear lower edge. }
837: \label{fig:m}
838: \end{center}
839: \end{figure}
840: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
841:
842: We have also considered estimates for the combinatorial background
843: that result from combining correctly identified bosons with the wrong
844: jet, which can be seen as tails in the blue signal distributions in
845: Figs.~\ref{fig:m_alpha}-\ref{fig:m_delta}. By combining accepted boson
846: and squark decay jet candidates randomly from all events we collect a
847: large mixed sample that should be representative of this part of the
848: background. However, we find that combining these two descriptions of
849: the background is difficult due to the limited statistics, and joint
850: fits to the events in the high invariant-mass region tend to favour
851: heavily the sideband sample. For the $\delta$ benchmark point the
852: sideband and mixed samples perform similarly in describing the
853: background in the vicinity of the edge, but the sideband is slightly
854: better at the high invariant masses used to set the scale of the
855: background distribution. As a consequence, the mixed sample is not
856: included in the fits shown in this Section, but investigations into
857: combining different descriptions of the background is certainly worthy
858: of further effort, in particular when data is available from the
859: experiments.
860:
861: After the sideband subtraction, the upper endpoints of the $qW$
862: distributions are clearly visible for all four benchmarks in
863: Fig.~\ref{fig:m}. To estimate their positions, we have performed fits
864: with a linear form for the signal, emulating the distributions of
865: Fig.~\ref{fig:imtruth}, with a free parameter for the cut-off at the
866: endpoint. These distributions are further smeared by a Gaussian to
867: model the limited jet energy resolution, using a smearing width
868: determined by the fit. The resulting values for the endpoints
869: $m_{qW}^{\rm max}$ can be found in Table~\ref{tab:qWmax}. We also show
870: in Fig.~\ref{fig:m_nosep} the $qW$ invariant mass distributions obtained
871: when we omit the cut on separation scale for the $W$ candidate, and
872: the corresponding fit values are also reported in
873: Table~\ref{tab:qWmax}.
874:
875: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
876: \begin{table}\begin{center}
877: \begin{tabular}{lcccc} \hline
878: Fit / Benchmark & $\alpha$ & $\beta$ & $\gamma$ & $\delta$ \\ \hline
879: Scale cut & $527.4\pm 8.7$ & $650.2\pm 6.5$ & $482.2\pm 4.0$ & $1219.0\pm 85.1$ \\
880: No scale cut & $532.7\pm 3.8$ & $651.5\pm 5.4$ & $481.7\pm 4.1$ & $1203.9\pm 34.5$ \\
881: Nominal & $519.9$ & $653.8$ & $468.6$ & $1272.1$ \\ \hline
882: \end{tabular}
883: \caption{Fitted endpoint values $m_{qW}^{\rm max}$ of $qW$ distributions,
884: measured in GeV, compared with the nominal values for the
885: corresponding benchmarks.}
886: \label{tab:qWmax}
887: \end{center}
888: \end{table}
889: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
890:
891: %mqWmin alpha: 306.522
892: %mqZmax alpha: 505.641
893: %mqZmin alpha: 358.504
894:
895: %mqWmin beta: 276.188
896: %mqHmax beta: 628.276
897: %mqHmin beta: 422.796
898: %mqZmax beta: 648.875
899: %mqZmin beta: 316.523
900:
901: %mqWmin gamma: 368.978
902:
903: %mqWmin delta: 246.799
904: %mqHmax delta: 1265.68
905: %mqHmin delta: 377.057
906: %mqZmax delta: 1269.64
907: %mqZmin delta: 280.802
908:
909: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
910: \begin{figure}
911: \begin{center}
912: \epsfxsize 15cm
913: \epsfbox{m_nosep.eps}
914: \caption{As Fig.~\ref{fig:m}, but omitting the cut on the sub-jet
915: separation scale.}
916: \label{fig:m_nosep}
917: \end{center}
918: \end{figure}
919: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
920:
921: We see that the endpoint estimates have statistical errors which are
922: in the ${\cal O}(1)$\% region for benchmarks $\alpha-\gamma$ and
923: slightly below ${\cal O}(10)$\% for $\delta$. The statistical errors
924: are in general smaller without the cut on the sub-jet separation
925: scale, particularly in the case of benchmark $\delta$, due to somewhat
926: larger statistics.
927:
928: In the case of the $\alpha$ benchmark point, the fitted values are
929: both fairly close to the nominal value, but with indications of larger
930: systematic errors when not using the scale cut. Both fits overestimate
931: the endpoint, which could be expected as the reconstructed jets from
932: squark decays have not been calibrated for the algorithm's tendency to
933: overestimate the jet. For $\beta$ both fits are very close to the
934: nominal value, while for $\gamma$ the fits again overestimate the edge
935: position. From the three benchmarks considered here this systematic
936: error seems to increase for lower values of the endpoint, barring
937: other more important hidden systematics. In the case of the $\delta$
938: benchmark both fits underestimate the endpoint, but here the
939: statistical errors are large. Considered as a whole, the results of
940: Table~\ref{tab:qWmax} indicate that the systematical errors on the
941: endpoints in our fitting procedure are comparable to the statistical
942: errors.
943:
944: For most of the distributions it is impossible to estimate the lower
945: endpoint, since any structure at low invariant masses is washed away
946: by the hard kinematical cuts on jet energies used to isolate the
947: signal. The exception is the relatively high lying lower endpoint of
948: the $qW$ distribution for the $\gamma$ benchmark, where a similar
949: fitting technique to the one used for the upper endpoint yields a
950: lower endpoint estimate of $m_{qW}^{\rm min}=363.0\pm 3.0$~GeV before
951: and $m_{qW}^{\rm min}=347.0\pm 2.9$~GeV after the sub-jet cut is
952: applied, compared to the nominal value of $369.0$~GeV. The difference
953: in endpoint estimate originates from a worse fit of the sideband
954: distribution at high invariant masses when the sub-jet cut is used.
955:
956: We have in addition estimated both the lower and upper edge of the
957: $qZ$ distribution for the $\alpha$ benchmark point, where we
958: reconstruct the $Z$ from a lepton pair. After a similar sideband
959: subtraction we estimate $m_{qZ}^{\rm max}=523.7\pm 10.6$~GeV and
960: $m_{qZ}^{\rm min}=324.5\pm 9.2$~GeV, to be compared to the nominal
961: values of $505.6$~GeV and $358.5$~GeV, respectively.
962:
963: For the $qh$ distributions, fitting is difficult with the low
964: statistics available both at the $\beta$ and $\delta$ benchmarks. From
965: sideband subtracted distributions we make estimates by visual
966: inspection, giving values of $m_{qh}^{\rm max}=540\pm 40$~GeV for
967: $\beta$ and $m_{qh}^{\rm max}=1450\pm 100$~GeV for $\delta$, compared
968: to the nominal values of $628.3$~GeV and $1265.7$~GeV, respectively. As
969: has been noted earlier, the feasibility of measuring these edges
970: depends on the $b$-tagging achievable, and on the use of the sub-jet
971: separation scale cut.
972:
973: \subsection{Mass Spectra}
974: \label{sec:mass}
975:
976: The positions of the edges of the invariant mass distributions may be
977: used to extract information on the spectrum of the SUSY particles
978: involved. Whilst the four equations giving the edges of the $qW$ and
979: $qZ/qh$ distributions may in principle be solved to obtain the four
980: masses involved, extracting accurate values of the absolute masses -
981: as opposed to mass differences - will be difficult in practise because
982: of the degeneracies between the masses of the bosons and of the
983: gauginos, respectively, and thus in the upper edges, as shown in
984: Fig.~\ref{fig:imtruth}. However, for the four benchmarks considered
985: here this is an academic problem, as neither of the benchmarks have four
986: measurable edges. Given both the lower and upper edge of a single
987: distribution the squark mass can be eliminated from
988: Eq.~\ref{eq:endqW}, and one can solve for the mass of the chargino or
989: the next-to-lightest neutralino in terms of the mass of the lightest
990: neutralino. This can in turn be used to arrive at the squark mass. We
991: show in Fig.~\ref{fig:masses_alpha} the squark and $\tilde\chi_2^0$
992: masses determined using the upper and lower $qZ$ edge measured for the
993: $\alpha$ benchmark point, as functions of the undetermined
994: $\tilde\chi_1^0$ mass. We also show the $1\sigma$ error bands on the
995: masses resulting from the statistical uncertainty of the edge
996: measurements, assuming that the errors on the two edges are
997: independent.
998:
999: With the squark mass known in terms of the $\tilde\chi_1^0$ mass, the
1000: upper edge of the $qW$ distribution can also be used to give the
1001: $\tilde\chi_1^\pm$ mass. Because of the quadratic nature of
1002: Eq.~\ref{eq:endqW} two solutions result, none of which can be rejected
1003: out of hand~\footnote{There are also in principle two solutions for
1004: the $\tilde\chi_2^0$ mass, but one can be rejected as unphysical since
1005: it is always less than the $\tilde\chi_1^0$ mass.}. We show both
1006: solutions in Fig.~\ref{fig:masses_alpha}. For values of the
1007: $\tilde\chi_1^0$ mass below $\sim 50$~GeV there are no solutions.
1008:
1009: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
1010: \begin{figure}
1011: \begin{center}
1012: \epsfxsize 15cm
1013: \epsfbox{masses_alpha.eps}
1014: \caption{The squark (red), $\tilde\chi_1^\pm$ (blue) and
1015: $\tilde\chi_2^0$ (black) masses as a function of the
1016: $\tilde\chi_1^0$ mass for the $\alpha$ benchmark point. The dashed
1017: lines show the $1\sigma$ statistical error bands. The vertical green
1018: line indicates the nominal $\tilde\chi_1^0$ mass.}
1019: \label{fig:masses_alpha}
1020: \end{center}
1021: \end{figure}
1022: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
1023:
1024: For the $\gamma$ benchmark point the same procedure can be followed
1025: for the two endpoints of the $qW$ distribution, and we show the
1026: resulting squark and chargino masses as functions of the lightest
1027: neutralino mass in Fig.~\ref{fig:masses_gamma}. For the remaining two
1028: benchmarks, extracting masses is more difficult since only one edge is
1029: well measured. However, the LHC experiments have the potential for
1030: measuring other SUSY mass-dependent quantities, such as the effective
1031: mass~\cite{Hinchliffe:1996iu,Tovey:2000wk}, whose determination would
1032: give complementary relations between the involved masses.
1033:
1034: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
1035: \begin{figure}
1036: \begin{center}
1037: \epsfxsize 15cm
1038: \epsfbox{masses_gamma.eps}
1039: \caption{The squark (red) and $\tilde\chi_1^\pm$ (blue) and
1040: masses as a function of the $\tilde\chi_1^0$ mass for the $\gamma$
1041: benchmark point. The dashed lines show the $1\sigma$ statistical error
1042: bands. The vertical green line indicates the nominal $\tilde\chi_1^0$
1043: mass.}
1044: \label{fig:masses_gamma}
1045: \end{center}
1046: \end{figure}
1047: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
1048:
1049: Given an accurate measurement of the $\tilde\chi_1^0$ mass, for
1050: example at a future linear collider, the heavier neutralino, chargino
1051: and squark masses could then be found with statistical errors in the
1052: range of $1-5$\%, using LHC data, even if the particles themselves are
1053: too heavy to be produced at the linear collider.
1054:
1055: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
1056: \section{Results and Conclusions}
1057: \label{sec:results}
1058: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
1059:
1060: We have shown in this paper, we believe for the first time, that it is
1061: possible to extract a SUSY signal solely from the hadronic decays of
1062: $W^\pm$ bosons produced in cascade decays involving charginos, and
1063: that the measurement of this signal may provide useful information
1064: about the sparticle mass spectrum. In each of the benchmarks studied,
1065: the upper edge of the $qW$ mass distribution in the decay chain
1066: $\tilde q_L\rightarrow\tilde\chi_1^\pm q \rightarrow\tilde\chi_1^0
1067: W^\pm q$ can be measured, with a statistical error of ${\cal O}(1)$\%
1068: for the relatively light sparticle masses of benchmarks
1069: $\alpha-\gamma$ and ${\cal O}(10)$\% for the heavier $\delta$, with
1070: squark and gluino masses above $1.5$~TeV. Systematic errors in the
1071: background and fitting are estimated to be of the same order as the
1072: statistical errors. In the case of benchmark $\gamma$, we are also
1073: able to extract the position of the lower edge of the $qW$ mass
1074: distribution, and in the case of point $\alpha$ we extract both the
1075: upper and the lower edges of the $qZ$ mass distribution, using the
1076: leptonic decays of $Z$ bosons. There are also clear indications in
1077: benchmarks $\beta$ and $\delta$ for the possibility of observing the
1078: upper edge of the $qh$ mass distribution using $h \to {\bar b} b$
1079: decays with collimated $b$-jets, but conclusions on such a signal
1080: require further analysis and better understanding of the $b$-tagging
1081: efficiency for these jets in a full detector simulation.
1082:
1083: The hadronic signals were extracted using the \kt\ algorithm for jet
1084: reconstruction, including the single-jet mass and the proposal
1085: of~\cite{Butterworth:2002tt} to improve the identification of hadronic
1086: decays of heavy bosons via a cut on the sub-jet separation scale. The
1087: \kt\ algorithm is shown to be well suited for reconstructing edge
1088: features in the considered invariant mass distributions. The sub-jet
1089: cut procedure improves quantitatively the signal-to-background ratio,
1090: and while the loss of statistics increases the statistical errors,
1091: fits to the edges of the $qW$ distribution show improved results with
1092: respect to the nominal values when using the sub-jet cut. For the $qh$
1093: distributions with their small signal-to-background ratio, the sub-jet
1094: cut is crucial in reducing the background to a manageable level.
1095:
1096: A detailed exploration of the capabilities of the LHC experiments to
1097: measure and use all the information that could be gained from these
1098: benchmark points, and hence a more detailed display of the value added
1099: to the global fits by the measurements described here, lies beyond the
1100: scope of this study. However, we do note that this technique provides
1101: novel information on chargino and neutralino spectroscopy, e.g., the
1102: difference between the $\tilde
1103: \chi_1^0$ and $\tilde\chi_1^\pm$ mases, and in at least one case the
1104: $\tilde\chi_2^0 - \tilde\chi_1^\pm$ mass difference. These pieces of
1105: information are useful for potentially constraining SUSY models and
1106: perhaps foreseeing the locations of interesting thresholds in $e^+
1107: e^-$ annihilation.
1108:
1109: As pointed out in~\cite{Butterworth:2002tt}, similar techniques for
1110: analysing hadronic final states arising from the decays of heavy
1111: particles should be useful in other situations. Examples include the
1112: analysis of top physics, the search for $R$-violating hadronic decays
1113: of sparticles, or in the isolation of a generic SUSY signal from QCD
1114: SM backgrounds. Indeed, some recent studies have highlighted the
1115: potential of single-jet mass cuts in exotic searches~\cite{studies},
1116: and we note that the $y$-scale cut should also be useful in these
1117: cases. We believe this to be an area meriting much further
1118: experimental and phenomenological study.
1119:
1120: \section*{Acknowledgements}
1121: The authors thank Jeppe Anderson for useful discussions. ARR
1122: acknowledges support from the European Community through a Marie Curie
1123: Fellowship for Early Stage Researchers Training, and from the
1124: Norwegian Research Council.
1125:
1126: \begin{thebibliography}{10}
1127:
1128: %\cite{Hinchliffe:1996iu}
1129: \bibitem{Hinchliffe:1996iu}
1130: I.~Hinchliffe, F.~E.~Paige, M.~D.~Shapiro, J.~Soderqvist and W.~Yao,
1131: %``Precision SUSY measurements at LHC,''
1132: Phys.\ Rev.\ D {\bf 55} (1997) 5520
1133: [arXiv:hep-ph/9610544].
1134: %%CITATION = HEP-PH 9610544;%%
1135:
1136: %\cite{edges}
1137: \bibitem{edges}
1138: H.~Bachacou, I.~Hinchliffe and F.~E.~Paige,
1139: %``Measurements of masses in SUGRA models at LHC,''
1140: Phys.\ Rev.\ D {\bf 62} (2000) 015009
1141: [arXiv:hep-ph/9907518];\\
1142: %%CITATION = HEP-PH 9907518;%%
1143: B.~C.~Allanach, C.~G.~Lester, M.~A.~Parker and B.~R.~Webber,
1144: %``Measuring sparticle masses in non-universal string inspired models
1145: %at the LHC,''
1146: JHEP {\bf 0009} (2000) 004
1147: [arXiv:hep-ph/0007009];\\
1148: %%CITATION = HEP-PH 0007009;%%
1149: C.~G.~Lester,
1150: %``Model independent sparticle mass measurements at ATLAS,''
1151: CERN-THESIS-2004-003;\\
1152: %\href{http://www.slac.stanford.edu/spires/find/hep/www?r=cern-thesis-2004-003}{SPIRES entry}
1153: B.~K.~Gjelsten, D.~J.~Miller and P.~Osland,
1154: %``Measurement of SUSY Masses via Cascade Decays for SPS~1a''
1155: JHEP {\bf 12} (2004) 003
1156: [arXiv:hep-ph/0410303];\\
1157: %%CITATION = HEP-PH 0410303;%%
1158: D.~J.~Miller, P.~Osland and A.~R.~Raklev,
1159: %``Invariant mass distributions in cascade decays,''
1160: JHEP {\bf 03} (2006) 034
1161: [arXiv:hep-ph/0510356].
1162: %%CITATION = HEP-PH 0510356;%%
1163:
1164: %\cite{Nojiri:2003tv}
1165: \bibitem{Nojiri:2003tv}
1166: M.~M.~Nojiri, G.~Polesello and D.~R.~Tovey,
1167: %``Measuring the mass of the lightest chargino at the CERN LHC,''
1168: arXiv:hep-ph/0312318.
1169: %%CITATION = HEP-PH 0312318;%%
1170:
1171: %\cite{DeRoeck:2005bw}
1172: \bibitem{DeRoeck:2005bw}
1173: A.~De Roeck, J.~R.~Ellis, F.~Gianotti, F.~Moortgat, K.~A.~Olive and L.~Pape,
1174: % ``Supersymmetric benchmarks with non-universal scalar masses or
1175: % gravitino dark matter,''
1176: [arXiv:hep-ph/0508198].
1177: %%CITATION = HEP-PH 0508198;%%
1178:
1179: %\cite{Butterworth:2002tt}
1180: \bibitem{Butterworth:2002tt}
1181: J.~M.~Butterworth, B.~E.~Cox and J.~R.~Forshaw,
1182: %``W W scattering at the LHC,''
1183: Phys.\ Rev.\ D {\bf 65} (2002) 096014
1184: [arXiv:hep-ph/0201098].
1185: %%CITATION = HEP-PH 0201098;%%
1186:
1187: %\cite{Catani:1993hr}
1188: \bibitem{Catani:1993hr}
1189: S.~Catani, Y.~L.~Dokshitzer, M.~H.~Seymour and B.~R.~Webber,
1190: % ``Longitudinally Invariant K(T) Clustering Algorithms For Hadron
1191: % Hadron Collisions,''
1192: Nucl.\ Phys.\ B {\bf 406} (1993) 187.
1193: %%CITATION = NUPHA,B406,187;%%
1194:
1195: \bibitem{fastjet}
1196: M.~Cacciari and G.~P.~Salam,
1197: %``Dispelling the N**3 myth for the k(t) jet-finder,''
1198: Phys.\ Lett.\ B {\bf 641} (2006) 57
1199: [arXiv:hep-ph/0512210].
1200: %%CITATION = HEP-PH 0512210;%%
1201:
1202: \bibitem{DGLAP}
1203: V.~N.~Gribov and L.~N.~Lipatov,
1204: %``Deep Inelastic E P Scattering In Perturbation Theory,''
1205: Sov.\ J.\ Nucl.\ Phys.\ {\bf 15} (1972) 438
1206: [Yad.\ Fiz.\ {\bf 15} (1972) 781].\\
1207: %%CITATION = SJNCA,15,438;%%
1208: G.~Altarelli and G.~Parisi,
1209: %``Asymptotic Freedom In Parton Language,''
1210: Nucl.\ Phys.\ B {\bf 126} (1977) 298. \\
1211: %%CITATION = NUPHA,B126,298;%%
1212: Y.~L.~Dokshitzer,
1213: % ``Calculation Of The Structure Functions For Deep Inelastic
1214: % Scattering And E+E- Annihilation By Perturbation Theory In Quantum
1215: % Chromodynamics. (In Russian),''
1216: Sov.\ Phys.\ JETP {\bf 46} (1977) 641
1217: [Zh.\ Eksp.\ Teor.\ Fiz.\ {\bf 73} (1977) 1216].
1218: %%CITATION = SPHJA,46,641;%%
1219:
1220: \bibitem{pythia}
1221: T.~Sjostrand, S.~Mrenna and P.~Skands,
1222: %``PYTHIA 6.4 physics and manual,''
1223: JHEP {\bf 0605}, 026 (2006) [arXiv:hep-ph/0603175];
1224: %%CITATION = HEP-PH 0603175;%%
1225: Note that the events were generated calling {\tt PYEVNT},
1226: which uses the same underlying event and
1227: parton shower models as the pre-6.3 versions of PYTHIA.
1228:
1229: \bibitem{alpgen}
1230: M.~L.~Mangano et al,
1231: JHEP {\bf 0307} (2003) 001
1232: [arXiv:hep-ph/0206293];
1233: %%CITATION = JHEPA,0307,001;%%
1234: M.~L.~Mangano, M.~Moretti and R.~Pittau,
1235: Nucl.\ Phys.\ B {\bf 632} (2002) 343
1236: [arXiv:hep-ph/0108069];
1237: %%CITATION = NUPHA,B632,343;%%
1238: F.~Caravaglios et al,
1239: Nucl.\ Phys.\ B {\bf 539} (1999) 215
1240: [arXiv:hep-ph/9807570].
1241: %%CITATION = NUPHA,B539,215;%%
1242:
1243:
1244: \bibitem{herwig} G.~Corcella et al, JHEP 0101 (2001) 010
1245: [arXiv:hep-ph/0011363]; [arXiv:hep-ph/0201201].
1246:
1247: \bibitem{jimmy}
1248: J.~M.~Butterworth, J.~R.~Forshaw and M.~H.~Seymour,
1249: Z.\ Phys.\ C {\bf 72} (1996) 637 [arXiv:hep-ph/9601371].
1250: %%CITATION = HEP-PH 9601371;%%
1251:
1252: \bibitem{tunes}
1253: C.~Buttar {\it et al.} in
1254: {\it HERA and the LHC - A workshop on the implications of HERA for LHC
1255: physics:
1256: Proceedings Part A}, pp.192-217
1257: [arXiv:hep-ph/0601012].
1258: %%CITATION = HEP-PH 0601012;%%
1259:
1260: \bibitem{tevjets}
1261: V.~M.~Abazov {\it et al.} [D0 Collaboration],
1262: Phys.\ Rev.\ D {\bf 65} (2002) 052008
1263: [arXiv:hep-ex/0108054];\\
1264: %%CITATION = HEP-EX 0108054;%%ITATION = HEP-EX 0505013;%%
1265: D.~Acosta {\it et al.} [CDF Collaboration], [arXiv:hep-ex/0505013].
1266:
1267: \bibitem{herajets}
1268: S.~Chekanov {\it et al.} [ZEUS Collaboration],
1269: Nucl.\ Phys.\ B {\bf 700} (2004) 3
1270: [arXiv:hep-ex/0405065].
1271: %%CITATION = HEP-EX 0405065;%%
1272:
1273: \bibitem{lepjets}
1274: G.~Abbiendi {\it et al.} [OPAL Collaboration],
1275: Eur.\ Phys.\ J.\ C {\bf 37} (2004) 25
1276: [arXiv:hep-ex/0404026];
1277: %%CITATION = HEP-EX 0404026;%%
1278: Eur.\ Phys.\ J.\ C {\bf 31} (2003) 307
1279: [arXiv:hep-ex/0301013];\\
1280: %%CITATION = HEP-EX 0301013;%%
1281: D.~Buskulic {\it et al.} [ALEPH Collaboration],
1282: Phys.\ Lett.\ B {\bf 384} (1996) 353.
1283: %%CITATION = PHLTA,B384,353;%%
1284:
1285: \bibitem{theses}
1286: S.~Allwood, Manchester University PhD thesis 2005.\\
1287: E.~Stefanidis, University College London PhD thesis 2006.
1288:
1289: %\cite{Bennett:2003bz}
1290: \bibitem{Bennett:2003bz}
1291: C.~L.~Bennett {\it et al.},
1292: %``First Year Wilkinson Microwave Anisotropy Probe (WMAP) Observations:
1293: %Preliminary Maps and Basic Results,''
1294: Astrophys.\ J.\ Suppl.\ {\bf 148} (2003) 1
1295: [arXiv:astro-ph/0302207].
1296: %%CITATION = ASTRO-PH 0302207;%%
1297:
1298: %\cite{Spergel:2003cb}
1299: \bibitem{Spergel:2003cb}
1300: D.~N.~Spergel {\it et al.},
1301: %``First Year Wilkinson Microwave Anisotropy Probe (WMAP) Observations:
1302: %Determination of Cosmological Parameters,''
1303: Astrophys.\ J.\ Suppl.\ {\bf 148} (2003) 175
1304: [arXiv:astro-ph/0302209].
1305: %%CITATION = ASTRO-PH 0302209;%%
1306:
1307: %\cite{Muhlleitner:2003vg}
1308: \bibitem{Muhlleitner:2003vg}
1309: M.~Muhlleitner, A.~Djouadi and Y.~Mambrini,
1310: %``SDECAY: A Fortran code for the decays of the supersymmetric particles
1311: in
1312: %the MSSM,''
1313: Comput.\ Phys.\ Commun.\ {\bf 168} (2005) 46
1314: [arXiv:hep-ph/0311167].
1315: %%CITATION = HEP-PH 0311167;%%
1316:
1317: %\cite{Lai:1999wy}
1318: \bibitem{Lai:1999wy}
1319: H.~L.~Lai {\it et al.} [CTEQ Collaboration],
1320: %``Global {QCD} analysis of parton structure of the nucleon: CTEQ5
1321: %distributions,''
1322: Eur.\ Phys.\ J.\ C {\bf 12} (2000) 375
1323: [arXiv:hep-ph/9903282].
1324: %%CITATION = HEP-PH 9903282;%%
1325:
1326: \bibitem{hztool}
1327: See {\it HERA and the LHC - A workshop on the implications of HERA for LHC
1328: physics: Proceedings Part B}, pp.617-621. [arXiv:hep-ph/0601013].\\
1329: %%CITATION = HEP-PH 0601013;%%
1330: http://projects.hepforge.org/hztool.
1331: %%CITATION = HEP-PH 0605034;%%
1332:
1333: %\cite{Tovey:2000wk}
1334: \bibitem{Tovey:2000wk}
1335: D.~R.~Tovey,
1336: %``Measuring the SUSY mass scale at the LHC,''
1337: Phys.\ Lett.\ B {\bf 498} (2001) 1
1338: [arXiv:hep-ph/0006276].
1339: %%CITATION = HEP-PH 0006276;%%
1340:
1341: \bibitem{studies}
1342: A.~L.~Fitzpatrick et al,
1343: arXiv:hep-ph/0701150;\\
1344: %%CITATION = HEP-PH/0701150;%%
1345: B.~Lillie, L.~Randall and L.~T.~Wang,
1346: %``The Bulk RS KK-gluon at the LHC,''
1347: arXiv:hep-ph/0701166;\\
1348: %%CITATION = HEP-PH/0701166;%%
1349: W.~Skiba and D.~Tucker-Smith,
1350: %``Using jet mass to discover vector quarks at the LHC,''
1351: arXiv:hep-ph/0701247;\\
1352: %%CITATION = HEP-PH/0701247;%%
1353: B.~Holdom,
1354: %``t' at the LHC: The physics of discovery,''
1355: arXiv:hep-ph/0702037.
1356: %%CITATION = HEP-PH/0702037;%%
1357:
1358:
1359: \end{thebibliography}
1360:
1361: \end{document}
1362: