hep-ex0508053/data-analysis-gamcut/doc.tex
1:  The gamma-ray cut is used for reduction of external gamma rays which 
2: mainly come from the surrounding rock, PMT glass, and stainless steel 
3: support structure of the detector.
4: These gamma rays are one of the major background especially for solar 
5: neutrino data.
6: 
7:  The distinctive feature of external gamma rays is that they travel from 
8: the outside edge of the SK volume and continue on to the inside.
9: In order to remove this kind of event, reconstructed direction and vertex 
10: information is used to define the effective distance from the wall,
11: d$_{eff}$, as shown in the inset of the left plot in Figure~\ref{effwall}.
12: As they emanate from the wall itself, the value d$_{eff}$ for external 
13: gamma ray events should be small.
14: Figure~\ref{effwall} also shows d$_{eff}$ distributions for data and MC.
15: The criterion of the gamma ray cut is determined by maximizing its 
16: significance; here the number of remaining events after reduction
17: in data and MC are used to define the significance, $ N_{MC}/\sqrt{N_{data}}$.
18: The gamma ray cut criteria are determined to be:
19: \begin{enumerate}
20:  \item d$_{eff}$ $\ge$ 450~cm ( for $E \ge$ 6.5~MeV )
21:  \item d$_{eff}$ $\ge$ 800~cm ( for 5.0MeV $\le$ $E <$ 6.5~MeV)
22: \end{enumerate}
23: %
24: %
25: \begin{figure}[htbp]
26:   \begin{center}
27: %  \includegraphics[width=4.2cm]{./data-analysis-gamcut/effwall65mev.eps}
28: %  \includegraphics[width=4.2cm]{./data-analysis-gamcut/effwall50-65mev.eps}
29:   \includegraphics[scale=0.45]{./data-analysis-gamcut/effwall.eps}
30:   \caption{Effective distance (d$_{eff}$) of $E$ $\ge$ 6.5~MeV (left)
31:            and 5.0MeV $\le$ $E <$ 6.5~MeV (right).
32:            Blank histogram shows data and hatched area shows solar neutrino MC.}
33:   \label{effwall}
34:   \end{center}
35: \end{figure}
36: %
37: 
38:  Figure~\ref{fig:gamver} and \ref{fig:gamdir} show the vertex and direction
39: distributions before and after the gamma ray cut for real data in the  
40: energy regions $E$ $\ge$ 6.5~MeV and 5.0~MeV $\le$ $E <$ 6.5~MeV.
41: The dead times introduced by this cut in the energy regions 
42: $E$ $\ge$ 6.5~MeV and 5.0~MeV $\le$ $E <$ 6.5~MeV are estimated 
43: to be 6.9\% and 22.0\%.
44: 
45:  Since the gamma ray cut uses reconstructed vertex and direction, the 
46: differences of vertex and angular resolution between data and MC 
47: can introduce systematic errors. In order to estimate this effect, we
48: shift the reconstructed vertex and direction of events 
49: within the difference of data and MC, which are measured by LINAC,
50: and apply the gamma ray cut to these events.
51: The difference of reduction efficiency before and after this shift
52: is taken to be the systematic error for the gamma ray cut.
53: The systematic error of this cut for the flux measurement is estimated 
54: to be $\pm 0.5\%$ , while for the energy spectrum it is estimated 
55: to be $\pm 0.1\%$.
56: %
57: %
58: \begin{figure}[htbp]
59:   \begin{center}
60:   \includegraphics[scale=0.45]{./data-analysis-gamcut/gamver.eps}
61:   \caption{Vertex distribution before (blank) and after
62: 	   (hatched) applying the gamma ray cut. The left plots 
63:             are R$^2$ for
64:             $|$Z$|<$1000~cm, and the right ones are Z for $|$R$|<$1000~cm.
65:             The upper plots show $E$ $\ge$ 6.5~MeV, and lower ones show 
66:             5.0~MeV $\le$ $E <$ 6.5~MeV.}
67:   \label{fig:gamver}
68:   \end{center}
69: \end{figure}
70: %
71: \begin{figure}[htbp]
72:   \includegraphics[scale=0.38]{./data-analysis-gamcut/gamdir.eps}
73:   \caption{Directional distribution (X, Y, Z) before (blank) and after
74: 	   (hatched) applying the gamma ray cut. The upper plots  
75: 	   show $E$ $\ge$ 6.5~MeV,
76:            and the lower plots show 5.0~MeV $\le$ $E <$ 6.5~MeV.}
77:   \label{fig:gamdir}
78: \end{figure}
79: 
80: 
81: 
82: 
83: 
84: 
85: 
86: