These endeavors have been both productive and fun, in large part due to my longtime collaborator in much of this work, Paul Roche. Since then, my 5.2 years at Caltech have been made much easier by the wonderful friends I have met here.
List of Figures
List of Tables
Note on Terms, Units, and Names
Chapter 1
Accretion-Powered Pulsars and BATSE
Overview and Project History
- Accretion-Powered Pulsars and the Role of BATSE
- Thesis Organization
The good (1.024 s) time resolution enables coverage of most of the pulsation phase space (P > 2 s) of known accretion pulsars. Interestingly, the spin-up and spin-down magnitudes differ by only 15%, with the neutron star spin varying on a time scale of Iv /vi ~ 5000 years in both states.
Background on Accretion-Powered Pulsars
- Early History 1
- Observed Characteristics
The emission mechanism in these spin-powered isolated pulsars is completely different from that of accretion-powered binary pulsars. A rough correlation between orbital period and rotation period has been observed in Be-star/X-ray binary pulsars ( Corbet 1986 ; Waters & van Kerkwijk 1989 ).
Chapter 2
Acquisition and Reduction of BATSE Data
The Compton Gamma Ray Observatory
Modules 0, 2, 4 and 6 are visible in this diagram. al. 1993) provides imaging observations in the 1–30 MeV range using a combined liquid/Na scintillator. The Energetic Gamma-ray Experiment Telescope (EGRET; Thompson et al. 1993) provides imaging in the 20 MeV–30 GeV range using a spark chamber detector.
The Burst and Transient Source Experiment
We also use the QUAL information to exclude intervals containing gamma-ray bursts from pulsar timing analysis. A variety of special data products are activated with high time and energy resolution when triggered by on-board identification of a gamma-ray burst in the BATSE LAD data stream.
Detector Response
The response does not drop to zero at 90° incidence due to the finite thickness of the crystal. The secondary maximum in the large-angle response for low energies is due to the geometry of the support structure at the edge of the LADs.
Optimal Combination of Detectors
The best overall sensitivity is obtained by adaptively choosing the detector weight based on the altazimuth position of the source. Celestial coordinates are shown in the altazimuth system with respect to the spacecraft axes. e) singles+poirs (f) singles+poirs+quods.
Earth Occultation
A source is obscured when the line-of-sight shock parameter relative to Earth is less than RtJJ + h. Given a position vector r = (x, y, z), the equatorial radius of the equipotential passing through that position is given by z.
Detector Background
It is instructive to consider the background power spectral properties of the BATSE in the frequency domain. The bottom curve is for the raw data after subtraction by Rubin et al.
2. 7 Sensitivity to Pulsed Signals
- General Principles
- BATSE Sensitivity as a Function of Energy
- Time Systems and Reference Frames
- Standard Analysis Procedure
- Chapter 3
The second panel shows the time series after subtraction of the ad hoc background model. The detector response matrices of the Burst and Transient Source Experiment (BATSE) at the Compton Gamma Ray Observatory.
Detection and Estimation of Periodic Pulsed Signals
- Introduction
- Detection of Periodic Pulsed Signals
- Time Domain: Epoch Folding
- Frequency Domain: Fourier Analysis
- Aliasing and Pulsed Sensitivity in Binned Data
- Estimation of Pulse Strength in Periodic Signals
- roo 2 n2 exp(~) (-a2)
- Flux and Spectral Estimation
- Chapter 4
Due to the finite length of the time series T, the power spectral response in the kth Fourier frequency bin (where Vk = k/T) to a periodic signal with frequency vo is given by. This function is plotted in the top panel of Figure 3.3 for various values of the true. For small values of the measured amplitude (a/n;::; 2), the distribution is highly skewed towards zero signal.
Discovery of the Orbit of the Accreting X-Ray Pulsar OAO 1657-415*
- Introduction
- Observations and Timing Analysis
- Results
- OAO 1 65 7 - 4 1 5 100
- Discussion
- Chapter 5
More recent observations have shown up-and-down rotation during the pulse period (Kamata et al. We have measured the eccentric binary orbit and eclipse duration of the accretion-driven X-ray pulsar OAO 1657-415. SIGMA observation of the pulsar OAO 1657-415: precise localization at hard X-ray energies rays and the discovery of spin down.
Discovery of the 18.7-Second Accreting X-Ray Pulsar GRO J1948+32*
- Introduction
- Observations and Analysis
- Discussion
- Chapter 6
GRO J1948+32 was initially detected in a routine search for the Fourier power spectra of this data for 1994 April 7. correspond to the eruption. The vertical bars show the 1 a statistical uncertainties, while the horizontal bars show the widths of the energy channels.
- Introduction
- Observations and Analysis
- Pulse Timing
- Phase Residual Analysis
- Pulse Spectroscopy
- Discussion
- Chapter 7
The resulting power spectral density of the pulse frequency fluctuation Pq is shown in the left panel of Figure 6.3. It is therefore puzzling that the magnitudes of up-spin and down-spin in 4U 1626-67 are almost the same. These contrasts may simply reflect differences in the stability of accretion flows in these systems.
Introduction
In the centrifugally driven wind (CDW) model of Arons et al. 1984), the magnetic field does not cross the disc but penetrates a narrow boundary layer. As an alternative, Makishima et al. suggested that the spin-down may be due to accretion from the dense, slow wind of the M giant after disruption of a previous accreting disc during spin-up. Some authors have claimed that this X-ray burst was accompanied by increased Ha emission (Manchanda et al. 1995;
7 .2 Observations and Analysis
Timing
34"red noise" component (a power spectral component that increases with decreasing frequency) in the pulse frequency fluctuations. Before calculating the power spectrum, we multiplied the pulse frequency residuals by a window function of the form Wj = cos4(j7r/N) with j = -N/2, · · ·, N/2, substantially suppressing the side response of the Fourier transform (Harris 1978) The resulting power spectral density of the pulse frequency fluctuations Pv is shown in the left panel of figure 7.5.
Pulse Profiles and Spectroscopy
A simultaneous measurement of the total (pulsed+non-pulsed) spectrum of GX 1+4 in the range 40-200 keV was performed with Compton/OSSE (Staubert et al. 1995). Furthermore, a head-to-head comparison of BATSE and OSSE observations of the Crab Nebula has shown that the BATSE fluxes are systematically::::::20% higher than the OSSE fluxes (Much et al. 1996). The dotted curve shows the best-fitting model for simultaneous OSSE observations of the total (pulsed+non-pulsed) flux, with corrections applied (see text for details).
Flux and Torque
The resulting history of the derivative of the pulse frequency (i.e. the spin-up rate), which is proportional to the net torque applied to the neutron star if we can neglect the orbital Doppler shifts, is shown in the middle panel of Figure 7.9. Clearly, most of the bright flares were accompanied by enhanced spin-down. We can formalize the correlation between torque and brightness by calculating the cross-correlation of the two time series.
7 .3 Discussion
Accretion
Both the MTD and CDW accretion models predict that a sufficiently increased mass transfer rate can cause a transition from spin-down to spin-up, consistent with our observations of the persistent bright state around MJD 49600. For consistency with the other data, we expect at least an initial improvement in spin-down torque. We can put a lower limit on the distance and X-ray luminosity from the steady spin-up of the source during the 1970s (see Appendix C).
Appendix A
Spacecraft Coordinates and Sky Coordinates
The vectors for the BATSE detector normals in spacecraft coordinates are given in Table A.l. The celestial coordinates for the spacecraft axes for each of the Compton Observatory paintings are given in Table A.2. Coordinates for more recent references are available from the Compton Observatory Science Support Center via WWW (http://cossc.gsfc.nasa.gov/cossc/cossc.html).
Appendix B
Decoherence Time Scales in Pulse Timing
Decoherence Due to Accretion Torque
For longer observations, phase coherence can be recovered by "accelerating" the time series to compensate for phase drift, essentially stretching or squeezing the size of the time bins (Middleditch 1989; Anderson et al. 1990; Wood et al. 1991; Johnston & Kulkarni 1991) .
Decoherence Due to Orbital Motion
Appendix C
Limits on Luminosity and Distance from Steady Spin-Up
Appendix D
Binary Orbits
Basic Concepts
The orientation of the line of apsides with respect to the line of sight is specified by the longitude of periastron w. By using the Doppler shifts of a binary pulsar to measure an orbit, we are only sensitive to motion projected along the line of sight. If sin i is sufficiently close to 1 to allow an eclipse of M2 at M1, then L = goo will roughly correspond to the center of the eclipse (see Deeter et al. 1g31 for some subtleties related to this point) .
Orbit-Fitting Equations
- Auxiliary Chain Rule Quantities
- Pulse Frequency Measurements
- Pulse Arrival Time Measurements
If one uses the alternative parameterization for low-eccentricity orbits, the partial derivatives with respect to T0 are replaced by those with respect to Trr ; 2,. A slightly different set of derivatives will be needed if the circuit is parameterized in terms of T1f;2 instead of To. A slightly different set of derivatives will be needed if the circuit is parameterized in terms of T1r ;2 instead of To.
Appendix E
Eclipse Constraints in OAO 1657-415
The coordinate system rotates about the z-axis in the ¢, one revolution per orbital period, so that the companion always lies on the x-axis with the neutron star at the origin. The line of sight from the neutron star to the observer will intersect the companion surface on the shown closed curve. However, because for an eccentric orbit the coordinate system no longer rotates at a uniform speed, we no longer have Oe =¢max· To relate the observed quantity Oe to equation (E.7), we can again exploit the location of periastron. halfway through the eclipse (¢ = 0).
Appendix F
Limits on the Orbit of GRO J1948+32
By doing this for a grid of fixed parameter values, we can calculate the allowed region of parameter space corresponding to a given confidence level by excluding parameter values for which x2 is also . Model with reduced x2 above this value is inconsistent with the data at the 953 confidence level. The corresponding results for eccentric orbit models are summarized in Table F.2 and shown in Figure F.3.
Appendix G
BATSE Localization of Faint Pulsed Sources*
Since Wis is dominated by harmonics of the spacecraft's orbital frequency VGRO « 2vo (where 1/vGRO ~ 93 min), we can generally assume IW2kl. The factor of E in the denominator of this definition appears because a fraction 1 - E of the time series { Q j} is set to zero.). Given the strength of the signal in these data, we will adopt 0.5° as the systematic precision limit of this technique with DISCLA data.
Appendix H
Introduction
Observations
The fundamental and second harmonics of the optical pulses, seen near 0.13 Hz and 0.26 Hz, are in good agreement with the BATSE pulse frequency ephemeris. Absolute time accuracy within ± 1.5 ms UTC-NIST was maintained using a Kinemetriks/Truetime model 468-DC satellite-synchronized clock in conjunction with the GOES-West satellite operated by the National Oceanographic and Atmospheric Administration.
Results
When corrected to the barycenter of the solar system, this agrees well with the BATSE epmeheris for X-ray pulsations (see Chapter 6). In addition to the main pulsation, a significant lower frequency sidelobe is also present at the mHz topocentric frequency. The mHz main peak to sidelobe spacing is almost identical to the sidelobe spacing reported by Middleditch et al.
Appendix I
IAU Circulars
Position of GRO J1948+32
Appendix J
Photon spectrum and period evolution of GX 1+4 as observed at hard X-ray energies by SIGMA. Timing of X-ray pulsars from data obtained with the ART-P telescope of the Granat Space Observatory in 1990-1992.
Colophon
