Printable Version 

Norfolk State University Astronomical Observatory

Summary

Norfolk State University (NSU) plans the construction of an Astronomical Observatory, housing a Rapid Response Robotic Telescope (RRRT), to enhance its participation in astronomical research and education. This project will increase diversity in the astronomical community and provide astronomy outreach for the Hampton Roads area (most populated in Virginia). The RRRT will be an important complement to an Astronomy education initiative at NSU directly to improve the offerings in a recently created Minor in Astronomy. It will also enhance our general public outreach trough collaborations with local public school systems and local amateur groups.  

We are planning the acquisition of a 0.6 cm (24”) RC telescope with equatorial fork mounting. It will be equipped with a CCD camera for UVBRI photometry and polarimetry. It will be robotic and fully automatic and controllable trough the web, allowing its operation from NSU’s campus and from the Hampton Roads area. 

For the operation of the RRRT we are establishing a close collaboration with the Department of Astronomy of the University of Virginia, Charlottesville VA. The proposed RRRT will be located at the Fan Mountain Observatory, Blue-ridge Mountains, near Charlottesville. NSU’s RRRT research activities will be centered on the early photometry and polarimetry of Gamma-ray Optical Afterglows (GRB-OA). This telescope will be part of an early-detection network directed by GSFC-NASA, tracking data produced by NASA’s Swift and GLAST missions. Intermediate-sized scopes with low inertia (fast slew), as the proposed here, will play a crucial role in future GRB observing programs.  Moreover, polarimetry capability on the ground is important to study the physics of the electromagnetic nature of the GRBs. The study of GRB-OA provides a natural evolution for our current high-energy astrophysics program, introducing more traditional observational astronomy that NSU would like to develop on the curriculum. 

The RRRT will also be used for other research and educational projects. Undergraduate students, from several science disciplines at NSU, will participate in research and students of introductory courses will be able to enhance their education thought direct astronomical observations. Another important educational component will target local high school science teachers and students. The Virginia Beach Public School System (VBPS), the largest in Virginia, has recently introduced an astronomy course in its curriculum (at the senior year). We plan a close collaboration with the VBPS to allow those students to use the RRRT. Moreover, a local amateur astronomer group is also interested in using the RRRT for Near Earth Objects (NEO), Variable Stars studies and Supernova searches; this group (the Back Bay Amateur Astronomers) will make for an important segment of the RRRT users.  

Through this program, NSU faculty and undergraduate students will directly participate in the design and construction of hardware and software in the forefront of observational astronomy. We plan to involve a minimum of two NSU undergraduate students during the construction phase. NSU is one of the leading HBCU in the country and its main campus is located at about 130 miles from Fan Mountain. This proposal will strengthen the national participation of undergraduate minorities in astronomical research. Throughout the use of this telescope we will make special efforts to attract historically underrepresented groups and to popularize the concepts of modern astronomy in the Hampton Roads area of Virginia.

A - Research Activities

The main research activity proposed for the Rapid Response Robotic Telescope (RRRT) is the photometric and polarimetric study of Gamma-ray Bursts Optical Afterglows (GRB-OA). We intent to measure UBVR and near Infrared magnitudes and polarization during the first few minutes after the gamma ray burst was detected by the Swift or GLAST missions.

Gamma-ray bursts are currently one of the most exciting topics in high-energy astrophysics, and perhaps of all of astrophysics. Their fundamental importance to cosmology and particle physics can be appreciated from the fact that GRBs are temporarily up to 10⁶ times more luminous in the optical than their host galaxies. They are detectable to the limits of the observable Universe.  Cosmologists appreciate the potential of the luminous GRBs to probe further back in space-time than even supernovae, to the first eras of star and galaxy formation [1]. Once their physical mechanisms are better understood, they may eventually surpass type Ia supernovae as the “standard candles” for probing high redshifts, thus helping to constrain the fundamental parameters in Cosmology [2], [3], [4]. The liberated energy in a GRB explosion and its afterglow phase can be extreme, assuming narrow jet opening angles, about 10^51-52 ergs in an interval of tens of seconds.            

The most popular class of models for the GRB phenomenon goes by the name “hypernova” – in analogy with the well-studied supernova classes ­– or “collapsar” [5], [6].  Such scenarios envision formation of a black hole near the time of supernova collapse, with the GRB following within seconds to a few days, the timescale depending on immediate or delayed accretion of residual debris from stellar collapse.  The burst emission is envisioned to result from ultrarelativistic particles that are accelerated in a narrowly collimated jet emerging from the spin axis of the nascent black hole.  We detect the cone of gamma-ray emission only if the jet axis is closely aligned to our direction.  Thus, in GRBs we may be seeing one of the most exotic processes in the Universe – the creation of a black hole.

GRBs are detected across the whole electromagnetic spectrum, with global alerts triggered by space-based gamma-ray and X-ray instrumentation.  In principle, transient alerts and positions, once telemetered to the ground, are distributed within seconds to ground observatories around the world operating in the optical, infrared, and radio bands. The GRB Coordinates Network (GCN) operated at GSFC is the center for GRB message distribution [7].

Space-based alerts provide ground observatories with a relatively large GRB localization error region – in the past, degree-sized; in the future with Swift, arc-minute (after 10–20 seconds) to arc-second-sized (after 1 minute).  The Swift’s Burst Alert Telescope (BAT) will be able to quickly determine the GRB position to within 4 arcmin radius and then distribute the information to the GCN within 12 seconds. The Swift’s XRT image coordinates (accurate to within 5 arcsec) would arrive within 62 seconds later for distribution via GCN. The time scale is critical, since some GRB afterglows are believed to decay below detectability in tens of seconds [8].  Without detection of the very accurate position of an optical/IR or radio afterglow, the host galaxy and redshift will not be found.

    Figure 1: R magnitude light curve of the GRB030329 [9].

Smaller scopes making the detection of a quickly decaying counterpart then pass the refined position to larger telescopes, which may observe the host galaxy’s redshift, even if the GRB itself has faded past the detection limit. Thus, intermediate-sized scopes with low inertia (fast slew) will play a crucial role in future GRB/OA observing programs.  Moreover, polarimetry capability on the ground is important to study the physics of the electromagnetic nature of the GRBs.   Phenomenally high polarization (~ 80%) has recently been reported in the prompt gamma-ray emission of one burst, and optical polarization of order 1–3% has been reported for several bursts. [10] Polarization measurements as a function of time throughout the afterglow are expected to help reveal the evolving structure of the fireball [11].

Of the roughly 190 GRB high-precision X-ray/gamma-ray positional alerts since 1997 (see J. Greiner’s website, http://www.aip.de/~jcg/grbrsh.html), only ~ 40 have led to discovery of optical afterglows and only about 24 have been reported to have radio afterglows.  These detection success rates have varied according to the capabilities of each particular spacecraft/instrument detection and alert system.  The bottom line is that only 32 redshifts have been associated with a GRB, either through the afterglow itself, or via the host galaxy. Thus a large fraction of GRB afterglows have gone undetected, with ~ 65% not seen at the longer wavelengths, and more than 80% with no redshift.

That most GRB sources have gone unidentified is due to a conspiracy of logistics, limits to present instrumentation, and the large dynamic ranges of GRB afterglow luminosities and decay time scales.  The optical/IR afterglow typically fades with a power-law dependence (see figure 1).  There is a distribution of power-law indices and we only know that part of the distribution for which afterglows are detected.  There could be a substantial tail in the distribution on the steeper, i.e., faster decay end. So just to detect the afterglow, rapid acquisition is required in some cases.

Also, from spectroscopic and image information, we understand that at least some GRB sources lie in obscured parts of their hosts – star forming regions – where optical extinction is high.  Many GRBs could be hidden in very high extinction regions, such that little optical emission penetrates to the exterior.  This argues for coverage into the infrared where extinction due to dust is significantly less than in the visible.  Most infrared telescopes are fairly large, with high inertia; hence, very rapid acquisitions in the optical with intermediate-size telescopes may serve to alert larger infrared facilities before decays beyond detection occur.  

Two separate classes of GRBs exist, long and short duration, with the transition near 2 seconds [12], [13].  Afterglow detection statistics are available for “long” bursts only (durations longer than ~ 2–3 s).  “Short” bursts do not trigger BeppoSAX, and only two short bursts have recently been detected by the IPN and reported to the Global Coordinates Network (GRBs 000326 and 000408: [14]).  No afterglow was found in either case.  One hypothesis, based on the tenuous statistics of these two non-detections, is that afterglow decay timescales are related to burst durations – total released energy – in which case very rapid acquisition on the ground will be essential for obtaining the redshifts of short bursts, assuming they also lie at cosmological distances? 

The proposed RRRT will fulfill a need for faster, more sensitive telescopes for acquiring GRB afterglows.  Implicit in the above discussion is that very large telescopes do not move fast enough to acquire GRBs when they are at their brightest, as we know from an example of one, GRB 990123 which reached 9^th optical magnitude during the burst [25]. The large inertia of large telescopes and their observing programs are not geared towards interruption to immediately observe these unexpected transients with power-law decays.  Since at gamma-ray energies some GRBs may last for only seconds to a few minutes, a network of ground scopes distributed in longitude/latitude is essential for efficient acquisition and follow-up studies 

The proposed RRRT will be located at Fan Mountain, Virginia [15] (latitude = +37.9^o and longitude = -78.69^o) and a height of about 800 ft. There are very few RRRTs in the USA East coast; the proposed telescope will fill this need for the GRB Coordinates Network. It will be also one of the few in making efforts to provide GRB-OA polarimetry during the very first minutes after the burst.

The proposed telescope will be directly connected to the GCN, such that a GRB detected over the telescope horizon will take priority over any other program, and the telescope will be ready in a few seconds to start photometry and polarimetry of the GRB-OA object. We expect to study afterglows up to 17-18 visible magnitudes.          

The GRBs themselves are brief, lasting from about 1 sec to about 100 sec. They are almost unbelievably energetic bursts of short wavelength radiation. The initially bright optical afterglows can last for days to months, decaying with power-law dependence in time [16]. The Swift mission [17] will be dedicated to detect GRBs, rapidly localize them in-orbit and relay this information to ground-based observatories in intervals of seconds. Swift is now scheduled to be launched by October 2004, and it is expected to be fully operational by January 2005. Swift localizes GRBs in stages (gamma rays to X‑ray to ultraviolet and optical) using a gamma-ray telescope with a wide field of view, and narrow X‑ray and optical telescopes with fine resolutions.  The proposed timescales associated with Swift are shown in table 1. Our telescope design has been done taking into account these parameters. The ground based rapid response telescopes will be most important during the first 150 sec (when the UVOT telescope in SWIFT is not activated).

Table 1: Swift’s time scales [17].

During the initial few tens of seconds, the optical afterglow begins to fade with power law dependence. The brightest optical emission occurs during the GRB, not hours after. The range of optical magnitudes at 100 s delay from GRB onset is of order 10^th to 20^th magnitude, accessible to a 24” telescope.  The prompt optical emission is millions of times brighter than the light of the host galaxy.  Larger observatories will be needed to study the host galaxy that is usually a very faint, > 23^th magnitude, highly redshifted galaxy.  In conclusion, earliest times (<100 sec) are the best to study GRBs directly.

As discussed before, invaluable physical insight into the GRB mechanism will be obtained by studying this prompt optical emission. Since a relatively coarse (4 arcmin radius) gamma-ray position will be provided by Swift and distributed in about 20 seconds to ground observers, these parameters determine the needed slewing and positioning precision.  Most GRBs will still be in progress when the position is relayed to ground observatories. Thus, the optical observer will need to make a fast determination of the possible position of the GRB source in this coarse positioning. 

An automated acquisition system combined with a relatively large field of view, rapid-slew telescope with polarimetry capability appears to be a good strategy for extending the study of GRBs.  Perhaps most important to this proposal is that GLAST will also have GRB localization capability, but it will produce much larger error boxes than Swift.  Since GLAST [18] is scheduled for launch in about 2006, and the two-year Swift mission may (or may not) overlap with GLAST for about one year, GLAST will definitely need ground-based assistance to determine accurate localizations, host galaxies, and GRB redshifts, and to study afterglow properties. The GLAST schedule is perhaps more fitted for this proposal, since we plan to have the proposed RRRT completely operational, calibrated and debugged at that time.

The expected rate of GRBs detected by Swift and GLAST over the RRT horizon is of about one every two-three days. Most of the times the telescope will be remain available for other type of research and/or educational purposes.

There is a large local amateur organization, the Back Bay Amateur Astronomers Association, which will collaborate with us. They plan to hunt for new objects that come near the Earth, called, NEO's, (Near Earth Objects). The research goal will be to find and track Asteroids and occasional Comets. They will use photometry (CCD camera) and the method known as "blinking" to identify Asteroids in the star field we are taking images. This program is naturally complementary to our main research objective since will use the same instrument and will allow a fast change of program. The group is also active in variable stars (including supernova searches) observations that are also among the most common among the amateur community and complementary to the other programs.

B - Description of the Research Instrumentation and Needs

We are planning the acquisition of a 0.6 cm (24”) rapid response robotic telescope (RRRT) to be located at Fan Mountain, Charlottesville, VA. (For location, see figures 2 and 3). As described in section A, the RRRT will study GRBs in the 10^th to 17^th magnitude range. It will be equipped with a CCD camera for UVBRI photometry and polarimetry. The telescope will be housed in its own enclosure and will be fully operational through the World Wide Web. Dr. Jay Norris from GLAST and Swift, a member of the LHEA-GSFC, will be a consultant for the project. Drs. David McDavid and Robert Rood, from the Department of Astronomy at the University of Virginia will be UVA’s point of contact. Staff and faculty members from UVA’s Department of Astronomy will collaborate in all stages of the project and will share operations.

The telescope will be robotic and completely automatic and mostly operated from the NSU campus.  Fan Mountain is about 130 miles from NSU’s campus. We expect to have; at least, weekly maintenance visits from NSU’s or UVA’s staff to the RRRT. There is a continuous people presence at Fan Mountain by UVA staff. We expect a strong collaboration and sharing of the use with the local Back Bay Amateur Astronomer (BBAA) association [19] that also expect site visit for manual use. We are planning periodical visits from NSU’s students for them to experiment the thrill of being at an astronomical observatory on a dark night, far from city lights (most of NSU’s students are inner city inhabitants).

The University of Virginia currently operates three telescopes at the Fan Mountain Observatory, about 15 miles south of Charlottesville. The Observatory complex was built in the mid 1960's [17]. The Observatory was used extensively for research up to the late 1980's and then fell in disuse as the astronomers moved their research to larger aperture telescopes. There are currently three (40”, 30” and 10”) telescopes at Fan Mountain but are not robotically operated. In recent years, extensive hardware upgrades and instrumentation efforts have transformed the observatory into a more modern research facility. Efforts are underway to expand the capabilities to include IR imaging and low-resolution spectroscopy with the 40” telescope.  A RRRT will be an important complement to the current infrastructure.

Figure 2: Fan Mountain Observatory with the possible location
for the planned NSU‘s rapid response robotic telescope (RRRT).

Figure 3: UVA’s Fan Mountain Observatory with the possible location
for the planned NSU‘s rapid response robotic telescope (RRRT). [17]

In cooperation with the University of Virginia (UVA), we will able to use the current installations at Fan Mountain (small machine shop and astronomer’s accommodations) and to get power and fast Internet access for our telescope. UVA in turn will be able to share the use of the telescope as schedule permits. 

Specifications for the Rapid Response Robotic Telescope 

The telescope should be capable of local (manual/real time) and remote (scheduled and real time) robotic operation. We favor a 0.6 cm (24”) aperture, Ritchey-Chretien, f/8 optics with an equatorial fork mounting. Fast slewing, precision positioning and tracking are the most important characteristics for the telescope and are specified accordingly to our research goals (GRB-OA photometric data and Swift time scales).

To maintain precision on the telescopic focus and position after fast slewing, it will necessary to require very small (zero) optics and mechanical mounting shifts, small detectable focal plane shifts versus focus position and small focus changes with temperature, zenithal distances and instrumental positions. These specifications could be achieved with good focal compensation or computerized focus control, which will be specified to the vendors.

The telescope must be fully automatic. All controllable elements must be Ethernet addressable, or “web based controlled”, such as: optical covers, optical assembly exhausts fans, telescope positioning, instrumentation (photometer and polarimeter controls), cameras, filter wheel, dome, scheduling, weather observation (cloud and seeing monitoring) and other observatory related hardware.

Positioning specifications

20 arcsec (absolute) pointing precision within 45 deg zenithal distance.
5   arcsec (relative) pointing precision for more than 10 degrees.
1   arcsec (relative) pointing repeatability for offset move of  less than 2 degrees

Tracking specifications

Track objects at sidereal rates (open-loop) better than 0.3 arcsec in Right Ascension and Declination per 100 seconds
Tracking jitter less than 0.3 arcsec (rms) in Right Ascension and Declination

Slewing specifications

Fast Slew rate: greater or equal to 5 deg/sec (better than 90 degree in 18 seconds)
Slow Slew Rate Selectable from 0.1 deg/sec to 0.5 deg/sec
Fine Set rate: Selectable 0.01 deg/sec to 0.05 deg/sec
Guide rate 1 arcsec nominal, adjustable zero to 10 arc/sec
Accelerations
0.5 degree/sec² nominal
1 to 2 degrees/sec² in special applications

Filter Wheel - 15 Position Filter Wheel for 2-inch diameter filters

We favor an equatorial fork mounting made of steel elements with stainless steel roller drives component. Gears and drives should be able to withstand fast slew and low maintenance. All mount and telescope’s structural elements should be constructed from like thermal coefficient materials for matching thermal expansion properties to better pointing and tracking.

We favor a commercial CCD camera, with very high quantum efficiency (i.e., Apogee Alta E42 – Back illuminated) without gaseous cooling (for easier robotic capabilities). The telescope operating system needs to be Linux based (for better in-house programming accessibility) with GUI interfaces and using the standard Astronomy Command Language (ACL) interface (to be able to use commercial astronomical software, i.e. The Sky) and TCP/IP link to the Internet.

For the RRRT enclosure we favor a steel dome and circular observatory (about 15’ Dome) with azimuthal rotation via computer integrated stepper motor. We will also consider a roll-off solution, but we favor the dome given the wind characteristics at Fan Mountain. A more protected enclosure (dome) will allow us to operate in higher wind conditions. The enclosure will be equipped with exhaust fans (remotely controlled), web cameras and total remote web based control.

As we have discussed previously, the early polarimetry of GRB-OA will be an important ingredient in our research program. RHESSI recently measured polarization of the prompt gamma-ray emission of ~ 80% [20].  This supplements several optical observations of afterglow polarization of order 1–3% that have been reported [21].  Polarization evolution will provide one of best diagnostics of the magnetic field and jet dynamics [5]. To perform polarimetry, the fewer the optical elements in the light path, a.k.a. the simpler the optical design, the better off you're likely to be as less corrections need to be performed on the acquired data. An equatorial mounting is required.

We will use the same CCD camera for imaging and polarimetry.  For polarimetry, we would need an additional module in front of the band pass filter wheel to carry the polarimeter analyzers - a pair of calcite Savart plates would be the simplest and least expensive (this will be done “in-house”).  A tailpiece could be designed with both filter wheels in place and with blank space(s) in the analyzer wheel for direct imaging.  It would need to be carefully designed to avoid vignetting, but that would not be a major difficulty. Dr. David McDavid, from UVA, that has already designed and built other similar polarimeters  [22,23] will be our consultant.

The measured GRB-OA polarization decay times are on the order of hours to days and have been done with bigger telescopes up to larger magnitudes. We want to measure brighter and in shorter times. It is not clearly know the degree of polarization in the early stages of a GRB-QA (~150 sec). If the degree of polarization is of a few to several percent, something like 10 to 20 minutes of integration time will give the necessary precision at magnitude 12 with a 24” in aperture.  Beyond that, the data can be continuously streamed to disk for later analysis.  You get one complete measurement of the polarization degree and position angle for every revolution of the analyzer, which are around 10 per second.  It is entirely possible to go back and analyze the recorded data stream by binning any time interval or any number of subsets of time intervals you like and simply adjusting your choices to get any desired level of precision and look for variations.  In other words this method of measurement is based on the "signal averaging" technique, where the precision improves according to how many measurements you combine.  The polarization parameters, when measured this way, are completely independent of the brightness of the object. Also by streaming data continuously to disk you have a record of brightness that can be analyzed separately from the polarization to look for variability.