1: \section{Event Reconstruction and Analysis}
2: \label{sec:analysis}
3:
4: \subsection{Track Reconstruction}
5: The first step of the event reconstruction is to search for
6: two-dimensional tracks in each view of SciBar using a cellular
7: automaton algorithm \cite{Maesaka:2005aj}. For tracking, the hit
8: threshold is set to two photoelectrons, corresponding to approximately
9: 0.2~MeV. Three dimensional tracks are reconstructed by matching
10: the timing and $z$-edges of the two dimensional tracks. The timing difference
11: between two two dimensional tracks is required to be less than 50 nsec, and the
12: $z$-edge difference must be less than 6.6~cm for upstream and
13: downstream edges. Reconstructed tracks are required to have at least
14: three-layer penetration, and therefore the minimum length of a
15: reconstructed track is 8~cm in the beam direction. According to the MC
16: simulation, 96\% of charged current interactions in SciBar are reconstructed to
17: have at least one track.
18:
19: To identify charged current events, we look for events in which
20: at least one reconstructed track in SciBar is matched with a track
21: or hits in the MRD. Such a track is defined as a SciBar-MRD matched
22: track. The most energetic SciBar-MRD matched track in any event is
23: considered as a muon candidate.
24: For matching a MRD track to a SciBar track, the upstream edge
25: of the MRD track is required to be on either one of the first two
26: layers of the MRD.
27: %The transverse distance between the extrapolated
28: %points at the first layer of the MRD from the SciBar and MRD tracks
29: %must be less than 30~cm.
30: The transverse distance between the two tracks at the first layer of
31: the MRD must be less than 30~cm.
32: The requirement on the difference between
33: track angles with respect to the beam direction is given by
34: $|\theta_{\rm MRD}~-~\theta_{\rm SB}|~<~\theta_{\rm max}$, where
35: $\theta_{\rm max}$ is a function of the length of the MRD track,
36: %the number of steel plates penetrated by the MRD track,
37: varying between 0.4~radian and
38: 1.1~radians. For track reconstruction in the MRD, at least two hit
39: layers in each view are needed, and thus this matching method is used
40: for tracks which penetrate at least three steel plates. If no MRD
41: track is found, we extrapolate the SciBar track to the MRD and search
42: for nearby contiguous hits in the MRD identifying a short muon track.
43: For matching MRD hits to a SciBar track, the MRD hit is required to be
44: within a cone with an aperture of $\pm$0.5~radian and a transverse
45: offset within 10~cm of the extrapolated SciBar track at the upstream
46: edge of the MRD. The timing difference between the SciBar track and
47: the track or hits in the MRD is required to be within 100 nsec. The
48: matching criteria impose a muon momentum threshold of 350~MeV/$c$.
49:
50: \subsection{Particle Identification}
51: The SciBar detector has the capability to distinguish protons from
52: muons and pions using $dE/dx$.
53: The particle identification variable, Muon Confidence Level (MuCL) is
54: calculated as follows. The confidence level at each plane is first
55: defined as the fraction of events in the expected $dE/dx$ distribution
56: of muons above the observed value, $(dE/dx)_{\rm obs}$. The expected
57: $dE/dx$ distribution of muons is obtained by using cosmic-ray muons.
58: Each plane's confidence level is combined to form a total confidence
59: level, assuming the confidence level at each layer is independent. The
60: MuCL is calculated as
61: \begin{eqnarray}
62: {\rm MuCL} = P \times \sum_{i=0}^{n-1} \frac{(-\ln P)^i}{i!}
63: \end{eqnarray}
64: where $n$ is the number of planes penetrated by the track,
65: $P = \prod_{i=1}^{n} {\rm CL}_{i}$, ${\rm CL}_{i}$ is the confidence level
66: at the $i$-th plane.
67:
68: Fig.~\ref{fig:dedx_mu_p.eps} shows the $dE/dx$ distributions of muon
69: and proton enriched samples. The predicted distributions of true muon
70: and proton tracks are shown as hatched histograms. To select muon
71: candidates for this study, we first select SciBar-MRD matched
72: tracks. According to the MC simulation, the sample is 94.7\% pure
73: muons with a small contamination of protons and charged pions. For
74: proton candidates, we select the second track in a charged current
75: quasi-elastic (CC-QE) scattering enriched sample made by cutting on a
76: kinematic variable described later. The fraction of protons in the
77: sample is 92.1\%, estimated with the MC simulation. The contamination
78: of charged pions and muons are estimated to be 5.5\% and 1.6\%,
79: respectively. Proton candidates are clearly separated from muon
80: candidates.
81:
82: \begin{figure}[tbp]
83: \begin{center}
84: \includegraphics[keepaspectratio=true,height=40mm]{figures/dedx_mu_p.eps}
85: \end{center}
86: \caption{(Color online) $dE/dx$ of muon enriched sample (left) and proton
87: enriched sample (right).}
88: \label{fig:dedx_mu_p.eps}
89: \end{figure}
90:
91: The MuCL distributions for the muon enriched sample and the proton
92: enriched sample are shown in Fig.~\ref{fig:mucl_mu_p.eps}. Tracks with
93: MuCL greater than 0.05 are considered muon-like (or pion-like) and the
94: others are classified as proton-like. The probability of
95: misidentification is estimated to be 1.1\% for muons and 12\% for
96: protons, averaged over track length in the muon and proton enriched
97: samples.
98:
99: \begin{figure}[tbp]
100: \begin{center}
101: \includegraphics[keepaspectratio=true,height=40mm]{figures/mucl_mu_p.eps}
102: \end{center}
103: \caption{(Color online) MuCL of muon enriched (left) and proton
104: enriched (right) samples.}
105: \label{fig:mucl_mu_p.eps}
106: \end{figure}
107:
108:
109: \subsection{Charged Current Event Selection}
110:
111: Events with at least one SciBar-MRD matched track are selected as charged current
112: event candidates. We reject events with hits associated with the muon
113: candidate on the most upstream layer of SciBar to eliminate incoming
114: particles due to neutrino interactions in the upstream wall or
115: soil. The hit threshold for this veto cut is set to two
116: photoelectrons. The neutrino interaction vertex is reconstructed as
117: the upstream edge of the muon track. The vertex resolution is
118: approximately 0.5~cm in each dimension, estimated with the MC
119: simulation. We select events whose vertices are in the SciBar fiducial
120: volume, defined to be $\pm$130~cm in both the $x$ and $y$ dimensions,
121: and 2.62~cm$<z<$157.2~cm, a total mass of 10.6~tons. The background
122: contamination due to neutrino events which occur in the EC and MRD is
123: 2.0\% and 0.5\%, respectively. Finally, the time of the muon
124: candidate is required to be within a 2~$\mu$sec window around the beam
125: pulse. The cosmic-ray background contamination in the beam timing
126: window is only 0.5\%, estimated using a beam-off timing
127: window. According to the MC simulation, the selection efficiency and
128: purity of true $\nu_\mu$ charged current events are 27.9\% and 92.8\%,
129: respectively. Impurity comes from $\nu_\mu$ neutral current events (3.0\%),
130: $\overline{\nu}_\mu$ charged current events (1.6\%), and neutrino events which
131: occur in the EC/MRD (2.5\%). The average neutrino beam energy for
132: true charged current events in the sample is 1.2~GeV. This SciBar-MRD matched
133: sample is our standard charged current data set and defines the MC normalization,
134: i.e. the MC distributions are normalized to the number of SciBar-MRD
135: matched events in data.
136:
137: Two sub-samples of the SciBar-MRD matched sample are further defined;
138: the MRD stopped sample and the MRD penetrated sample. Events with the
139: muon stopping in the MRD are classified as MRD stopped events,
140: in which we can measure the muon momentum. Events with the muon exiting
141: from the downstream end of the MRD are defined as the MRD penetrated
142: sample, in which we can measure only a part of the muon momentum. The
143: average neutrino beam energy for true charged current events in the MRD stopped and
144: MRD penetrated samples are 1.0~GeV and 2.0~GeV, respectively, enabling
145: a measurement of charged current coherent pion production at two different mean neutrino
146: energies.
147:
148: The slopes of the muon angles with respect to the beam in the two
149: SciBar views are used to calculate the three dimensional muon angle
150: with respect to the beam ($\theta_\mu$).
151: The kinetic energy of the muon is calculated by
152: the range and expected energy deposition per unit length ($dE/dx$) in
153: SciBar, the EC and the MRD,
154: \begin{eqnarray}
155: E_{\rm kin} &=& E^{\rm SB} + E^{\rm EC} + E^{\rm MRD} \nonumber \\
156: &=& \left.\frac{dE}{dx}\right|_{\rm SB} L_{\rm SB}
157: +\frac{\Delta E_0^{\rm EC}}{\cos \theta_\mu} + E^{\rm MRD}(L_{\rm MRD})
158: \end{eqnarray}
159: where $E^{\rm SB}$, $E^{\rm EC}$, and $E^{\rm MRD}$ are the energy
160: deposition in each detector. $L_{\rm SB}$ and $L_{\rm MRD}$ are the
161: track length of the muon in SciBar and the range in the MRD,
162: respectively. We set $dE/dx|_{\rm SB}$ to 2.04~MeV/cm, and $\Delta
163: E_0^{\rm EC}$, which is the energy deposited in the EC by a
164: horizontally traversing minimum ionizing particle, is set to 91~MeV,
165: estimated with the GEANT4 simulation. $E^{\rm MRD}$ is calculated
166: from a range to energy lookup table based on the MC simulation. For
167: muons stopping in the MRD, the average muon momentum and muon angular
168: resolutions are 50~MeV/$c$ and 0.9~degree, respectively. For muons
169: exiting the MRD, only a lower limit on muon momentum is obtained,
170: while the muon angle is determined with the same resolution as that of
171: stopping muons. The systematic uncertainty in the muon momentum scale
172: is estimated to be 2\% which is dominated by the difference among
173: various calculations of the range to energy lookup table.
174:
175: \subsection{Event Classification}
176: \label{sec:event_classification}
177: The MRD stopped and MRD penetrated samples are further divided into sub-samples
178: with the same selection criteria.
179: Once the muon candidate and the neutrino interaction vertex are
180: reconstructed, we search for other tracks originating from the
181: vertex. For this purpose, the track edge distance is defined as the 3D
182: distance between the vertex and the closer edge of another
183: reconstructed track. Tracks whose edge distance is within 10~cm are
184: called vertex-matched tracks. Fig.~\ref{fig:ntracks.eps} shows the
185: distribution of the number of tracks at the vertex for the MRD
186: stopped sample. For the MC simulation, the contributions from
187: charged current coherent pion, charged current resonant pion, charged current quasi-elastic, and other interactions
188: are shown separately.
189: Most events are reconstructed as either one track or two track events.
190: The two track sample is further divided based on particle identification. We
191: first require that the MuCL of the SciBar-MRD matched track is greater
192: than 0.05 to reject events with a proton penetrating into the MRD.
193: Then the second track in the event is classified as a pion-like or a
194: proton-like track with the same MuCL threshold. Fig.~\ref{fig:mucl.eps}
195: shows the contributions to the second track from
196: true proton, pion, muon, and electron tracks as predicted
197: by the MC simulation.
198:
199: In a charged current resonant pion event, $\nu p\to \mu^- p\pi^+$, the proton is
200: often not reconstructed due to its low energy, and such an event is
201: therefore identified as a two track $\mu+\pi$ event. To separate charged current
202: coherent pion events from charged current resonant pion events, additional protons
203: with momentum below the tracking threshold are instead detected by
204: their large energy deposition around the vertex, so-called vertex
205: activity. We search for the maximum deposited energy in a strip
206: around the vertex, an area of $12.5 \ {\rm cm} \times 12.5 \ {\rm cm}$
207: in both views. Fig.~\ref{fig:vtxedep.eps} shows the maximum energy
208: for $\mu+\pi$ events in the MRD stopped sample.
209: A peak around 6~MeV corresponds to the energy deposited in the strip containing the vertex
210: by two minimum ionizing particles, and a high energy tail is mainly due to the low energy proton.
211: To simulate such protons, we consider re-interactions of nucleons in the nucleus as described
212: in Sec.~\ref{sec:intra_nuclear} as well as ones outside the nucleus
213: (described in Sec.~\ref{subsec:detector_mc}).
214: De-excitation gamma-rays from the carbon nucleus do not affect the distribution since most of
215: the gamma-rays first interact outside the vertex region.
216: Events with energy deposition greater than 10 MeV are considered to have activity at the
217: vertex. Charged current coherent pion candidates are extracted from the sample of
218: $\mu+\pi$ events without vertex activity. Four sub-samples, the
219: one track events, $\mu+p$ events, $\mu+\pi$ events with vertex activity,
220: and $\mu+\pi$ events without vertex activity in the MRD stopped sample
221: are used for constraining systematic uncertainties in the
222: simulation, described next.
223:
224: \begin{figure}[tbp]
225: \begin{center}
226: \includegraphics[keepaspectratio=true,height=60mm]{figures/ntracks.eps}
227: \end{center}
228: \caption{(Color online) Number of vertex-matched tracks for the MRD stopped sample.
229: The MC distribution shown here is before tuning.}
230: \label{fig:ntracks.eps}
231: \end{figure}
232:
233: \begin{figure}[tbp]
234: \begin{center}
235: \includegraphics[keepaspectratio=true,height=60mm]{figures/mucl_2nd.eps}
236: \end{center}
237: \caption{(Color online) MuCL for the second track in the two-track sample.
238: The MC distribution shown here is before tuning.}
239: \label{fig:mucl.eps}
240: \end{figure}
241:
242: \begin{figure}[tbp]
243: \begin{center}
244: \includegraphics[keepaspectratio=true,height=60mm]{figures/vtxedep.eps}
245: \end{center}
246: \caption{(Color online) Maximum deposited energy in a strip around the vertex
247: for the $\mu+\pi$ events. The MC distribution shown here is before tuning.}
248: \label{fig:vtxedep.eps}
249: \end{figure}
250:
251: \subsection{Tuning the Monte Carlo Simulation}
252: The MC simulation includes systematic uncertainties due to the detector response,
253: nuclear effects, neutrino interaction models, and neutrino beam
254: spectrum, and these uncertainties affect background estimation. The sources of
255: systematic uncertainty are summarized in Section~\ref{subsec:systematic}.
256: In order to constrain these uncertainties, the MC distributions of
257: the square of the four-momentum transfer ($Q^2$) are fitted to the distributions
258: of the four aforementioned data samples. The reconstructed $Q^2$ is calculated as
259: \begin{equation}
260: Q^{2}_{\rm rec} = 2 E_{\nu}^{\rm rec} ( E_{\mu} - p_{\mu} \cos \theta_{\mu} ) - m_{\mu}^2
261: \label{eq:q2rec}
262: \end{equation}
263: where $E_\nu^{\rm rec}$ is the reconstructed neutrino energy calculated by assuming
264: charged current quasi-elastic kinematics,
265: \begin{equation}
266: E_{\nu}^{\rm rec} = \frac{1}{2}
267: \frac{(m_p^2-m_\mu^2)-(m_n-V)^2+2E_\mu(m_n-V)}{(m_n-V)-E_\mu+p_\mu \cos \theta_\mu}
268: \label{eq:enurec}
269: \end{equation}
270: where $m_p$ and $m_n$ are the mass of proton and neutron,
271: respectively, and $V$ is the nuclear potential, which is set to 27~MeV.
272: The one track events, $\mu+p$ events, and $\mu+\pi$ events with and without
273: vertex activity are fit simultaneously. Each $Q^2_{\rm rec}$
274: distribution is fit in bins of width 0.05~(GeV/$c$)$^2$ up to
275: 1~(GeV/$c$)$^2$.
276:
277: We introduce eight fitting parameters; the normalization factor of the
278: MRD stopped sample ($R_{\rm norm}$), the resonant pion scale factor
279: ($R_{\rm res}$), the scale factor of other non-QE interactions
280: ($R_{\rm other}$), the ratio of the number of two track events to the
281: number of one track events ($R_{\rm 2trk/1trk}$), the ratio of the
282: number of $\mu+p$ events to the number of $\mu+\pi$ events
283: ($R_{p/\pi}$), the ratio of the number of low vertex activity
284: $\mu+\pi$ events to the number of high vertex activity $\mu+\pi$
285: events ($R_{\rm act}$), the muon momentum scale ($R_{\rm pscale}$),
286: and a charged current quasi-elastic Pauli-suppression parameter~$\kappa$. All parameters are
287: ratios to nominal values in the MC simulation, i.e. all parameters
288: are set to 1 in the default MC simulation.
289:
290: The parameters $R_{2trk/1trk}$, $R_{p/\pi}$, and $R_{\rm act}$
291: represent possible migrations between subsamples due to systematic
292: uncertainties. The parameter $R_{\rm pscale}$ changes the scale of
293: the reconstructed muon momentum for the MC simulation. The parameter
294: $\kappa$, which was first introduced by MiniBooNE \cite{:2007ru},
295: controls the strength of Pauli-blocking and thus
296: suppresses low $Q^2$ charged current quasi-elastic events. We employ this parameter in the
297: fitting because a deficit of data is found at low $Q^2$ in the one track
298: sample where the charged current quasi-elastic interaction is dominant.
299:
300: The $\chi^2$ function to be minimized is given by:
301: \begin{eqnarray}
302: \chi^2 = \chi^2_{\rm dist} + \chi^2_{\rm sys}.
303: \end{eqnarray}
304: The term $\chi^2_{\rm dist}$ is calculated using a binned likelihood defined as~\cite{Baker:1983tu}:
305: \begin{eqnarray}
306: \chi^2_{\rm dist} &=& -2\sum_{i,\ j}
307: \ln \frac{P(N_{ij}^{\rm obs};N_{ij}^{\rm exp})}{P(N_{ij}^{\rm obs};N_{ij}^{\rm obs})} \nonumber \\
308: &=& 2 \sum_{i,\ j} \left( N_{ij}^{\rm exp}-N_{ij}^{\rm obs}+N_{ij}^{\rm obs}
309: \times \ln \frac{N_{ij}^{\rm obs}}{N_{ij}^{\rm exp}} \right)
310: \end{eqnarray}
311: where $P(n,\nu)=\nu^n e^{-\nu}/n!$ is the Poisson probability of finding $n$ events with
312: a expectation value $\nu$,
313: $N_{ij}^{\rm obs}$ and $N_{ij}^{\rm exp}$ are the observed and expected
314: number of events in the $i$-th $Q^2$ bin in subsample $j$ ($j=$one track, $\mu+p$,
315: $\mu+\pi$ with high and low vertex activity), respectively.
316: %\begin{eqnarray}
317: % \chi^2_{\rm dist} = 2 \sum_{i,\ j}
318: % \left( N_{ij}^{\rm exp}-N_{ij}^{\rm obs}+N_{ij}^{\rm obs}
319: % \times \ln \frac{N_{ij}^{\rm obs}}{N_{ij}^{\rm exp}} \right)
320: %\end{eqnarray}
321: %where $N_{ij}^{\rm obs}$ and $N_{ij}^{\rm exp}$ are the observed and expected
322: %number of events in the $i$-th $Q^2$ bin in subsample $j$ ($j=$one track, $\mu+p$,
323: %$\mu+\pi$ with high and low vertex activity), respectively.
324: The expected number of events for each sample is given by:
325: \begin{eqnarray}
326: N_{i,\ {\rm 1trk}}^{\rm exp}
327: &=& R_{\rm norm} \nonumber \\
328: & & \cdot \left[ n_{i, {\rm 1trk}}^{\rm QE}+R_{\rm res} n_{i, {\rm 1trk}}^{\rm res}
329: +R_{\rm other} n_{i, {\rm 1trk}}^{\rm other} \right] \\
330: N_{i,\ \mu p}^{\rm exp}
331: &=& R_{\rm norm} \cdot R_{\rm 2trk/1trk} \cdot R_{p/\pi} \nonumber \\
332: & & \cdot \left[ n_{i,\mu p}^{\rm QE}+R_{\rm res} n_{i,\mu p}^{\rm res}
333: +R_{\rm other} n_{i,\mu p}^{\rm other} \right] \\
334: N_{i,\ \mu\pi {\rm H}}^{\rm exp}
335: &=& R_{\rm norm} \cdot R_{\rm 2trk/1trk} \nonumber \\
336: & & \cdot \left[ n_{i,\mu\pi {\rm H}}^{\rm QE}+R_{\rm res} n_{i,\mu\pi {\rm H}}^{\rm res}
337: +R_{\rm other} n_{i,\mu\pi {\rm H}}^{\rm other} \right]
338: \end{eqnarray}
339: \begin{eqnarray}
340: N_{i,\ \mu\pi {\rm L}}^{\rm exp}
341: &=& R_{\rm norm} \cdot R_{\rm 2trk/1trk} \cdot R_{\rm act} \nonumber \\
342: & & \cdot \left[ n_{i,\mu\pi {\rm L}}^{\rm QE}+R_{\rm res} n_{i,\mu\pi {\rm L}}^{\rm res}
343: +R_{\rm other} n_{i,\mu\pi {\rm L}}^{\rm other} \right]
344: \label{eq:num_mupi}
345: \end{eqnarray}
346: where $n_{i,\ j}^{\rm QE}$, $n_{i,\ j}^{\rm res}$, $n_{i,\ j}^{\rm other}$
347: are the number of charged current quasi-elastic, charged current resonant pion, and other events
348: in each bin in each subsample, respectively.
349: $R_{\rm pscale}$ and $\kappa$ do not appear explicitly in these equations,
350: but $R_{\rm pscale}$ causes migration between $Q^2$ bins and $\kappa$ changes
351: $n_{i,\ j}^{\rm QE}$.
352:
353: The term $\chi^2_{\rm sys}$, added to constrain systematic parameters, is
354: calculated as:
355: \begin{eqnarray}
356: \chi^2_{\rm sys} = (\bm{P_{sys}}-\bm{P_0}) \bm{V}^{-1} (\bm{P_{sys}}-\bm{P_0})
357: \end{eqnarray}
358: where $\bm{P_{sys}}$ represents the set of systematic parameters and
359: $\bm{P_0}$ is the set of parameter values before fitting, expressed as:
360: \begin{eqnarray}
361: \bm{P_{sys}} = \left(
362: \begin{array}{c}
363: R_{\rm res} \\
364: R_{\rm 2trk/1trk} \\
365: R_{p/\pi} \\
366: R_{\rm pscale}
367: \end{array}
368: \right)
369: \quad , \quad
370: \bm{P_0} = \left(
371: \begin{array}{c}
372: 1 \\
373: 1 \\
374: 1 \\
375: 1
376: \end{array}
377: \right).
378: \end{eqnarray}
379: $\bm{V}$ is a covariance matrix estimated by considering the possible variations
380: due to systematic uncertainties in the detector responses, nuclear effects,
381: neutrino interaction models, and neutrino beam spectrum.
382: We prepare several MC event sets by changing each underlying physics parameter,
383: i.e. the source of systematic uncertainty, by $\pm 1 \sigma$. The covariance
384: between two systematic parameters $p_i$ and $p_j$ is calculated as:
385: \begin{eqnarray}
386: V_{ij}\equiv {\rm cov}[p_i,p_j]
387: = \sum_{\rm source} \frac{\Delta p_i \Delta p_j |_+ + \Delta p_i \Delta p_j |_-}{2}
388: \end{eqnarray}
389: where $\Delta p_i \Delta p_j |_{+(-)}$ is the product of variations of two parameters
390: when the underlying physics parameter is increased (decreased) by the size of its
391: uncertainty. The covariance matrix is estimated to be:
392: \begin{eqnarray}
393: \bm{V} = \left(
394: \begin{array}{cccc}
395: \ (0.20)^2 & -(0.09)^2 & +(0.10)^2 & 0 \\
396: -(0.09)^2 & \ (0.09)^2 & -(0.07)^2 & 0 \\
397: +(0.10)^2 & -(0.07)^2 & \ (0.15)^2 & 0 \\
398: 0 & 0 & 0 & (0.02)^2
399: \end{array}
400: \right).
401: \end{eqnarray}
402: $R_{\rm norm}$, $R_{\rm other}$, $R_{\rm act}$, and $\kappa$ are unconstrained in the fit.
403:
404: Events with $Q^2_{\rm rec}<0.10 \ ({\rm GeV}/c)^2$ in the $\mu+\pi$
405: sample with low activity are not included in the fit to avoid charged current
406: coherent pion signal events. A data excess is observed in the region
407: with $Q^2_{\rm rec}<0.15
408: \ ({\rm GeV}/c)^2$ in the $\mu+p$ sample.
409: Further investigation reveals that the second track in the excess
410: events is emitted at a relatively large angle with respect to the beam
411: direction and has large $dE/dx$, thus the events have an
412: additional large energy deposition at the vertex. Each of these events
413: seems to have a muon and a proton with additional activity, and
414: therefore the excess is not expected to affect the charged current coherent pion
415: analysis. A possible candidate for the excess is charged current resonant
416: pion production where the pion is absorbed in the nucleus. In such an
417: event, two or more additional nucleons should be emitted after the
418: pion is absorbed, which is currently not simulated. The excess cannot
419: be explained with the introduced fitting parameters, and therefore the
420: region is not used in the fit.
421:
422: Fig.~\ref{fig:q2rec_after.eps} shows reconstructed $Q^2$ after the
423: fitting for the one track, $\mu+p$, and $\mu+\pi$ events with and without
424: vertex activity. The best fit values and errors of the fit
425: parameters are summarized in Table~\ref{table:best_fit}. These same
426: fit parameters are also applied to the MRD penetrating sample. The
427: $\chi^{2}/$d.o.f before the fit is $473/75=6.31$. The
428: $\chi^{2}/$d.o.f after the fit is $117/67=1.75$. Even after fitting,
429: the reduced $\chi^2$ is relatively large, which indicates that the
430: introduced parameters are not sufficient in fully reproducing the data. To take
431: into account the incompleteness of our simulation, we enlarge the errors
432: on the fitting parameters by a factor of $\sqrt{\chi^2/{\rm d.o.f}}$.
433:
434: \begin{figure}[htbp]
435: \begin{center}
436: \includegraphics[keepaspectratio=true,height=53mm]{figures/q2rec_1trk.eps}
437: \includegraphics[keepaspectratio=true,height=53mm]{figures/q2rec_mup.eps}
438: \includegraphics[keepaspectratio=true,height=53mm]{figures/q2rec_mupi_act.eps}
439: \includegraphics[keepaspectratio=true,height=53mm]{figures/q2rec_mupi_noact.eps}
440: \end{center}
441: \caption{(Color online) Reconstructed $Q^2$ after fitting for (a) the one track,
442: (b) $\mu+p$, (c) $\mu+\pi$ with activity, and (d) $\mu+\pi$ without activity samples.}
443: \label{fig:q2rec_after.eps}
444: \end{figure}
445:
446: \begin{table}[htbp]
447: \caption{Best fit values and errors of the fitting parameters}
448: \label{table:best_fit}
449: \begin{center}
450: \begin{tabular}{lrrr}
451: \hline \hline
452: Parameter \quad & \quad Value & \quad Error \\
453: \hline
454: $R_{\rm norm}$ & 1.103 & 0.029 \\
455: $R_{\rm 2trk/1trk}$ & 0.865 & 0.035 \\
456: $R_{p/\pi}$ & 0.899 & 0.038 \\
457: $R_{\rm act}$ & 0.983 & 0.055 \\
458: $R_{\rm pscale}$ & 1.033 & 0.002 \\
459: $R_{\rm res}$ & 1.211 & 0.133 \\
460: $R_{\rm other}$ & 1.270 & 0.148 \\
461: $\kappa$ & 1.019 & 0.004 \\
462: \hline \hline
463: \end{tabular}
464: \end{center}
465: \end{table}
466:
467: \subsection{Charged Current Coherent Pion Event Selection}
468: Charged current coherent pion candidates are extracted from both the MRD stopped and MRD penetrated
469: samples with the same selection criteria. In this section, we describe the event
470: selection for the MRD stopped sample. The event selection for the MRD penetrated
471: sample is summarized later.
472:
473: After selecting $\mu+\pi$ events which do not have vertex activity,
474: the sample still contains charged current quasi-elastic events in which a proton is
475: misidentified as a minimum ionizing track. We reduce this charged current quasi-elastic background by
476: making use of kinematic information in the event. Since the charged current quasi-elastic
477: interaction is a two-body interaction, one can predict the proton
478: direction from the measured muon momentum $p_\mu$ and muon angle
479: $\theta_\mu$;
480: \begin{equation}
481: \vec{p}_p = (-p_{\mu x},-p_{\mu y},E_\nu^{\rm rec}-p_\mu \cos \theta_\mu)
482: \label{eq:expected_proton_direction}
483: \end{equation}
484: where $p_{\mu x}$ and $p_{\mu y}$ are the projected muon momentum in the $x$ and $y$
485: dimension, respectively. $E_\nu^{\rm rec}$ is the reconstructed neutrino energy given by
486: Equation~\ref{eq:enurec}.
487: For each two-track event, we define an angle called $\Delta\theta_p$ as the angle between
488: the expected proton track direction given by Equation~\ref{eq:expected_proton_direction} and
489: the observed second track direction.
490: Fig.~\ref{fig:dthetap.eps} shows the $\Delta\theta_p$ distribution for $\mu+\pi$ events
491: in the MRD stopped sample. Events with $\Delta\theta_p$ larger than 20 degrees are selected.
492: With this selection, 48\% of charged current quasi-elastic events in the $\mu+\pi$ sample are
493: rejected, while 91\% of charged current coherent pion events pass the cut according to the MC simulation.
494:
495: \begin{figure}[tbp]
496: \begin{center}
497: \includegraphics[keepaspectratio=true,height=60mm]{figures/dthetap.eps}
498: \end{center}
499: \caption{(Color online) $\Delta \theta_p$ for the $\mu+\pi$ events in the
500: MRD stopped sample after fitting.}
501: \label{fig:dthetap.eps}
502: \end{figure}
503:
504: Further selections are applied in order to separate charged current coherent pion
505: events from charged current resonant pion events which are the dominant backgrounds
506: for this analysis. Fig.~\ref{fig:theta2nd.eps} shows the angular
507: distribution of pion candidates with respect to the beam direction.
508: In the case of charged current coherent pion events, both the muon and pion tracks
509: are directed forward. Events in which the track angle of the pion
510: candidate with respect to the beam direction is less than 90~degrees
511: are selected.
512:
513: \begin{figure}[tbp]
514: \begin{center}
515: \includegraphics[keepaspectratio=true,height=60mm]{figures/theta2nd.eps}
516: \end{center}
517: \caption{(Color online) Track angle of the pion candidate with respect to
518: the beam direction for the $\mu+\pi$ events after the charged current quasi-elastic rejection after
519: fitting.}
520: \label{fig:theta2nd.eps}
521: \end{figure}
522:
523: \begin{figure}[tbp]
524: \begin{center}
525: \includegraphics[keepaspectratio=true,height=60mm]{figures/q2rec_mrdstop.eps}
526: \end{center}
527: \caption{(Color online) Reconstructed $Q^2$ for the $\mu+\pi$ events in the MRD
528: stopped sample after the pion track direction cut and after fitting.}
529: \label{fig:q2rec_mrdstop.eps}
530: \end{figure}
531:
532: Fig.~\ref{fig:q2rec_mrdstop.eps} shows the reconstructed $Q^2$
533: distribution for the $\mu+\pi$ events after the pion track direction
534: cut. Although a charged current quasi-elastic interaction is assumed, the $Q^2$ of charged current coherent
535: pion events is reconstructed with a resolution of 0.016~(GeV/$c$)$^2$
536: and a shift of -0.024~(GeV/$c$)$^2$ according to the MC simulation.
537: Finally, events with reconstructed $Q^2$ less than 0.1 (GeV/$c$)$^2$
538: are selected. The charged current coherent pion event selection is summarized in
539: Table~\ref{table:selection_mrdstop}. In the signal region, 247 charged current
540: coherent pion candidates are observed, while the expected number of
541: background events is 228$\pm$12. The error comes from the errors on
542: the fitting parameters summarized in Table~\ref{table:best_fit}. The
543: background in the final sample is dominated by charged current resonant pion
544: production. The ``other'' background is comprised of 50\% charged current DIS,
545: 32\% neutral current, and 18\% $\overline{\nu}_\mu$ events. The selection
546: efficiency for the signal is estimated to be 10.4\%.
547:
548: \begin{table}[tbp]
549: \caption{Event selection summary for the MRD stopped charged current coherent pion sample.}
550: \label{table:selection_mrdstop}
551: \begin{center}
552: \begin{tabular}{lrrrrr}
553: \hline \hline
554: Event selection & DATA & \multicolumn{2}{c}{MC} & Coherent $\pi$ \\
555: & & Signal & B.G. & Efficiency \\
556: \hline
557: Generated in SciBar fid.vol.& & 1,939 & 156,766 & 100\% \\
558: SciBar-MRD matched & 30,337 & 978 & 29,359 & 50.4\% \\
559: \hline
560: MRD stopped & 21,762 & 715 & 20,437 & 36.9\% \\
561: 2 track & 5,939 & 358 & 6,073 & 18.5\% \\
562: Particle ID ($\mu+\pi$) & 2,255 & 292 & 2,336 & 15.1\% \\
563: Vertex activity cut & 887 & 264 & 961 & 13.6\% \\
564: CC-QE rejection & 682 & 241 & 709 & 12.4\% \\
565: Pion track direction cut & 425 & 233 & 451 & 12.0\% \\
566: Reconstructed $Q^2$ cut & 247 & 201 & 228 & 10.4\% \\
567: \hline \hline
568: \end{tabular}
569: \end{center}
570: \end{table}
571:
572: \subsection{MRD penetrated Charged Current Coherent Pion Events}
573: The same selection is applied to the MRD penetrated sample to extract
574: charged current coherent pion candidates at higher energy. Fig.~\ref{fig:q2rec_mrdpenetrate.eps}
575: shows the reconstructed $Q^2$ distribution of the MRD penetrated charged current
576: coherent pion sample. The reconstructed $Q^2$ and $E_\nu$ for the MRD
577: penetrated sample are calculated from muon angle and
578: partially-reconstructed muon energy, using Equation~\ref{eq:q2rec} and
579: Equation~\ref{eq:enurec}, respectively. Although only a part of the
580: muon energy is observed, the $Q^2$ reconstruction performance is essentially
581: same because of the small muon angle. The event selection is
582: summarized in Table~\ref{table:selection_mrdpenetrate}. In the signal
583: region, 57 charged current coherent pion candidates are observed, while the expected
584: number of background events is 40$\pm$2.2. The background in the
585: final sample is dominated by charged current resonant pion production. The
586: ``other'' background is comprised of 75\% charged current DIS, and 25\%
587: $\overline{\nu}_\mu$ events. The selection efficiency for the signal
588: is estimated to be 3.1\%.
589:
590: \begin{figure}[tbp]
591: \begin{center}
592: \includegraphics[keepaspectratio=true,height=60mm]{figures/q2rec_mrdpenetrate.eps}
593: \end{center}
594: \caption{(Color online) Reconstructed $Q^2$ for the $\mu+\pi$ events in the MRD
595: penetrated sample after the pion track direction cut after fitting.}
596: \label{fig:q2rec_mrdpenetrate.eps}
597: \end{figure}
598:
599: \begin{table}[tbp]
600: \caption{Event selection summary of MRD penetrated charged current coherent pion sample.}
601: \label{table:selection_mrdpenetrate}
602: \begin{center}
603: \begin{tabular}{lrrrrr}
604: \hline \hline
605: Event selection & DATA & \multicolumn{2}{c}{MC} & Coherent $\pi$ \\
606: & & Signal & B.G. & Efficiency \\
607: \hline
608: Generated in SciBar fid.vol.& & 1,939 & 156,766 & 100\% \\
609: SciBar-MRD matched & 30,337 & 978 & 29,359 & 50.4\% \\
610: \hline
611: MRD penetrated & 3,712 & 177 & 4,375 & 9.1\% \\
612: 2 track & 1,029 & 92 & 1,304 & 4.7\% \\
613: Particle ID ($\mu+\pi$) & 418 & 78 & 474 & 4.0\% \\
614: Vertex activity cut & 167 & 71 & 186 & 3.6\% \\
615: CC-QE rejection & 134 & 67 & 135 & 3.5\% \\
616: Pion track direction cut & 107 & 66 & 109 & 3.4\% \\
617: Reconstructed $Q^2$ cut & 57 & 60 & 40 & 3.1\% \\
618: \hline \hline
619: \end{tabular}
620: \end{center}
621: \end{table}
622: