0801.2417/ms.tex
1: \documentclass[
2:     ,final            % use final for the camera ready runs
3: %%  ,draft            % use draft while you are working on the paper
4: %%  ,numberedheadings % uncomment this option for numbered sections
5: %%  ,                 % add further options here if necessary
6:   ]
7:   {aipproc}
8: 
9: \layoutstyle{6x9}
10: 
11: \newcommand{\Ep}{\ensuremath{E_{peak}}}
12: 
13: 
14: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
15: %% FRONTMATTER
16: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
17: 
18: \begin{document}
19: 
20: \title{RHESSI Spectral Fits of \textit{Swift} GRBs}
21: 
22: %%\classification{<Replace this text with PACS numbers; choose from this list:
23: %%                \texttt{http://www.aip..org/pacs/index.html}>}
24: 
25: %98.70.Rz 	gamma-ray sources;gamma-ray bursts
26: %95.55.Ka 	X- and gamma-ray telescopes and instrumentation
27: %95.85.Pw (Astronomical observations--gamma ray)
28: 
29: \classification{98.70.Rz; 95.85.Pw}
30: \keywords      {gamma-rays: bursts}
31: 
32: \author{Eric C. Bellm}{
33:   address={UC Berkeley Space Sciences Laboratory, 7 Gauss Way,
34:   Berkeley, CA 94720-7450, USA}
35:   ,altaddress={Department of Physics, UC Berkeley} % additional visiting address
36: }
37: 
38: \author{Mark E. Bandstra}{
39:   address={UC Berkeley Space Sciences Laboratory, 7 Gauss Way,
40:   Berkeley, CA 94720-7450, USA}
41:   ,altaddress={Department of Physics, UC Berkeley} % additional visiting address
42: }
43: 
44: \author{Steven E. Boggs}{
45:   address={UC Berkeley Space Sciences Laboratory, 7 Gauss Way,
46:   Berkeley, CA 94720-7450, USA}
47:   ,altaddress={Department of Physics, UC Berkeley} % additional visiting address
48: }
49: 
50: \author{Wojtek Hajdas}{
51:   address={Paul Scherrer Institute, Villigen PSI, Switzerland}
52: }
53: 
54: \author{Kevin Hurley}{
55:   address={UC Berkeley Space Sciences Laboratory, 7 Gauss Way,
56:   Berkeley, CA 94720-7450, USA}
57: }
58: 
59: \author{David M. Smith}{
60:   address={Santa Cruz Institute for Particle Physics, Santa Cruz, CA, USA}
61: }
62: 
63: \author{Claudia Wigger}{
64:   address={Paul Scherrer Institute, Villigen PSI, Switzerland}
65: }
66: 
67: \begin{abstract}
68: One of the challenges of the \textit{Swift} era has been accurately determining
69: \Ep\ for the prompt GRB emission.  RHESSI, which is sensitive from
70: 30 keV to 17 MeV, can extend spectral coverage above the \textit{Swift}-BAT bandpass.
71: Using the public \textit{Swift} data, we present results of joint spectral fits for 26
72: bursts co-observed by RHESSI and \textit{Swift}-BAT through May 2007.  
73: We compare these fits to estimates
74: of \Ep\ which rely on BAT data alone.  A Bayesian \Ep\
75: estimator gives better correspondence with our measured results than an
76: estimator relying on correlations with the \textit{Swift} power law indices.
77: \end{abstract}
78: 
79: \maketitle
80: 
81: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
82: %% MAINMATTER
83: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
84: 
85: \section{GRB Prompt Emission Spectroscopy with RHESSI}
86: 
87: The Reuven Ramaty High-Energy Solar Spectroscopic Imager (RHESSI)
88: \citep{lin02} is a dedicated solar observatory.  
89: RHESSI's nine germanium detectors are sensitive
90: from 30 keV to 17 MeV, with excellent resolution in energy (1-5 keV) and
91: time (1 binary $\mu$s) \citep{smit02}.  Each of the nine coaxial
92: detectors is electronically segmented into front and rear segments.
93: Because the detectors are unshielded, RHESSI observes emission from
94: astrophysical sources like GRBs with a $\sim$2$\pi$ field of view.
95: 
96: We perform Monte Carlo simulations using the MGEANT package \citep{stur00} 
97: to determine RHESSI's response 
98: to off-axis sources like GRBs.  RHESSI's response varies with off axis angle, 
99: so we create responses every 15 degrees.  For each response, we simulate 
100: monoenergetic photons in 192 logarithmic energy bins ranging from
101: 30 keV -- 30 MeV.  Since RHESSI's per-detector response also varies during
102: the spacecraft's four second spin period, 
103: we bin the annular response in six azimuthal bins and weight these bins by the 
104: total burst lightcurve to create the final response.
105: 
106: The RHESSI data are extracted in SSW-IDL.  We fit and subtract a 
107: time-varying background, allowing for possible periodic modulation with 
108: the spin period.  We perform spectral fitting in ISIS \citep{citeisis}, 
109: a forward-fitting package analogous to 
110: XSPEC\footnote{http://heasarc.gsfc.nasa.gov/docs/xanadu/xspec/},
111: which is extensively programmable and allows 
112: computation of rigorous fluence error estimates via exploration of 
113: the parameter space.
114: 
115: Energetic charged particles over time have caused radiation damage to 
116: RHESSI's germanium detectors.  This radiation damage causes broadened 
117: spectral lines due to hole trapping and a general loss of active volume.  
118: %Without modeling the effect of the radiation damage on the continuum 
119: %response relevant for GRB spectroscopy, we are forced to 
120: In this work, we restrict our 
121: analysis to those detectors which do not exhibit signs of radiation damage.
122: At the time of \textit{Swift}'s launch, six RHESSI segments were usable for
123: spectroscopy; by May 2007 only two segments remained undamaged.  
124: In November 2007, 
125: RHESSI underwent an annealing procedure to reverse the effects of the
126: radiation damage.  The anneal restored some lost sensitivity, but 
127: analysis of future bursts will require more sophisticated modeling of
128: the remaining effects of radation damage on RHESSI's spectral response.
129: 
130: \section{RHESSI-BAT Joint Fits}
131: 
132: We attempted simultaneous spectral fitting for all RHESSI-observed GRBs
133: appearing in the first BAT Catalog \citep{saka07}.
134: Of 46 candidate bursts, 26 had sufficient RHESSI counts for spectral
135: analysis and produced acceptable joint fits.  We selected analysis time 
136: intervals manually from the RHESSI lightcurve using a S/N criterion. 
137: The resulting intervals were usually
138: shorter than those used in the BAT Catalog.  We generated BAT spectra and
139: responses for our intervals with the standard analysis 
140: procedures.\footnote{http://swift.gsfc.nasa.gov/docs/swift/analysis/threads/bat\_threads.html}
141: 
142: Typically, the joint fits did not require a normalization 
143: offset between the RHESSI and BAT data.  (Of the four bursts that 
144: needed an offset for acceptable fitting, one was in a period of highly modulated
145: RHESSI background, and three were after December 2006 when radiation 
146: damage was becoming severe.) 
147: %The BAT fit energy band for all bursts was 14 keV-151 keV.  
148: For the RHESSI data, we generally fit over the full 30
149: keV-17 MeV energy band.  For bursts coming from the rear of RHESSI, we
150: raised the lower energy bound to $\sim$60 keV, as the additional passive
151: material of the RHESSI cryostat can influence the low energy data.  For
152: GRB 061007, we omitted the RHESSI data above 3 MeV; there were no
153: significant counts above that level, but the residuals showed systematic
154: deviation which biased the fit.
155: 
156: Our results show good correspondence with comparable fits reported by
157: Konus-Wind and Suzaku-WAM, which also are sensitive in the MeV range
158: \citep[e.g.,][]{krim06, kgcn061007, sgcn061007, kgcn061121, page07}.
159: For 16 of the 26
160: bursts, the joint RHESSI-BAT best fit found additional model parameters
161: (\Ep\ and/or $\beta$) compared to the BAT-only fit.
162: 
163: We report the results of the joint fits in Table \ref{tab-fitpars}.
164: 
165: \section{Testing \Ep\ Estimators}
166: 
167: The peak energy \Ep\ of the prompt GRB spectrum is crucial to determining 
168: overall burst energetics, and it plays a key role in several proposed 
169: luminosity indicators.
170: The narrow passband of \textit{Swift}-BAT prevents determination of \Ep\ for 
171: many bursts.  Accordingly, a number of attempts have been made to infer
172: \Ep\ from the BAT data alone.  
173: \citet{butl07} used a Bayesian fit method with priors determined from the 
174: BATSE catalog to estimate \Ep.  \citet{zhan07} derived an \Ep\ -- BAT 
175: power law index correlation using hardness ratios (see also
176: \citep{zhan07b}).
177: In Figure \ref{fig-cfepeak}, we compare the predictions of these models 
178: to our joint fit results.
179: 
180: The Bayes model shows good correspondence with the measured values.  
181: Above $\sim$600 keV, the predicted values of \Ep\ tend to be low, 
182: although their error bars reach near the measured values.
183: %Predicted values of \Ep\ tend to be low when the measured value is above 
184: %$\sim$600 keV, but the measured value is near the upper bound of the predicted 
185: %error bar.  
186: This deviation is exaggerated somewhat, as our data
187: are for shorter, more intense (and typically harder) burst
188: intervals than used in \citep{butl07}.
189: 
190: Comparison of the Zhang et al. correlation-predicted \Ep\ to our
191: measured RHESSI-\textit{Swift} values shows that
192: this correlation appears to systematically underpredict the 
193: measured value of \Ep, especially at high energy.
194: 
195: \begin{figure}
196:   \resizebox{\textwidth}{!}{\includegraphics{fig1}}
197:   \caption{Comparison of the measured RHESSI+BAT \Ep\ values with those
198:   predicted by \citep{butl07} (left) and those found with the 
199:   correlation of \citep{zhan07} (right).
200:   Points marked with a cross have fit $\beta > -2$, and hence \Ep\ is only a
201:   formal value.  Points in the right plot marked with a diamond have
202:   BAT-only power
203:   law indices outside the range ($-2.3 < \alpha < -1.2$) likely to yield 
204:   accurate predictions of \Ep\ \citep{zhan07b}.  
205:   The overplot lines represent equality of the measured and predicted values.}
206:   \label{fig-cfepeak}
207: \end{figure}
208: 
209: %\subsection{<A subsection>}
210: %Some url test \url{http://www.world.universe}.
211: %\paragraph{<A subsubsubsection>}
212: %lamentabile\footnote{Here we test footnotes.} regnum cruerint Danai;
213: 
214: %\cite{Mittelbach/Schoepf:1990} a lacrimis? Et iam
215: %nox umida caelo praecipitat, suadentque \cite{Wang} cadentia
216: 
217: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
218: %% Sample figure:
219: %%
220: %% The option [height=...] scales the picture to the given height,
221: %% without it it would be printed at its nominal size
222: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
223: 
224: %\begin{figure}
225: %  \includegraphics[height=.3\textheight]{golfer}
226: %  \caption{Picture to fixed height}
227: %\end{figure}
228: 
229: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
230: %% BACKMATTER
231: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
232: 
233: \begin{theacknowledgments}
234: This work was supported by the \textit{Swift} AO-3 GI grant NNX08AE86G.
235: \end{theacknowledgments}
236: 
237: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
238: %% You may have to change the BibTeX style below, depending on your
239: %% setup or preferences.
240: %% For The AIP proceedings layouts use either
241: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
242: 
243: \bibliographystyle{aipproc}   % if natbib is available
244: %\bibliographystyle{aipprocl} % if natbib is missing
245: 
246: \bibliography{grb}
247: 
248: %fit parameters
249: \input{tab1} 
250: 
251: \end{document}
252: 
253: \endinput
254: