Our Cataclysmic-Variable Network

Joseph Patterson
Sky and Telescope
October 1998, Page 77
(used with permission, see note below)

Most long-time visual observers discover, at some point in their lives, the allure of cataclysmic variable stars. For many nights you check a little patch of sky where a chart says an eruptive variable resides. The spot is always blank. Then one night you swing the telescope to the spot again, and bang! A new star is shining smack at the correct point, in a field that you have completely memorized by now.

When such stars attain naked-eye brilliance, they can profoundly influence people’s lives. Tycho Brahe’s observations of the supernova of 1572, for example, earned him financial support from the Danish monarch and guaranteed his place in astronomical history. The little specks we see in telescopes are usually not that dramatic, but they do provide a personal thrill to be remembered
for a long time.

Cataclysmic variables (CVs) are stars that erupt in brightness more or less unpredictably. They include novae, dwarf novae, recurrent novae, symbiotic stars, and others. Much of our physical understanding of them comes from the pioneering work done in the 1950s and ’60s by Merle Walker, Jesse Greenstein, Robert Kraft, and Jozef Smak at Palomar and Lick observatories. CVs are close binary systems in which a fairly normal, low-mass star gently pours a stream of gas toward a white dwarf. The stream settles into orbit around the white dwarf and forms an “accretion disk,” releasing its gravitational energy as it gradually spirals its way down to the white dwarf’s surface. Perhaps surprisingly, we don’t usually see the white dwarf or its companion star directly. The system’s light output is ordinarily dominated by the hot, brilliant accretion disk alone. Thus CV s are well suited as natural laboratories for studying accretion-disk physics, which has relevance in other contexts around the universe.

Small changes in the gas transfer rate, disk structure, or accretion pattern will manifest themselves as slight, rapid fluctuation in the system’s brightness. Sudden collapses of disk structure cause dwarf-nova flare-ups. The nuclear ignition of hydrogen that has piled up on the white dwarf’s surface leads to the outbursts of novae and recurrent novae.

Virtually all our data about the frequency and amplitude of CV outbursts comes from visual observers. Continuous records spanning nearly a century are kept by the American Association of Variable Star Observers (AAVSO), the British Astronomical Association, and the Royal Astronomical Society of New Zealand, among others. Thousands of observers belonging to these organizations have contributed millions of observations to the AAVSO International Database. So I’m quite certain that amateur visual observers will continue to dominate this field for the foreseeable future.

But the 1990s have brought to the backyard observer cheap, powerful CCD detectors and desktop computers. A 10-inch telescope can now routinely carry out stellar photometry to 17th magnitude,
which means many research programs traditionally done by professional astronomers can now be delegated to amateurs using their own equipment.

This is an exciting prospect for the future of variable-star astronomy, if you consider a few basic numbers. There are perhaps 300 professional researchers worldwide who study variable stars, and each spends maybe four weeks a year at telescopes — meaning that an effective total of about 20 telescopes is devoted to this task. But there are thousands of amateur astronomers around the world who actively use telescopes for variable-star work. What would happen if many of them were equipped with computers and CCDs and were organized to carry out coordinated research programs? By sheer
numbers they would probably dominate the field.

A Fruitful Collaboration

In 1980 I became acquainted with David Skillman, an amateur from Laurel, Maryland, who had built in his backyard a 12-inch reflector equipped with a photomultiplier-based photometer and controlled by an Apple II computer to perform automated observations of variable stars (S&T: January 1981, page 71, and May 1993, page 83). We began to collaborate on research projects and called ourselves the Center for Basement Astrophysics — since all the construction and telescope control took place in Dave’s basement, and since we liked the image of dark, subversive, unsanctioned activity. We wrote a few research papers, but our photometer’s limiting magnitude of 11.5 hampered our observations. The instrument could not reliably find and center fainter stars. We then upgraded to a 26-inch reflector and CCD camera, which improved our limiting magnitude to more than 16 and gave us a practically unlimited number of targets.

By the early 1990s the quality of CCDs and computers continued to rise sharply while prices fell . Soon we found ourselves talking with similarly equipped photometrists around the world. We began to realize that we could establish a small but powerful and highly efficient collaborative network for studying variable stars. At any given site you can study a star for only a few hours per night; you lose about 13 to 18 hours to daylight and twilight and more when the object sets early or rises late. And you often lose whole nights to bad weather. This leads to very inadequate coverage of variables that fluctuate on time scales of 0.5 to 3 days. Nature is surely full of deep secrets that play out on such time scales, but our rotating and cloud-plagued home planet is a poor platform from which to probe those secrets.

Enter the Center for Backyard Astrophysics (CBA), so renamed because too few of our members have basements. At present we have more than a dozen dedicated observers from the United States, Canada, the United Kingdom, Denmark, Belgium, New Zealand, Australia, Japan, Italy, South Africa, Israel, Slovakia, the Czech Republic, Russia, and Ukraine, in addition to many more who occasionally join us for observing campaigns. About a third are professional astronomers, a third are amateurs, and the rest are difficult to classify. Our instruments range from Jerry Gunn’s 8-inch Schmidt-Cassegrain in East Peoria, Illinois, to the 60-inch Cassegrain occasionally used by Pieter Meintjes in Bloemfontein, South Africa. Our wide geographical distribution gives us an unequaled ability to detect periodic signals in frequency ranges near 24 hours that are poorly sampled at any one site. Also, by sheer weight of coverage, the CBA is a great research network for studying transient phenomena, relatively free from the ravages of local weather.

We have found that CVs are an excellent subject for study since their light curves are boiling with activity on time scales from minutes to years. The most interesting phenomena are the periodic signals. These reveal the rotation period of the white dwarf, the orbital period of the binary system, less obvious phenomena such as “superhump” periods, and precession periods of the accretion disk. To date we have explored these matters in 22 research papers published in astronomical journals.

Observing Campaigns

About twice a year we send out a CBA newsletter over the Internet with updates on past, present, and future observing programs, as well as the status of scientific papers awaiting publication. Roughly every month we send out a list of suggested targets and declare observing campaigns for particular times on particular stars. Typically, the observer posts the reduced data a few days later by email or on floppy disks. At the end of an observing season or perhaps two, the crop of data is usually ready for harvest in a scientific paper.

Our observing technique is simple: We take a long “time series” of differential photometry of the variable with respect to a comparison star in the same field. A 5-hour run is typical, and we usually amass 50 to 70 nights on a star before a study is completed. The focus on periodic signals serves us well. These signals usually occur at all wavelengths and can therefore be studied in “white light” (no filters). That suits us just fine, because our small telescopes need all the photons they can get.

An example of recent work is shown on page 78. It’s a two-day light curve of the eclipsing dwarf nova DV Ursae Majoris. During its superoutburst in April 1997 a large-amplitude photometric wave (“superhump”) appeared in the light curve with a period of 2.18 hours. The deep eclipses continued to occur on their strict 2.06-hour period, so each eclipse took place at a slightly different phase in the 2.18-hour cycle. As a result the depth and shape of the eclipse varied with the “beat period” with which these two cycles lap each other (2 .2 days). This can be used to infer the varying shape of the eclipsed object, the accretion disk.

Another example is the two-week light curve of the cataclysmic variable V803 Centauri. This star consists of two helium-burning white dwarfs in very close orbit, with rapid variability that was said to be irregular. With our observing stations in New Zealand, South Africa, and Chile, we managed to keep a constant watch on the star from April to May 1997. We found the remarkable light curve shown on page 79, proving that the fundamental period of variability is about 23 hours. Despite the short period and the exotic composition of the accretion disk (nearly pure helium), this is actually the recurrence period for dwarf-nova eruptions. A dwarf nova made of nearly pure helium! The recurrence period is a record, the shortest known. This remained unrecognized for 20 years because observations were so bedeviled by the Earth’s rotation period.

Our biggest thrill came with discovery of the orbital period of the strange but wonderful helium star AM Canum Venaticorum. The light curve exhibits a 1,051-second wave, which has puzzled astronomers since its discovery in 1962. At first it seemed possible to interpret this as the underlying orbital period of the binary. But radial-velocity searches never confirmed this, and recent studies demonstrate that the variation is erratically unstable and therefore cannot possibly be the true orbital period.

So what is the orbital period? In 1997 we conducted a global photometry campaign to find out. The results are shown on page 79. We searched for periods by calculating the light curve’s power spectrum (a graph showing the strength of variations at every possible period). A spike in the power spectrum indicates a periodicity; the flanking spikes are “aliases” caused by clustering of observations at certain longitudes (usually the Maryland-Arizona cluster). The spike at 1,028.74 seconds has never been seen before, but a theory published in 1993 requires that exact orbital period! We then returned to archival light curves and found that this signal has actually been present for years, hovering near previous detection limits. So the long search for the orbital period, by far the shortest among all CVs, is over.

The future seems bright for networked programs of this type. On the technical side it’s inevitable, since all the required hardware is improving and getting cheaper. But the Earth will remain as round and cloudy as ever. We need to expand our network to reach the goal of full 24-hour coverage. This will require greater vigilance in observing, recruitment, and fundraising. Up until now, a grant from the Research Corporation has enabled us to keep surviving and publishing.

The CBA’s weakest points are the lack of observers at certain longitudes, including those of the world’s most populous countries (like China and India), and our lack of regular representation at the best astronomical sites (Chile and Hawaii). Our other priority is to establish interest groups on other variable-star topics, especially multiperiodic stars such as nonradial pulsators.

But I’ve found that the critical limiting factors are usually not technical but human. Only an unusual person gets excited about making 5,000 sequential measurements of the same star. It takes an enduring love of stargazing and of science. But astronomy seems to attract people with high tenacity as well as technical skill, and it has been my privilege to find a couple dozen such people. We hope to find many others out there and bring them together in a network that will form an important new research tool in variable-star astronomy.

Copyright © 1998 Sky Publishing Corporation; used with permission. This material may not be reproduced in any form, either printed or electronic, without first obtaining permission from Sky Publishing Corp., P.O. Box 9111, Belmont, MA 02178-9111, USA.