Back to the GRB Research page

Prompt Emission GRB Simulations, the University of Wyoming GRB Program, and Image Reduction

S. L. Savage
Graduate Student
Dept of Physics & Astronomy, UW

Advisor: J. P. Norris
NASA GSFC


Abstract

One of the most recent and rapidly-developing areas of research in astrophysics is gamma ray burst (GRB) study because of the mysterious nature of GRBs and the implications they could have on cosmology, theories surrounding stellar evolution, and other controversial investigations. Despite efforts from a growing GRB research field, details describing the true nature of such phenomena have yet to be convincingly put forth; however, one of the most successful models thus far is the collapsar model involving a massive dying star and the birth of a black hole – the focus of my research. My primary research project for the summer was to begin utilizing a 3D relativistic hydrodynamic code to simulate GRB jet development arising from such a scenario in order to predict and understand their lightcurves, time evolution, and magnetic field connection. Complementing this project were two others: preparing a poster for a GRB conference and developing an image reduction code. The poster describes the efforts being pursued by the University of Wyoming’s GRB afterglow follow-up program to rapidly react to, locate, obtain data from, and monitor GRBs using two upgraded observatories and three new IR instruments: two imagers and one spectrograph. The image reduction code was needed to calibrate one of these new instruments, the Goddard InSb IR camera, for consequent availability to the GRB program. Due to the extended nature of these projects, none have been completed but all are progressing successfully and will be continued alongside my graduate studies for possible future thesis work.


Gamma Ray Burst Prompt Emission Simulations

Gamma ray bursts (GRBs) are the most energetic phenomena in the universe with the ability to outshine an entire galaxy one-thousand fold! As indicators of the deaths of the most massive stars, they provide insight into the creation of blackholes from stellar progenitors. This exciting field, however, is barely emerging from its infancy for it was a mere thirty-six years ago in 1967 that GRBs were accidentally discovered by the Vela satellites which had been deployed to monitor illegal atomic bomb tests in space during the Cold War.

Unfortunately, those satellites only revealed their existence. What they are and from where they originated were questions left completely unanswered. Several attempts were made to localize them and models describing their origin abounded encompassing the belief that they were local to our universe. This belief was so popular because in order for them to be extragalactic, their source must be of astounding energies, on order of 1050 ergs or more – a difficult leap of faith.

A great debate ensued thereafter led by Don Lamb and Bohdan Paczynski over the distance scale of these explosions. Paczynski was convinced of their cosmological origins while Lamb, and most of the astronomical community, insisted that they were linked to the Milky Way and its local surroundings. The argument in favor of extragalactic distances was largely based on the randomness of the distribution, but at that time, there was no way to test this theory because once the burst was gone, all that was left was a lightcurve which tells nothing of its origin or distance.

The introduction of the Italian BeppoSAX mission changed all that though. It was eventually able to detect a burst fast enough and was designed in such a way as to be able to localize it to within an error box so that ground based telescopes could slew to that patch of sky and try to find something new. After several failed attempts, the afterglow of a burst was first detected and followed on February 28, 1997. This marked a new era in the field of GRBs! Amazingly enough, the location of the afterglow coincided with a distant galaxy which added weight to the cosmological distance theory.

The nail in the coffin, though, came on May 8, 1997. Another GRB was detected and telescopes around the world frantically began searching for the afterglow which fades very rapidly and is usually not very bright. The search was not in vain for the optical transient (i.e. afterglow, OT) was detected and its spectrum was actually measured with the 10-m telescope in Hawaii which reveled that its redshift was at least 0.835. In other words, it was more than 6 billion lightyears away!

Every question answered adds at least ten more questions unsolved. Even though the distance riddle has been solved, many more questions remain. Two prevailing ones are what are they? and what could possibly create so much energy as to rival the Big Bang?

There are many competing theories trying to explain the origin of classical long GRBs – all generally involve the collapse of a massive star into a blackhole. The collapsar model, which has claimed the most popularity among scientists, can actually be subdivided into two scenarios.

I) The iron core of a massive helium star collapses, becomes a neutron star, and then a blackhole. The matter surrounding the core is slowed and an accretion disk forms around the blackhole due to angular momentum conservation and other factors and promptly falls into the blackhole. Consequently, jets of radiation and material moving near the speed of light out of the core result from the infalling matter and run into the outer layers of the star’s atmosphere and other material in the immediate vicinity creating shock waves which are seen as the burst and the afterglow. The supernova reaches its peak weeks after the GRB.

II) The core of a massive helium star collapses into a neutron star; however, there is not enough energy to eject the outer layers which results in that matter falling back onto the neutron star. The mass of the neutron star eventually reaches its limit to end cataclysmically as a blackhole. An accretion disk and jets form as in the first case and cause extreme shock waves. The supernova reaches its peak weeks after the GRB.2,5,9

There is also another proposed scenario (supranova model) in which the supernova actually occurs first and leaves a rapidly rotating neutron star whose mass is actually over the critical limit for forming a blackhole but whose angular momentum keeps it intact. Eventually, it loses enough momentum to reach the critical frequency and hence collapses. Once again, material accretes and jets form and pierce the surrounding medium.3 There are proposals to monitor supernovae in order to detect delayed GRBs. Observations thus far tend to favor the previous collapsar models – notably from the recent GRB 030329.8

Because of their possible very high redshifts, GRBs can be used for cosmological purposes by mapping out the Lyman-Alpha forest, indicating star formation rates, and providing insight into the large-scale structure of the universe – holy grails for the cosmologist.46

In order for this to be successful, however, GRB lightcurves must be understood in more detail through the use of simulations and observations. I was only able to make a few runs of the code because each run takes on order of days and much time had to first be put into understanding its procedures and developing a strategy for varying input parameters. Computer memory has also been an issue that will be resolved in the coming months. By becoming familiar with this code and its output, we are preparing for the influx of Swift low-energy burst data and high-energy burst data from GLAST. The following is a table of plots that shows some examples of the code’s output.



Table 1: The above table of images shows the output of the 3D RHD code. The various parameters being plotted are noted along the y-axis. Each individual plot is a combination of the graphs from different time steps as the jet is allowed to evolve along the x-axis of a cylinder. Initial conditions: jet density = 1.0, jet pressure = 1.0, gamma = 1000, opening angle = 0, x velocity = 1.0, y velocity = 0, z velocity = 0, medium density = 1.0e-5, medium pressure = 1.0e-4. Notice how the gamma factor increases dramatically in the forward shocked region.



The next progressive step for the code will be to incorporate magnetohydrodynamics to determine an appropriate magnetic field strength.


The University of Wyoming GRB Afterglow Follow-Up Program

The Interplanetary Network (IPN) of satellites has been set up so that once a burst is detected, ground-based telescopes all over the world rush to detect the OT. The University of Wyoming’s Physics and Astronomy department is now involved with this collaboration! While running the RHD code, I was able to start designing a poster to be presented at the GRB conference in Santa Fe in September describing the University of Wyoming’s GRB afterglow follow-up program. The university owns two research observatories that are now involved in the program. Until recently, only Red Buttes Observatory (RBO) has been used to hunt these afterglows and add to the successful detection probability as well as to complement filter selections from other telescopes.

RBO is located only 19 km south of Laramie on a dark site which, combined with the dry, thin atmosphere (~ 2200-m elevation), enables relatively deep imaging with its Apogee AP8p 10242 CCD camera mounted to a 0.6-m Cassegrain DFM reflector. 10-minute exposures can yield typical limiting magnitudes of ~ 19.5 (19) for V and R (B). The large 18 arcmin FOV is especially useful for targeting GRBs and relaying the locations to WIRO for deeper, more sophisticated imaging.

Demands from involved undergraduate and graduate students along with faculty to use the facility have greatly increased for various scientific endeavors in addition to GRB afterglow follow-up study. To compensate, RBO has been extensively renovated and upgraded. Among the upgrades are refurbishment of the telescope platform and facility, acquisition of faster computers, establishment of a microwave link, and installation of a GPS clock, flat-field lights/screen, a weather station, and an all-sky camera for local weather monitoring. All of these improvements are directed towards the successful implementation and completion of making RBO completely automated so that it will respond to burst alerts without human intervention. Remote operation will also be available.

Our team has contributed to the observation of several bursts from RBO and has been documented in eight GCN archives. The most important one thus far occurred recently on March 29, 2003. It was the brightest burst to be found with an afterglow and is now known to be the second closest with a redshift of z = 0.1685!10 Our group observed this burst with the rest of the world for several nights with RBO. The light curve of this burst (GRB 030329) was unusual in that there were many “bumps” and rebrightenings indicative of an underlying supernova. Spectra taken of the OT with other telescopes strongly correlate with that of a hypernova (supernova involving a very massive star). Images of this burst as taken from RBO and its lightcurve are shown below in Figure 1.

Figure 1. Above right are images taken from RBO of the brightest and one of the closest bursts – GRB 030329, and the lightcurve is that of the same burst as measured from RBO. The results from RBO monitoring of this burst have been contributed and archived as GCN notices.

Now joining RBO in the hunt for afterglows is the Wyoming Infrared Observatory. WIRO, situated at 2944 m 40 km southwest of Laramie in semi-arid conditions, is an optimal site for optical and mid-infrared observations. The 2.3-m telescope is one of the largest with extensive availability to afterglow follow-up research. Without compromising much collecting power, its medium size allows for fast acquisition times compared to large telescopes such as Hawaii’s Keck Observatory. WIRO has been undergoing renovation similar to that of RBO with upgraded instruments and facility improvements. Recent additions are WIRO’s three primary instruments: WIRO-Prime, WIRO-Spec, and the Goddard IR camera.

WIRO-Prime is a 20482 prime-focus CCD camera with a 20 arcmin diameter FOV (f/2.1). A 5-minute exposure reaches point sources as faint as 24th magnitude in V in 1-arcsec seeing. WIRO-Spec is an integral field, holographic spectrometer which utilizes Volume-Phase-Holographic gratings with a 20482 CCD detector. A 15 X 20 array of 1 arcsec optical fibers will allow simultaneous spectroscopy over an equivalent region on the sky for rapid follow-up spectroscopy of GRB afterglows. The high system efficiency (~ 40%) should enable us to reach S/N ~ 10 for a 20th mag point source in a 10-min exposure with a typical resolution of ~ 5 Angstrom. The Goddard IR Camera is a 2562 InSb camera (FOV ~ 108 arc sec) mounted at Cassegrain and operated at ~ 15K. Available filters for GRB observations include R, I, J, H, and K’. Beyond GRB study, WIRO is now operational and available to the faculty and the renewed graduate student program for various scientific interests such as quasars, cataclysmic variables, globular clusters, etc. For efficiency the instrument in use at the time of a burst alert will be the one planned for use to promptly pursue afterglows with imaging or spectroscopy.

Because of the transient nature of bursts, GRB follow-up research must be performed on a target-of-opportunity basis. Both observatories, within an hour of Laramie at the furthest, are actually owned by the university making swift and numerous prompt observations more accessible as well as long- term follow-up monitoring feasible. Short bursts have yet to be associated with optical counterparts due to their predicted steep light curves and intrinsically low initial luminosity.1 Acquiring light curves for short bursts with the rapid reaction capabilities of RBO would be a tremendous contribution. Beyond imaging, however, spectroscopy is available for determining redshifts of hi z bursts – salient information for several models and research goals as well as an effective method for comparing bursts to supernovae for determining burst properties. GRBs are predicted to occur at considerable rates out to z &ge ~ 5 thus making them primary candidates as probes for cosmology in the study of the Lyman-&alpha forest, the epoch of reionization, the evolution of metallicity, and large-scale, high-redshift structure.4,7

To complement the GRB response from Swift, the combined efforts of the two research observatories at the University of Wyoming will enable rapid reaction to burst alerts, offer a large FOV for targeting bursts, and provide wavelength coverage ranging from optical through mid-infrared. Automated response of RBO, with a slewing rate comparable to Swift’s, coupled with a large 18 arcmin FOV increases the chances of targeting a burst in its early afterglow emission stages and possibly detecting the elusive short burst afterglow – a feat yet to be accomplished. The large collecting power of the 2.3-m WIRO telescope and mid-infrared optimization ensure deep imaging in wavelengths not seen by Swift; nevertheless, WIRO and RBO will be able to overlap with Swift’s data in the optical regime. The team collects data in collaboration with the Follow-Up Network (FUN) GRB group in primary affiliation with Dan Reichart of the University of North Carolina, Chapel Hill. WIRO will be able to continue the lightcurve analysis beyond the capabilities of RBO and possibly reveal host galaxies and star formation rates along with redshifts. The success of the GRB program will expectantly revitalize and promote other programs within the department.


Image Reduction

Prior to beginning my summer internship, plans were made for calibrating one of the new imaging instruments used at WIRO, the Goddard InSb IR camera. Following its recent development, the dewar/camera system has been used to observe edge-on galaxies to study their morphological characteristics; however, data reduction for the camera had been minimally performed. In order to make the camera available to the UW GRB group, I took on the image reduction/analysis project. Writing procedures in IDL accounted for a vast majority of my 10-week stay. Example results are shown in the following figures.


Figure 2: Top left is the original, unaltered FITS image taken of NGC 5746 with the Goddard InSb IR camera using the I filter. A histogram of its intensity values (mV) is given to its right. The bottom figures are that of the dark-subtracted FITS image. Note that the histogram is shifted to eliminate the offset due to dark current in the original image. The image appears the same because the values have only been shifted – not rescaled.


Figure 3: Above left is the masked, flattened version of the original FITS image from Figure 2. The mask eliminates all pixels outside the field of view and uses a threshold routine to replace hot or dead pixels with an average of the surrounding intensity values. The threshold routine’s success is filter dependent. The above right image, obtained from NED, is an image of the same galaxy, NGC 5746, taken with the B filter. Imaging in infrared wavelengths reduces blockage of the nucleus by dust obscuration.


Calibrating the instrument to obtain magnitudes of point sources and extended objects followed image reduction. Unfortunately though, only one standard IR star was available for useful calibration and its standard magnitude values overlap with only two of the filters used to image it – J and K; therefore, only the J and K filters were able to be calibrated with any confidence. Below are examples of the aperture photometry methods used to obtain magnitudes.


Figure 4: Top left is the reduced image of standard IR star, p266-c, taken in the J filter. This star was used to calibrate the J and K filters. As a check of the code, the procedure to find magnitudes was used on p266-c. The boxes around the star in the top image correspond to the vertical lines in the bottom plot which indicate the inner and outer apertures used to separate the signal from the surrounding background. The outer aperture allows a background signal to be calculated in order to subtract that from the star’s point spread function within the inner aperture. The conversion method for determining magnitudes uses a mV – Flux conversion determined from the standards using a calibration procedure. The comparison method uses direct comparison of the signal with that of a standard.


Figure 5: The above images are that of the galaxy NGC 5775 taken with the J filter and use the same principle described in Figure 4 except that the source is an extended object and the aperture must encompass it. The bottom plot depicts the smaller box in the top figure and is used to determine the background signal.

Because of the lack of standard calibration stars, the error analysis is poor. The next steps to take in the image reduction will be to incorporate photon statistics, perform more sophisticated error analysis, determine signal to noise ratios, map coordinates, register and stack, calculate filter ratios, and possibly create color images.

This project, along with the other two, will be continued at the University of Wyoming.




Acknowledgements: I would like to thank Drs. J. P. Norris, R. Canterna, A. Kutyrev, and M. Pierce for their advice and encouragement as well as USRA for the internship opportunity.


References

1Bromm, V. & Loeb, A. 2002 ApJ, 575, 111

2Galama, T.J. et al. 1998 Nature, 395, 670

3Konigl, A. & Granot, J. 2002 ApJ, 574, 134

4Lamb, D.Q. & Reichart, D.E. 2000 ApJ, 536, 1

5MacFayden, A.I. et al. 2001 ApJ, 550, 410

6Norris, J.P. 2003 Astrophysics of GW Sources Conference proceedings

7Panaitescu, A., Kumar, P., Narayan, R. 2001 ApJ, 561, L171

8Willingale, R. et al. 2003 astro-ph/0307561

9Woosley, S.E. & Eastman, R.G. 1999 ApJ, 516, 788

10GCN 2020



Back to the GRB Research page