astro-ph0005532/ms.tex
1: %\documentstyle[12pt,aaspp4]{article}
2: \documentstyle[11pt,aaspp4]{article}
3: 
4: %\lefthead{LEFT head}
5: %\righthead{RIGHT head}
6: 
7: \begin{document}
8: %\Large
9: 
10: \title{EUVE Observations of Hercules X-1 During a Short High State Turn-On}
11: 
12: \author{D. A. Leahy}
13: \affil{Dept. of Physics, University of Calgary, University of Calgary,
14: Calgary, Alberta, Canada T2N 1N4}
15: \author{H. Marshall} %\altaffilmark{1}}
16: \affil{Massachusetts Institute of Technology, Cambridge, MA 02139}
17: %\altaffiltext{1}{}
18: \authoremail{leahy@iras.ucalgary.ca}
19: \and
20: \author{D. Matthew Scott \altaffilmark{1}}
21: \affil{Space Science Directorate, SD-50, NASA/Marshall Space Flight
22: Center, Huntsville, AL 35812}
23: \altaffiltext{1}{Universities Space Research Association}
24: \authoremail{scott@gibson.msfc.nasa.gov}
25:  
26: \begin{abstract}
27: Observations of Hercules X-1 by the Extreme Ultraviolet Explorer (EUVE)
28: covering low state and the early part of the Short High state 
29: are reported here. This is the first EUV observation of this part of
30: the 35-day cycle of Her X-1.
31: The low state portion of the EUV light curve (prior to the start of the Short 
32: High state) has similar properties as that following the end of the Short High 
33: state (\cite{lea99}). This is evidence that the low state EUV emission is 
34: primarily due to EUV reflection from the companion star HZ Her.
35: The EUV lightcurve 
36: during the Short High state is pulsed and closely resembles the average 
37: 2-12 keV X-ray Short High state lightcurve indicating that
38: the EUV emission, like the X-ray emission, originates near the neutron star. 
39: The Short High state EUV spectrum is consistent with a blackbody of 
40: temperature $\simeq 0.13$ keV and radius $\simeq230$ km. The Short High state 
41: EUV spectrum and pulse shape are similar to that in the soft X-rays 
42: (0.1 $-$ 1 keV). The most likely origin of the EUV emission is reprocessed 
43: X-rays from the inner edge of the accretion disk, and the radius of 
44: the inner edge of the accretion disk is likely to be small, consistent with 
45: that determined from analysis of the X-ray pulse shape evolution (\cite{sco00}).
46: \end{abstract}
47:  
48: \keywords{binaries: eclipsing --- star: neutron --- stars:
49: individual (HZ Her/ Her X-1)}
50: 
51: \section{Introduction}
52: 
53:       Hercules X-1 is one of the brightest, and most studied, of the 
54: persistant X-ray binary pulsars. The system displays a great variety of 
55: phenomena at many timescales, including pulsations at 1.24 seconds, eclipses 
56: at the orbital period of 1.7 days, and a 35-day cycle in the X-ray intensity 
57: that normally consists of a Main High state lasting 10-12 days and a Short 
58: High state lasting 5-7 days separated by 8-10 day long low states.   
59: Recent discussions of the properties of the 35-day cycle are given by
60: \cite{sco99} and \cite{sha98}. Her X-1 is reviewed by \cite{sco93}.
61: The X-ray pulse profile evolution is discussed 
62: in \cite{dee98}. Recent X-ray spectra of Her X-1 are given by \cite{oos98} and
63: \cite{dal98} (from BeppoSAX) and \cite{cho97} (from ASCA). An updated set of
64: binary parameters is given by \cite{lea98}. Analysis of ultraviolet spectra
65: of Her X-1 are presented by \cite{bor97} and \cite{vrt96}. Optical signatures 
66: of reprocessing on the companion and accretion disk are discussed by 
67: \cite{sti97}.
68:  
69: Her X-1 has the further advantage of a high galactic latitude and hence 
70: a low interstellar hydrogen column density, making extreme ultraviolet (EUV)
71: observations feasible.
72: Her X-1 has previously been observed several times in the EUV energy range 
73: (\cite{lea99}; \cite{roc94}; \cite{vrt94}). 
74: \cite{roc94} detected Her X-1 during a declining phase of the Short High
75: state with the Rosat Wide Field Camera.
76: The \cite{vrt94} observations occurred over the 35-day phase 0.14$-$0.245,
77: normally associated with the peak and flux decline of an average Main High 
78: state (\cite{sco99}). They state that their observations were during an 
79: ``anomalous low state''. \cite{lea99} observed Her X-1 at 35-day phase 
80: 0.76-0.88 which covers the end of the Short High state and the low state
81: \footnote{The 35-day phases quoted above are based on nearby Main High state
82: Turn-ons observed with the Burst and Transient Source Experiment (BATSE) 
83: on the Compton Gamma Ray Observatory (CGRO) at JD $-$ 2440000.0 = 9205.14 and 
84: 9936.22 for \cite{vrt94} and \cite{lea99}, respectively, rather than the values 
85: quoted in those papers that rely on longterm phase extrapolations.}.
86: From RXTE/ASM  observations, the average Short High state ends at 35-day phase 
87: 0.76-0.80 (\cite{sco99}).  
88: The count rate during the \cite{lea99} observation was $\sim0.02$ c/s and 
89: strongly modulated at the binary period. 
90: They concluded that most of the observed EUV emission was reflected emission
91: from the companion star HZ Her. In addition to the EUVE observations, soft 
92: X-ray observations with BeppoSAX of the middle and latter part of a Short 
93: High state are reported in \cite{oos00}.
94: 
95: Emission from Her X-1/HZ Her covers the optical, ultraviolet, EUV and
96: X-ray regime and models for the, often coupled, emission processes
97: must ultimately be consistent.
98: The hard X-rays ($>1$ keV) are believed to arise mostly as a result of 
99: mass accretion onto the neutron star and are modulated by the neutron star 
100: rotation and obscuration by the accretion disk (e.g. \cite{sco00} and 
101: references therein), companion star and moving gas ``blobs'' that cause 
102: the well known absorption dips 
103: (e.g. \cite{cro80}). 
104: A small reflected/reprocessed X-ray component 
105: is also present that is observable during the low state and eclipses 
106: (e.g. \cite{cho94}, \cite{lea95}). 
107: A major portion of the observed optical/ultraviolet emission is believed to 
108: arise from X-ray heating of HZ Her and the accretion disk. The X-ray heating 
109: causes the surface temperature of HZ Her facing the neutron star to be 
110: approximately 10,000 degrees higher than the cooler shadowed side 
111: (\cite{cheng95}). 
112: Observations of the broad band optical emission of HZ Her/Her X-1
113: have been presented by \cite{dee76} and \cite{vol90}, among others. The
114: broadband optical emission exhibits a complex, systematic variation pattern
115: over the course of the 35-day cycle in addition to the orbital modultation 
116: due to X-ray heating of HZ Her. This pattern can be explained as a consequence 
117: of disk emission and disk shadowing/occultation of the heated face of HZ Her 
118: by a tilted, counter-precessing accretion disk (\cite{ger76}). The precessing 
119: disk also causes the alternating pattern of High and low X-ray intensity 
120: states by periodically blocking the neutron star from view. The initial rapid 
121: flux rise over a few hours marks the start of a High state and is known as 
122: the ``turn-on''. This event is generally believed to be the emergence of the 
123: neutron star from behind the moving outer disk edge (see \cite{sco00} and 
124: references therein). Between the hard X-ray and 
125: optical/ultraviolet band lies the soft X-ray/extreme ultraviolet band 
126: ($\sim0.016$ $-$ 1 keV). A blackbody spectral component has been detected by 
127: many previous observations with a temperature of about 0.1 keV 
128: (e.g. \cite{shul75}; \cite{oos00}) and generally has been attributed to 
129: reprocessing of hard X-rays in the inner region of the accretion disk 
130: (e.g. \cite{mcc82}, \cite{oos00}).  
131:  
132: Here we report the results of an analysis of EUVE observations of Her X-1 
133: covering two complete orbital cycles including the first EUVE observation of
134: a turn-on to the Short High state of Her X-1.
135: 
136: \section{Observations}
137: 
138: Her X-1 was observed with the Extreme Ultraviolet Explorer (EUVE)
139: on July 25-29, 1997 
140: (Truncated Julian Day (TJD) = JD $-$ 2440000.0 = 10654.7 $-$ 10659.0).
141: See \cite{mal93} for a description of the EUVE instruments.
142: The 35-day phase interval of the EUVE observation is 0.49-0.62
143: based on the 35-day Turn-on time of the preceding Main High state 
144: from \cite{sco99} and a 35-day period $P_{35} = 20.5 P_{orb}$ where
145: $P_{orb}$ is the orbital period.
146: Her X-1 was detected during these observations as a source in  
147: the Deep Survey (DS) Spectrometer, with the Lexan/B filter. 
148: 
149: \section{Lightcurve}
150: 
151: The DS lightcurve of Her X-1 is given in  Fig. 1, which 
152: shows the net source count rate in the Deep Survey instrument for the 
153: observation period plotted as a function of time, 35-day phase and
154: orbital phase. The error bars are $\pm1\sigma$.
155: There is a sharp rise in the EUV flux (by a factor of 
156: $\sim10$) at 35-day phase 0.56.
157: This is consistent with the expected turn-on phase of a Short High
158: state following an ``0.7'' turn-on Main High state as this Short High does 
159: (\cite{sco99}). The Main and Short High states typically cover the 
160: 35-day phase intervals $\simeq 0-0.30$ and $\simeq 0.56-0.79$, respectively.
161: 
162: Orbital modulation at the 1.7 day orbital period of Her X-1 is clearly 
163: seen both before and after the turn-on. 
164: The EUV flux is within $2\sigma$ of zero during the orbital phase of 
165: the High state eclipse as measured, for example, by Ginga (\cite{lea95}).
166: Binary phase is defined here so that binary phase 0.0 
167: is the center of the eclipse of the neutron star by the companion.
168: 
169: The faint part of the lightcurve prior to the Turn-on is 
170: comparable in intensity to the faint phases of the EUVE lightcurve observed 
171: by \cite{vrt94} after TJD 9212.4 (see also \cite{vrt96}) and the bright part 
172: is comparable to the bright phases observed by \cite{vrt94} before TJD 9212.4
173: (peak count rate $\sim 0.5$ c/s). 
174: 
175: The orbital phase dependence of the faint part of the lightcurve is
176: similar to that observed in 1995 following a Short High state
177: (\cite{lea99}). A much broader, shallower, dip near orbital phase 0.5 is
178: observed rather than the deeper, narrower dips seen in 1995. The
179: peak count rate is approximately twice the 1995 peak count rate.
180: 
181: \section{Spectrum Analysis and Pulsation Search}
182: 
183: EUVE SWS spectra were extracted for both the faint and bright phases.
184: The faint phase has too few counts to give a useful spectrum.
185: The bright phase spectrum is shown in Figure 2, and covers the
186: wavelength range 65 to 179 \AA\ %Angstroms 
187: in 38 bins of 3 \AA\ %Angstrom 
188: width, which translates to 0.069 keV to 0.191 keV. 
189: 
190: The spectrum was fitted with a blackbody model, giving best-fit parameters
191: of $kT = 0.133  \pm  0.034$ keV, 
192: $N_H = 8.6 \pm 1.3 \times 10^{19}$ $\rm cm^{-2}$ and 
193: blackbody radius of $230\pm 40$ km assuming a distance of $6.6$ kpc to Her X-1.
194: %note errors are 90% errors
195: The fit was statistically good, with a $\chi^2$ of $34.2$ for $35$ degrees of
196: freedom. 
197: This fit is plotted as the histogram in Figure 2 (top panel), with the fit
198: residuals shown in the bottom panel.
199: We also fit the spectrum with a power law model  
200: with photon index $1.0$, giving best-fit parameters   
201: $N_H =  1.10 \pm 0.15 \times 10^{20}$ $\rm cm^{-2}$ and 
202: normalization of $0.32 \pm 0.09$ photons/keV/$\rm cm^2$/s at $1$ keV.
203: This fit was also statistically good, with a $\chi^2$ of $34.0$ for 
204: $35$ degrees of freedom. The power law model is indistinguishable by
205: eye from the blackbody model shown in Fig. 2.
206: Addition of a low temperature component to either the blackbody model
207: or power law model does not improve the fit. The present data does not
208: support the existence of a second, low temperature component.
209: The interstellar absorption to Her X-1 (e.g. from the blackbody fit)
210: is enough that a low temperature component (e.g. $\le 0.01$ keV)  would 
211: be reduced from its unabsorbed value by several orders of magnitude.
212: 
213: The bright phase data was used to perform a pulsation 
214: search using the epoch folding method with maximum likelihood test. 
215: The data times were first corrected to the solar system barycenter 
216: and then corrected for the orbit of Her X-1. The faint phase (prior to 
217: turn-on) has too few counts to perform a useful search. When all of the bright 
218: phase data is folded, best fit period and an alias are found at 
219: $1.237732(1)$ sec and $1.237742(1)$ sec.
220: The Her X-1 pulse frequencies for the Main High states immediately before and 
221: after the current Short High state, as measured with the standard BATSE 
222: pulsed monitor, are respectively: 1.237730 seconds and 1.237731 seconds. 
223: Based on these measurements, we choose the first of the two EUV periods for 
224: making the folded light curve shown in Fig. 3. The pulse
225: profile has a modulation of 10\%, using a definition of modulation as
226: the ratio of the equivalent amplitude of a sinusoidal variation to its mean 
227: value. 
228: %The best pulse period measured from the first orbit after 
229: %turn-on (TJD 10656.993 - 10657.676) is $1.237728\pm5$ sec with a $9\%$ 
230: %modulation. For data from the second orbit after turn-on 
231: %(TJD 10658.654 - 10658.885), the best pulse period is at $1.237751\pm4$ second
232: %with a $16\%$ modulation.  
233: 
234: \section{Discussion}
235: 
236: \subsection{EUVE low state observation (prior to Short High state turn-on)}
237: 
238: We have observed a Short High state turn-on in Her X-1 with the EUVE DS/SWS. 
239: The observations include just over one orbit of data during low state
240: prior to turn-on, and about 2.4 orbits of data after turn-on.
241: BATSE monitoring revealed normal Main High states flanking
242: this Short High state, so there is no reason to believe it is atypical.
243: 
244: The low state light curve is modulated at the orbital period (Fig. 1).
245: It was argued in \cite{lea99} that the low state EUV orbital modulation is 
246: caused by the changing orientation of the face of HZ Her reflecting EUV 
247: emission from the pulsar towards the observer. The EUV emission from the 
248: X-ray heated face of HZ Her is too dim to account for the observed EUV 
249: intensity, while EUV
250: reflection from the accretion disk would produce a constant EUV emission
251: component modulated only by eclipses and slow changes over the 35-day
252: cycle. The primary EUV emission originates in a relatively small region near 
253: the neutron star and is observable as the High state EUV emission, while the 
254: EUV emission reflected from HZ Her is modified by the disk shadow and by disk 
255: occultation. This model was described in \cite{lea99} and 
256: applied to the 1995 observations at the end of the Short High state 
257: (35-day phase 0.76-0.88). The occultation of HZ Her by the accretion disk 
258: causes the narrow dips in the lightcurve at orbital phase 0.5.
259: 
260: The peak intensity for the current EUV low state observation, corrected for 
261: the dip at orbital phase 0.5 (35-day phase 0.53), is a factor $\simeq2$ 
262: higher compared to the corrected peak intensities of the 1995 observations
263: in which orbital phase 0.5 occurred at 35-day phase $\simeq0.79$ and $\simeq0.84$.
264: The current low state light curve also lacks a strong, narrow dip at orbital 
265: phase 0.5.  
266: 
267: The shadow pattern cast by the disk onto HZ Her repeats every 35-days
268: at a given orbital phase as the disk precesses. From the point of view
269: of HZ Her the pattern repeats every 1.62 days: the beat period of the 
270: orbit period and the 35-day period. 
271: From disk precession, we expect the disk shadow pattern on HZ Her to be
272: equivalent to a shift by $\sim0.23-0.3$ in orbital phase between 
273: the 35-day phase 0.76-0.88 observations and the current 35-day phase 0.53 
274: observation. The shadow would appear later in orbital phase for the 0.53 
275: observation since the disk is counterprecessing.
276: 
277: The observed orbital light curve of X-rays reflected from HZ Her
278: requires an detailed calculation including the disk, star and observer geometry.
279: We have constructed a sample model for shadowing of HZ Her by the accretion disk
280: for a  twisted disk geometry with the following parameters.
281: The inner disk is tilted at $15^\circ$ from the binary plane and the outer 
282: disk is tilted at $25^\circ$. The outer disk line of nodes trails the line of 
283: nodes for the inner disk by $100^\circ$ in azimuth. These parameters were 
284: chosen in order to be approximately consistent with the observed 35-day 
285: phases of the start 
286: and end of both Main High and Short High states in Her X-1.
287: An illustration of the disk geometry is given in Fig. 4. It is drawn from the
288: perspective of the observer ($5^\circ$ above the orbital plane) at 35 day phase
289: 0.06. The resulting shadowing of the celestial
290: sphere, as seen from the neutron star, is shown in Fig. 5. The reference for
291: disk azimuth ($0^\circ$) is the line of nodes of the inner disk and elevation
292: angle is with respect to the binary plane.
293: Also shown in Fig. 5 is the limb of HZ Her, which is approximately a
294: circle of radius $25^\circ$ (elongated vertically  in Fig. 5 due to the
295: different angular scales on the two axes). 
296: Every 1.621 days, HZ Her moves $360^\circ$ in 
297: azimuth with respect to the disk (and disk shadow).
298: The disk shadow model was used to calculate the reflected intensity from HZ Her
299: for face-on geometry (i.e. orbital phase 0.5) as a function of 35-day phase 
300: (disk phase). The result is shown in Fig. 6, normalized to the maximum reflected
301: intensity. 
302: There is a modulation of the reflected intensity by a factor of 
303: $\sim2.7$ over disk phase. This  is nearly 
304: sinusoidal (with a period of $\pi$ radians, or 0.5 cycles, in disk phase) 
305: despite the complicated shadow pattern, 
306: due to the integration over the large angular size of HZ Her 
307: (as seen from the neutron star).
308:  
309: On Fig. 6 we also indicate the phase when the disk shadow on HZ Her is 
310: minimum (A). 
311: For orbital phase 0.5, this also coincides with minimum shadowing in
312: the direction of the observer, i.e. with the peak of the Short High state at 
313: 35-day phase $\simeq0.58$. With this reference, we can mark the 35-day phases 
314: when HZ Her is at orbital phase 0.5 during the EUVE low state 
315: observations:
316: B marks the 35-day phase for the 1997 low state and C and D mark the two 
317: 35-day phases for the two orbits of data for the 1995 low state observations.
318: The observed intensities for B, C and D at orbital phase 0.5 were 
319: corrected by a small amount for disk occultation of HZ Her (which is not
320: included in the model). The intensities
321: %are $\sim0.055$ c/s,  $\sim0.030$ c/s and $\sim0.027$ c/s.
322: were scaled by a common factor derived from normalizing the B intensity to the 
323: reflection model. On Fig. 6, these three intensities are marked by asterisks 
324: to compare to the model relative intensity. 
325: The agreement is good:  the increased shadowing of HZ Her by the accretion disk
326: at 35-day phases 0.79 and 0.84  compared to the shadowing  at 35-day phase 0.53 
327: is approximately reproduced by the model calculations. 
328: Another small ($\sim10\%$) roughly constant contribution to flux in the model,  
329: due to reflection of X-rays from the disk,  would reduce the amplitude of
330: the model curve and bring it into agreement with the observations.
331: Thus the current observation of Her X-1 prior to the turn-on is in qualitative
332: agreement with the idea that the
333: EUV emission in the low state is primarily due to reflection of X-rays from HZ Her.
334: 
335: %\subsection{Comparison of reflection model to previous optical observations}
336: 
337: We now consider whether the EUV X-ray reflection model is 
338: consistent with optical observations of HZ Her/Her X-1. \cite{vol90} present 
339: 16-year-average U, B and V light curves (Fig. 1 in that paper). They find  
340: the average U, B and V light curves differ by
341: a maximum of 20 to 22\% between ON state (which they defined as 35-day phase 
342: 0.85-0.15) and OFF state  (defined as 35-day phase 0.20-0.85).
343: However due to the large 35-day phase intervals, this is not very useful. 
344: Much more useful for comparison is the data presented by \cite{dee76}.
345: Optical B band fluxes as a function of orbital phase are presented in their 
346: Fig. 1. 
347: %separately for 20 different orbits of the 35-day cycle.  
348: The variation between different 35-day phases at a fixed orbital phase is
349: large: from a factor of $\simeq1.7$ around orbital phase $0.5$ to
350: a factor of $\simeq1.3$ around orbital phase $0.0$.
351: \cite{ger76} reproduce the optical orbital lightcurve and its
352: variation with 35-day period using a model including radiation due to
353: X-ray heating of HZ Her shadowed by the disk, emission by the disk, and
354: occultation of HZ Her by the disk. 
355: 
356: %We have done no detailed calculation of the optical radiation, but
357: %Based on the model of \cite{ger76} 
358: 
359: We argue that one expects the optical and EUV lightcurves to be different.
360: The estimated EUV heating flux is far too faint to explain the observed EUV 
361: observations (\cite{lea99}).
362: The EUV reflection model can produce enough flux, and uses the disk
363: shadow to cause a factor of 
364: $\sim2.7$ change in reflected intensity at orbital phase 0.5.  
365: The EUV reflection flux 
366: has a different distribution over the stellar surface than the 
367: optical flux, which is due to X-ray heating:
368: the optical flux is more concentrated toward the L1 point. 
369: The optical emission from the disk is significant,  
370: based on the model of \cite{ger76} for the observed optical flux.
371: %whereas the EUV flux from the disk appears to be smaller.
372: For an estimate of the disk component of the B flux, Fig. 7 of \cite{ger76}
373: gives a disk B flux of $\sim1.5-2.5$ (units defined in  \cite{ger76})
374: which is generally out of phase
375: with the stellar heating contribution (B flux $\sim0.5-4$). On average the
376: disk, in their model, contributes $\sim 40\%$ of the heating flux.
377:  The heating
378: flux alone varies by a factor of up to $\sim 7$ with orbital phase. If
379: we consider orbital phase 0.5, it varies by a factor $\sim7$ over 35 day phase,
380: which is larger than the observed variation in EUV flux, as expected
381: if the EUV flux is due to reflection.
382: %If the distribution of heating and reflection were the same then the optical 
383: %radiation would be less modulated due to the extra accretion disk component.
384: %On the other hand, the heating flux is more concentrated toward the L1 point,
385: %which makes the optical heating flux more variable that the EUV flux.
386: 
387: Thus the variation in EUV and optical emission
388: with orbital and 35-day phase are expected to be different.  
389: Empirically, one can compare the observed optical and EUV lightcurves.
390: A basic result found by \cite{dee76} was that the total optical radiation,
391: when averaged over an orbit, varied by less than a few percent over the
392: 35-day cycle. In contrast, the low state EUV orbital lightcurves observed 
393: here and in 1995 (\cite{lea99}) varied by roughly a factor of two in
394: total intensity.
395: %Thus there is no inconsistency of the EUV reflection model with the optical
396: %observations. 
397: 
398: The detailed form of both optical and EUV lightcurves depend on 
399: the shadowing geometry and the amount of disk radiation.
400: Thus EUV data over a wider 35-day phase range (in low state) and
401: more detailed modeling would be very useful to constrain the disk geometry 
402: using both EUV and optical data. 
403: 
404: \subsection{EUVE observation of the Short High state and the Short High state turn-on}
405: 
406: This EUVE observation is only the second high 
407: resolution observation of a Short High state turn-on performed to date as 
408: far as the authors know. 
409: A five-day long observation of the middle and latter 
410: part of a Short High state in soft X-rays has been made with BeppoSAX 
411: (\cite{oos00}).
412: Fig. 7 shows the EUV lightcurve (the histogram) 
413: at the time of turn-on as a function of time and of 35-day phase.
414: The smooth line in Fig. 7 is a fit to the Short High turn-on observed
415: by GINGA (\cite{sco00}, \cite{lea00}) 
416: and is plotted at the observed GINGA 35-day phase
417: without any phase shift to match the EUVE lightcurve.
418: The timescale for the EUV turn-on transition is 
419: $\simeq 3.5$ hours, the same as the timescale of turn-on observed by 
420: GINGA in 2-37 keV X-rays for both Short High state and Main High state 
421: (\cite{sco00}). This confirms that the Short High state turn-on is indeed sharp 
422: as opposed to the gradual rise predicted in some disk models for Her X-1 
423: (e.g. \cite{sch94}). There is strong evidence that the X-ray turn-on 
424: transition is due to reduced absorption as the outer edge of the disk moves 
425: out of the line-of-sight (e.g. \cite{beck77}; \cite{sco00}). 
426: Thus the increase in the EUV emission at turn-on is likely due to the same 
427: cause and hence the size of the EUV emission region, like the X-ray emission 
428: region, must be small compared with the scaleheight of the outer edge of the 
429: disk.
430: 
431: The first orbit after turn-on shows a strong dip at orbital phase 0.5
432: followed by a second dip after orbital phase 0.65. 
433: The flux rises soon after the eclipse in the second orbit after turn-on,
434: with a dip at orbital phase 0.19 before reaching a maximum.
435: The pattern is similar to that of the first orbit of ``0.2'' turn-on Main 
436: High states that show a turn-on, anomalous dip at orbital phase 0.5-0.6, 
437: preeclipse dip and a ``post-eclipse recovery'' (\cite{cro80}, \cite{sco99}).
438: The average RXTE/ASM Short High state lightcurves displayed in 
439:  Fig. 3 of \cite{sco99} also exhibits a similar pattern of
440: turn-on, anomalous dip, and preeclipse dip followed by a 
441: ``post-eclipse recovery''.
442: The BeppoSAX Short High state observation of \cite{oos00}
443: is very similar to the middle and latter part of the average
444: RXTE/ASM Short High state lightcurve and shows that the pattern of anomalous 
445: dip, preeclipse dip and ``post-eclipse recovery'' continues for each orbit 
446: of the Short High state.   
447: The very close similarity of the EUV, soft X-ray and 2-12 keV X-ray light 
448: curves implies that the same occulting structures cause the 
449: modulation observed in both the EUV and X-ray light curves, and the EUV and 
450: X-ray emission regions are much smaller in size than the occulting structures.
451: These occulting structures are probably the outer disk edge in the case
452: of the turn-on (for either High state type) or ``blobs'' of matter that
453: pass between the observer and the neutron star in the case of the dips 
454: (e.g. \cite{cro80}, \cite{sco00}). 
455: 
456: Our EUVE SWS Short High state bright phase spectrum is the highest
457: resolution spectrum yet presented in the $0.05$ to $0.2$ keV energy 
458: range. It demonstrates that the emission from Her X-1 is dominated by
459: continuum and not line emission.
460: Due to previous observations, we reject the power law model in favor of the 
461: blackbody model. Previous observations of the low energy (below 1 keV)
462: spectrum of Her X-1 have been made during both the Main High state and
463: the Short High state.  
464: The BeppoSAX LECS spectrum  at 35-day phase 0.1 (\cite{dal98})
465: is dominated below 1 keV by a blackbody component with $kT\simeq 0.092$ keV
466: and $N_H \simeq 5.1 \times 10^{19}$ $\rm cm^{-2}$, with blackbody
467: radius $\simeq 353$ km (for $d=6.6$ kpc).
468:  Similarly, a BeppoSAX observation of
469: a Short High state peak found a blackbody component with $kT\simeq 0.094$ keV 
470: and a blackbody radius $\simeq 209$ km (for $d=6.6$ kpc).
471: With the current Short High state bright phase spectrum, the various
472: measurements of temperature are consistent with  the blackbody component having
473: a constant temperature of $0.09$ keV throughout Main High and Short High states. 
474: The blackbody radius declines during the High States, 
475: and the early Short High state 
476: has approximately half the radius compared to the peak of Main High state. 
477: 
478: The current detection of EUV pulsations at the X-ray pulsation period 
479: during the Short High state implies that the EUV emission is coming from a 
480: compact region near the pulsar, less than several tenths of a light second 
481: across. The folded EUV pulse profile has a symmetric quasi-sinusoidal
482: form (Fig. 3) consistent with the pulse shape below 0.4 keV in X-rays 
483: found during the during Main High state (e.g. \cite{mcc82}; \cite{oos00}) and 
484: the Short High state (\cite{oos00}). The similarity of the pulse form
485: and spectrum of the EUV emission with the pulse form and spectrum of the 
486: Short High state soft X-ray emission suggests that these are simply the 
487: same emission component. 
488: The BeppoSAX observations of \cite{oos00}
489: clearly show the pulse changing form from a symmetrically peaked
490: quasi-sinusoid below 1 keV to an asymmetric quasi-sinusoid
491: above 1 keV. 
492: The change in form of the pulse with energy suggests that the 
493: soft X-ray/EUV emission and the harder X-ray quasi-sinusoidal pulsed emission 
494: originate from different physical regions, an idea reinforced by the 
495: corresponding change in the energy spectrum from a thermal form below 1 keV to 
496: a power-law form above 1 keV.  
497: 
498: The main pulsed emission above 1 keV during the peak of the Main High state
499: has a complicated profile and probably originates on the pulsar or
500: within a few radii of the neutron star (\cite{sco00}). The pulsed emission 
501: above 1 keV with a quasi-sinusoidal pulse profile has a power-law energy 
502: spectrum like the main pulsed emission so it is likely not reprocessed, but
503: rather scattered. A probable scattering location with a large solid angle
504: for radiation emitted close to the neutron star is the accretion column,
505: which may be optically thin to Thomson scattering for the conditions
506: applicable in Her X-1 (\cite{bra91}).
507: The thermal spectrum of the X-ray/EUV emission below
508: 1 keV suggests that it is reprocessed hard X-ray emission. 
509: The reprocessing region must be close to the neutron star to be pulsed,
510: yet not on the neutron star itself because the blackbody radius is too
511: large. Neither could it be an optically thin accretion column.   
512: The probable reprocessing location is the innermost region of the accretion 
513: disk. The deduced blackbody 
514: radius of $\sim230$ km from the EUVE/SWS spectrum supports this. 
515: This implies a quite small inner edge: only a few hundred
516: km, depending on the details of the geometry. The 
517: analysis of the pulse shape evolution (\cite{sco00}) also requires a small
518: radius for the inner edge of the accretion disk  (about $\sim240-480$ km).
519: Thus we have good evidence that the $0.1$ keV blackbody component from Her X-1 
520: is reprocessed emission from the inner edge of the disk as first suggested
521: by \cite{mcc82}.
522: 
523: The details of the reprocessing depend on the beam pattern from the neutron 
524: star and the geometry of the innermost region of the accretion disk. 
525: We take a simple disk edge and an inner disk tilt of $\simeq11^\circ$ discussed
526: by \cite{sco00} to explain the pulse evolution. Excluding disk self-occultation,
527: the reprocessed emission should have about the same intensity in either the 
528: Main or the Short High state peaks since the difference in projected area 
529: for an observer elevation ($5^\circ$) is only a few percent 
530: (viewing angles of $16^\circ$ versus 
531: $6^\circ$ for the Main and Short High states respectively). 
532: However,  disk self-occultation is important: during the
533: Short High state a sizeable fraction of the inner disk edge is blocked 
534: from the observers view, whereas a more open view exists during the Main
535: High state (see \cite{sco00}).
536: This situation is consistent with the ratio of $\sim0.25$ for
537: the observed peak of the Short High state blackbody emission (the current
538: observation and \cite{oos00}) to
539: the peak of the Main High state blackbody emission (\cite{dal98}).
540: 
541: The Short High state EUV emission shows sharp dips at different orbital phases. 
542: The dips may be due to the accretion stream. The stream moves in synchronism 
543: with the binary, rather than at the 35-day period, up until the point where it 
544: impacts the disk. The dips could also be due to the splash at the impact point 
545: which is periodic at the binary period but has a complicated path over the 
546: disk surface which depends in detail on the disk tilt and twist.
547: Above we saw that the Short High state EUV emission is blocked from the
548: observer's line-of-sight after orbital phase 0.75, like the X-ray emission.  
549: Feasible explanations are blockage of the line-of-sight
550: by the accretion stream or disk impact point splash region. 
551: The splash region subtends a much larger angle at the neutron
552: star than the stream so is a better candidate for the extended blockage
553: after orbital phase 0.72. It is highly desirable to obtain
554: further observations, such as for later orbits in the Short High state in
555: both EUV and X-ray bands, to test the origin of the dips.
556:                     
557: \acknowledgments
558: 
559: This work was supported in part by the Natural Sciences and Engineering 
560: Research Council of Canada. The authors thank Robert B. Wilson for
561: providing help with the analysis of the BATSE data used in this paper.
562: 
563: \begin{thebibliography}{}
564: 
565: \bibitem[Becker et al. 1977]{beck77}
566: Becker, R., Boldt, E., Holt, S., Pravdo, S., Rothschild, R., Serlemitsos,
567: P., Smith, B., \& Swank, J. 1977, \apj, 214, 879 
568: 
569: \bibitem[Boroson et al. 1997]{bor97}
570: Boroson, B., Blair, W. P., Davidsen, A. F., et al.  1997, \apj, 491, 903
571: 
572: \bibitem[Brainerd \& Meszaros 1991]{bra91}  Brainerd, J., \& M\'esz\'aros, P.,
573:  1991, \apj, 369, 179
574: 
575: \bibitem[Cheng et al. 1995]{cheng95} 
576: Cheng, F. H., Vrtilek, S. D. \& Raymond, J. C. 1995, \apj, 452, 825
577: 
578: \bibitem[Choi et al. 1994]{cho94}
579: Choi, C., Dotani, T., Nagase, F., Makino, F.,  Deeter, J. \& 
580: Min, K. 1994, ApJ, 427, 400
581: 
582: \bibitem[Choi et al. 1997]{cho97}  Choi, C., Seon, K., Dotani, T.,
583: \& Nagase, F. 1997, \apj, 476, L81  
584: 
585: \bibitem[Crosa \& Boynton 1980]{cro80} Crosa, L., \& Boynton, P. E. 1980, 
586:  \apj, 235, 999
587: 
588: \bibitem[Dal Fiume et al. 1998]{dal98}
589: Dal Fiume, D., Orlandini, M., Cusumano, G., et al. 1998, \aap, 329, L41
590: 
591: \bibitem[Deeter et al. 1976]{dee76}
592: Deeter, J., Crosa, L., Gerend, D., \& Boynton, P. 1976, \apj, 206, 861
593: 
594: \bibitem[Deeter et al. 1991]{dee91}
595: Deeter, J., Boynton, P., Miyamoto, S., Kitamoto, S., Nagase, F.,
596: Kawai, N. 1991, \apj, 383, 324
597: 
598: \bibitem[Deeter et al. 1998]{dee98} Deeter, J., Scott, D. M., Boynton, P.,
599: et al. 1998, \apj, 502, 802
600: 
601: \bibitem[Gerend and Boynton 1976]{ger76}
602: Gerend, D., \& Boynton, P. 1976, \apj, 209, 652
603: 
604: \bibitem[Leahy 1995]{lea95}
605: Leahy, D.A. 1995, \apj, 450, 339
606: 
607: \bibitem[Leahy 2000]{lea00}
608: Leahy, D.A. 2000, \mnras, accepted
609: 
610: \bibitem[Leahy \& Scott 1998]{lea98}  
611: Leahy, D., \& Scott, D. M. 1998,  \apj, 503. L63 
612: 
613: \bibitem[Leahy \& Marshall 1999]{lea99}
614: Leahy, D. A., \& Marshall, H., 1999, \apj, 521, 328
615: 
616: \bibitem[Lubow \& Shu 1975]{lub75}
617: Lubow, S., \& Shu, F. 1975, \apj, 198, 383
618: 
619: \bibitem[Malina et al. 1993]{mal93}
620: Malina, R. et al. 1993, AJ, 107, 751 
621: 
622: \bibitem[Mavromatakis 1993]{mav93}
623: Mavromatakis, F. 1993, \aap, 273,147
624: 
625: \bibitem[McCray et al. 1982]{mcc82}
626: McCray, R., Shull, M., Boynton, P., Deeter, J., Holt, S., White, N. 
627: 1982 \apj 262, 301 
628: 
629: \bibitem[Oosterbroek et al. 1997]{oos97}
630: Oosterbroek, T., Parmar, A., Martin, D., \& Lammers, U. 1997 \aap 327, 215
631: 
632: \bibitem[Oosterbroek et al. 1998]{oos98}
633: Oosterbroek, T., Parmar, A., Martin, D., \& Lammers, U. 1998 \aap 329, L41 
634: 
635: \bibitem[Oosterbroek et al. 2000]{oos00}
636: Oosterbroek, T., et al. 2000, A\&A, 353, 575 
637: 
638: \bibitem[Reynolds et al. 1997]{rey97}
639: Reynolds, A., Quaintrell, H., Still, M., Roche, P., Chakrabarty, D. \&
640: Levine, S. 1997, \mnras, 288, 43 
641: 
642: \bibitem[Rochester et al. 1994]{roc94}
643: Rochester, G., Barnes, J., Sidher, S., Sumner, T., Bewick, A., Corrigan, R.,
644: \& Quenby, J. 1994 \aap, 283, 884
645: 
646: \bibitem[Schandl \& Meyer 1994]{sch94} Schandl, S., \& Meyer, F. 1994,
647: \aap, 289, 149
648: 
649: \bibitem[Scott 1993]{sco93} Scott D. M. 1993, PhD Thesis, University 
650: of Washington
651: 
652: \bibitem[Scott \& Leahy 1999]{sco99} Scott, D. M., \& Leahy, D. 1999, 
653: \apj, 510, 974
654: 
655: \bibitem[Scott, Leahy \& Wilson 2000]{sco00} Scott, D. M., Leahy, D., 
656: \& Wilson, R. 1999, \apj, accepted (astro-ph/0002327)
657: 
658: \bibitem[Shakura et al. 1998]{sha98} Shakura, N., Postnov, K., \&
659: Prokhorov, M. 1998, \aap, 331 , L37
660: 
661: \bibitem[Shulman et al. 1975]{shul75} Shulman, S., Friedman, H., Fritz, G.,
662: Henry, R. C., \& Yentis, D. J. 1975, \apjl, 199, L101
663: 
664: \bibitem[Still et al. 1997]{sti97}
665: Still, M., Quaintrell, H., Roche, P., Reynolds, A. 1997, \mnras, 292, 52 
666: 
667: \bibitem[Voloshina et al. 1990]{vol90}
668: Voloshina, I., Lyutyi, V., \& Sheffer, K. 1990,
669: Sov. Astron. Lett. 16, 257
670: 
671: \bibitem[Vrtilek et al. 1994]{vrt94}
672: Vrtilek, S., Mihara, F., Primini, A., et al. 1994, \apj, 436, L9
673: 
674: \bibitem[Vrtilek \& Cheng 1996]{vrt96}
675: Vrtilek, S. D., \& Cheng, F. H.  1996, \apj, 465, 915
676: 
677: \end{thebibliography}
678: 
679: \newpage 
680:  
681: \figcaption[fig1.ps]
682: {Observed EUVE DS lightcurve of Her X-1 (with $\pm1\sigma$ error bars)
683: at the beginning of Short High state during July 25-29, 1997:
684: top panel- plotted as a function of time (lower x-axis) 
685: and 35-day phase (upper x-axis); 
686: bottom panel- expanded scale and plotted as a function of time (lower x-axis) 
687: and orbital phase (upper x-axis).
688:  }
689: \label{fig.1}
690: 
691: \figcaption[fig2.ps]
692: {Top panel:
693: Observed EUVE SW bright phase spectrum (horizontal bars indicating channel 
694: bins, with vertical error bars) with best-fit blackbody model
695: spectrum (histogram). Bottom panel: residuals between the data and model.}
696: \label{fig.2}
697: 
698: \figcaption[fig3.ps]
699: {Folded light curve at the best-fit pulse period.
700: %Top panel: for the first binary orbit after Short High turn on (Orbit 1); 
701: %middle panel: for the second binary orbit after Short High turn on (Orbit 2);
702: %bottom panel: for the combined data (Both orbits).
703: }
704: \label{fig.3}
705: 
706: \figcaption[fig4.ps]
707: {Illustration of the tilted-twisted disk. 
708: The view is from $5^o$ above the orbital plane
709: at 35 day phase 0.06. The rings drawn on the disk surface have thickened lines 
710: for the portion of the disk closer to the observer than the neutron star 
711: (near side of disk); the shaded part of the disk is the bottom side of the disk.
712: }
713: \label{fig.4}
714: 
715: \figcaption[fig5.eps]
716: {Shadow on the celestial sphere by the twisted accretion disk, as viewed from the
717: neutron star. The limb of HZ Her is also indicated at an arbitrarily chosen 
718: disk azimuth ($90^\circ$) of the line-of-centers.}
719: \label{fig.5}
720: 
721: \figcaption[fig6.ps]
722: {Relative reflected intensity from HZ Her as a function of the accretion disk
723: phase, for HZ Her seen at orbital phase 0.5. The letter A indicates the phase that
724: the disk is most open to HZ Her. B, C and D indicate disk phases of the times of
725: orbital phase 0.5 during the EUVE observations in 1997 (B) and 1995 (C and D).
726: The asterisks mark the observed EUV intensities.
727: }
728: \label{fig.6}
729: 
730: \figcaption[fig7.ps]
731: {Observed EUVE DS lightcurve of Her X-1 (with $\pm1\sigma$ error bars)
732: at turn-on of Short High state at JD =2450656.9. The smooth curve is
733: a fit to the GINGA 1989 Short High lightcurve plotted against 35-day
734: phase.
735: }
736: \label{fig.7}
737: 
738: \end{document}
739: 
740: 
741: 
742: