"…while the stars that oversprinkle all the heavens seem to twinkle with a crystalline delight…"

Throughout the ages, observers of the sky have pondered the nature of the countless points of light sparkling like crystals against the velvety backdrop of the night sky. Indeed, the twinkling stars are a beautiful sight to behold. They have been the source of inspiration for artists and astronomers alike. While artists have been inspired to create beautiful paintings and poetry, astronomers have been inspired to understand what causes twinkling, what can be learned from twinkling, and how to take the twinkle out.

Scintillation, the technical term for the twinkling of stars, is defined as the rapid and irregular variation of intensity of celestial objects. In this paper I will present a historical overview of the study of scintillation and the theories put forth to explain the phenomenon. I will discuss different types of scintillation and describe what can be learned about objects through this phenomenon. I will discuss the negative impact of scintillation on astronomic observing and imaging, and review the various techniques that have been developed to minimize the deleterious effects of scintillation on astronomic research.

History

There is no doubt that prehistoric cultures had theories on the nature of the twinkling stars. Their stories, passed down through countless generations, are still known of today. The Australian Aborigines of 10,000 years ago believed that the stars marked the location of distant tribes. The flickering points of light were produced by their campfires ⁽¹⁾.

Ancient Greek philosophers had their theories as well. One of the more popular explanations, put forth by Aristotle around 300 BC, attributed stellar scintillation to a weakness of vision. Aristotle wrote, "The planets are near, and our vision reaches them. The fixed stars are too far, and their distance causes our vision to waver."

Roger Bacon, a 13^th century monk who is often referred to as the forerunner of the modern scientific method, wrote extensively on astronomy and optical science. In the second volume of his most popular work, Opus Majus, he discussed the phenomenon of shimmering stars. Bacon, like Aristotle, attributed the scintillation of stars to an error of human vision.

The belief that the twinkling of stars was in the eye of the beholder appears to have been a popular one. In the 16^th century, Leonardo da Vinci studied "the nature of the luminous ray", his term for the light that was responsible for the twinkling of the stars. He believed that scintillation was not inherent in the stars themselves, but simply an optical illusion ⁽²⁾.

The theories put forth by the great astronomers of the late 16^th century, Tycho Brahe and Johannes Kepler, were no more accurate. Brahe felt that stellar scintillation was related to the movement of the stars in the celestial sphere. He is quoted as saying, "The enormous space to the stars takes part in the celestial daily motion. Planets do not thus rotate, and therefore do not scintillate." Kepler had a different explanation. He was certain that the twinkling of the stars was not an illusion, but an actual change in the brightness and color of the stars. Unlike many before them, both Brahe and Kepler incorrectly assumed that scintillation was a product of the stars themselves. It would be another century before the true nature of the phenomenon would begin to be understood.

The true cause of stellar scintillation was finally determined by Isaac Newton in the early 18^th century. He realized that it was the very air that we breathe that was responsible for the twinkling of the stars. In a discussion on his design of a new type of telescope that employed a mirror as its primary objective instead of a lens, he described how the turbulence in the atmosphere was responsible for altering the path of the incoming star light. The following passage, taken from Opticks, a work he published in 1704, describes the process in his own words:

"If the Theory of making Telescopes could at length be fully brought into Practice, yet there would be certain Bounds beyond which Telescopes could not perform. For the Air through which we look upon the Stars, is in a perpetual Tremor; as may be seen by the tremulous Motion of Shadows cast from high Towers, and by the twinkling of the fix’d Stars. But these Stars do not twinkle when viewed through Telescopes which have large apertures... Long Telescopes may cause Objects to appear brighter and larger than short ones can do, but they cannot be so formed as to take away that confusion of the Rays which arises from the Tremors of the Atmosphere. The only Remedy is a most serene and quiet Air, such as may perhaps be found on the tops of the highest Mountains above the grosser Clouds."

Newton’s suggestion that telescopes should be placed upon high mountains was an excellent one. However, for logistic and financial reasons most observatories continued to be built in or near cities ⁽³⁾. The Lick Observatory, built in 1888 on Mt. Hamilton in Northern California, was the first major observatory constructed on a mountain to take advantage of the superior observing conditions found at higher altitudes. Many sites were considered for the observatory, including Mount St. Helena and other peaks in the San Francisco Bay area. To aid in the decision making process, S. W. Burnham, an accomplished double-star observer, was hired to test the ‘seeing’ (a term denoting how well or poorly the atmosphere allows an image to appear) at Mt. Hamilton. The conditions at the site exceeded Burnham’s highest expectations, as evidenced by the entries in his observing notes such as, "First class seeing. No wind at all… Splendid night – absolutely still "⁽⁴⁾. It was therefore determined that Mt. Hamilton would be an excellent site for the new observatory. Thus, Lick Observatory set the precedent for taking a location’s seeing into account as a primary consideration for the construction on an observatory.

George Ellery Hale was familiar with the concept of placing telescopes as high above the ground as possible. When he set out to build his new solar observatory, he began looking for a suitable mountain. Hale considered many sites rumored to possess excellent seeing, including Mt. Hamilton and Mt. Palomar, before choosing Mt. Wilson, a 6000’ peak overlooking the Los Angeles Basin. Since Hale planned to construct not only solar telescopes but also the largest reflecting telescope that had yet been built, a site with excellent observing conditions was of prime concern.

Hale was willing to go to any lengths, or in this case, heights, to determine whether the observing conditions on Mt. Wilson were acceptable. During his first site surveying expedition, he could often be found scrambling up tall pine trees, dragging his 3-inch telescope with him. The following is an excerpt from an entry Hale made in his diary during that expedition:

"Seeing poor at tree…at 32 feet and 68 feet, seeing better…" ⁽⁵⁾

Hale had discovered that getting just an extra 20 meters above the ground resulted in improved seeing conditions, especially for solar observing. It is therefore no surprise that when he constructed his solar telescopes on Mt. Wilson, he placed the optics on platforms high above the ground, first at 60 feet (~18 meters) and later at 150 feet (~45 meters).

Ever since the days of these great observatories, Lick and Mt. Wilson, telescopes have continued to be built upon mountains, the higher the better. Mauna Kea on the island of Hawaii, is home to optical, infrared and radio telescopes. At an altitude of about four kilometers (~14,000 feet) and therefore above two-thirds of the Earth’s atmosphere, the telescopes are far above the level of the lower atmosphere where much of the turbulence occurs. Getting above as much of the atmosphere as possible is the first line of defense from the negative effects caused by a turbulent atmosphere.

As Newton correctly determined, the apparent twinkling of the stars is caused by the Earth’s atmosphere. Temperature and density variations in the atmosphere result in the random refraction of the incoming star light.

Light rays are bent or refracted when passing through mediums of different density. Light from a star travels through the near vacuum of space relatively unhindered. When the light reaches our atmosphere, however, it is refracted by various amounts and in different directions, depending upon the density of the air it encounters. When the light is refracted away from us, the intensity of the star is diminished, and when it is refracted toward us the intensity is increased. The end result is the starlight’s apparent twinkling. The more vigorous the turbulence (the movement of the different density layers in the atmosphere), the more pronounced the scintillation or twinkling.

Not only does the starlight vary in intensity, it varies in color as well. White light is composed of all colors from violet and blue, through green and yellow, to orange and red. The pockets of air of different densities act like prisms to disperse the light. Dispersion, the spreading out of white light into its constituent colors, occurs because different wavelengths of light are refracted to different degrees, with the shorter wavelengths of violet and blue being refracted more than longer wavelengths of orange and red. For this reason, the twinkling stars often appear to rapidly flash through many different colors.

Atmospheric turbulence occurs in three main regions: the free atmosphere, the atmospheric boundary layer, and in the immediate vicinity of the telescope ⁽⁶⁾. Turbulence in the free atmosphere is due to temperature gradients and high wind speeds associated with the jet stream, located about 12 kilometers (~4.5 miles) above the surface of the Earth. The atmosphere exists in layers of differing densities in this region, and are in constant motion. This type of turbulence is generally referred to as horizontal turbulence.

The atmospheric boundary layer is the area between the Earth’s surface and the atmosphere. The turbulence in this region is due in large part to the frictional drag of air against the ground. This being the case, a relatively smooth terrain will not produce as much turbulence as a rough one.

The turbulence that occurs in the immediate vicinity of the telescope is often responsible for creating the majority of poor seeing, but is also the most easily controlled. In short, anything that traps or stores heat in or around the telescope will cause poor seeing. For example, electronic equipment such as computers, used to control the telescope and imaging systems produce heat. If the areas where these electronic components are located are not properly ventilated, vertical turbulence can be produced. Even the motors responsible for moving the telescope into position and keeping the telescope pointed at an object as it travels across the sky can produce heat. The type of ground cover relates to the amount of turbulence in this region as well. During the day, the ground absorbs solar radiation, which is stored as heat. This stored heat is re-radiated back into space at night, resulting in vertical turbulence. A dark ground cover, such as asphalt, for example, will absorb and store much more heat than dirt or a surface covered by grass, resulting in poor seeing.

Astronomers refer to the amount of scintillation as astronomical ‘seeing’. Minimal twinkling is referred to as good seeing; conversely, a high degree of scintillation is referred to as poor seeing. Seeing is quantitatively defined to be the diameter of a star image in arcseconds, caused by atmospheric turbulence. While there are many factors that influence the quality of seeing, including sky brightness and transparency, the main factor is scintillation.

Resolution, the amount of detail that can be discerned in an image measured in arcseconds, is determined by both the optics of the telescope (larger apertures yield higher resolutions), and by the Earth’s atmosphere. For large telescopes, the limit on resolution is determined by the atmosphere rather than by the telescope optics.

When a star is observed through a telescope at high magnification, the star seems to dance about wildly. This is referred to as image motion. When the star is recorded photographically or digitally, the resulting image of the star appears not as a point of light, but as a blurry disk. This effect is caused by the presence of air pockets of different densities and therefore different refractive indices in the atmosphere. The majority of these air pockets in the atmosphere are no bigger than about 15-20 centimeters across. In a small telescope, there may be only one such air pocket covering the field of view. But a large telescope such as Keck (10 meters) could have hundreds. In this case not only does the star image move, but multiple star images are produced, many of them distorted. It is therefore impossible to discern small scale details of less than 1–2 arcseconds in an image produced under these conditions, which severely limits the success of the observations.

Stellar scintillation is not always looked upon as a negative, as there are certain determinations that can be made about the nature of an object through studying the phenomenon. For example, the occultation of a star by a planet can reveal the presence of a planetary atmosphere. A planet’s atmosphere can be detected by the increase in scintillation of the star in the moments before it is completely occulted by the planet. Both the density and the depth of a planetary atmosphere can be determined from such observations ⁽⁷⁾. More often than not, however, optical astronomers focus their efforts upon minimizing the amount of scintillation utilizing a variety of techniques.

 
Taking the Twinkle Out

As was discussed previously, placing telescopes on high mountains maximizes the potential for good seeing by getting above a large percentage of the atmosphere. There are other considerations in site selection as well. Some of the finest sites for observatories are mountain tops that face into prevailing winds that have crossed thousands of miles of cool oceans. The reasoning behind such a choice is that water does not change temperature to a great degree between day and night. Vertical turbulence resulting from the re-radiation of stored heat is at a minimum under such conditions. Mt. Wilson, Mt. Palomar, and Mauna Kea are just a few examples. Even a smaller body of water such as a lake can produce similar results. Big Bear Solar Observatory resides on a protrusion built into Big Bear Lake to take advantage of this effect.

In addition to site location, the conditions in the immediate vicinity of a telescope have a large influence on the quality of seeing. The domes that house telescopes are painted white in an effort to reflect as much solar radiation as possible, and are constructed of materials that absorb minimal heat. The lower floors of a dome beneath the telescope are cooled by fans or refrigeration units to decrease the likelihood of rising warm air currents, which interfere with observations. Also, it is customary to open the dome some hours before nightfall to allow the warm air that has collected inside the dome over the course of the day to escape long before observing commences. The temperature is therefore given time to drop closer to the outside temperature, once again minimizing the amount of rising air currents responsible for vertical turbulence that result in poor seeing.

Since we know that atmospheric turbulence causes poor seeing, and getting above much of the atmosphere improves seeing, then the next logical step is to get above the atmosphere entirely. The Hubble Space Telescope (HST) is the result of this line of reasoning. Orbiting above the Earth at an altitude of about 600 kilometers (~370 miles), HST has a clear view of the heavens. HST has produced images of unprecedented clarity, limited not by atmospheric effects but by the telescope’s optics. Unfortunately, the astronomic price tag of $2.1 billion (US dollars) attached to this instrument is a major deterrent against launching more of these optical telescopes into orbit.

There are, however, more affordable means of improving seeing for ground-based telescopes. Adaptive Optics (AO) is a means by which the incoming planar wavefront of light which becomes corrugated by its passage through the atmosphere, can be once again made planar ⁽⁸⁾. An AO system employs a wavefront reconstructor to accomplish this task. The wavefront reconstructor consists of a high speed camera that is capable of analyzing the structure of the distorted wavefront. The information obtained by the camera is used to provide a list of electronic corrections that are passed on to a deformable mirror. The deformable mirror is then commanded into a shape that matches that of the corrugated wavefront. In this manner, the distortions in the wavefront are cancelled out, and the wavefront is returned to the planar shape it had before passing through the atmosphere.

Zeiss, the manufacturer of planetarium projectors as well as fine optics, has been working on the other side of the coin. In a recent press release by the renovated Hayden Planetarium in New York, they were proud to announce the purchase of the one-of-a-kind Zeiss Mark IV planetarium projector, capable of recreating the realistic effect of stellar scintillation via the use of fiber optics. It is ironic that while optical astronomers are working on cost effective techniques to take the twinkle out, public educators have found a way to put the twinkle in for a mere four million dollars!

 
Other Types of Scintillation

We have discussed the scintillation of visible light sources caused by the turbulence of the Earth’s atmosphere. Many objects observed in longer wavelengths of millimeters to meters, radio wavelengths, scintillate as well. The radiation from distant radio sources can be effected by the different densities of the material in the interstellar medium (ISM). Turbulent clouds of gas present in the ISM can cause a compact radio source to scintillate on a time scale ranging from minutes to days, depending upon the angular diameter of the source and the wavelength of observation. Therefore, this phenomenon, known as interstellar scintillation (ISS) can provide information about the angular diameter and transverse motion of a scintillating radio source, as well as the density and dynamics of the gasses that make up the interstellar medium.

Interplanetary scintillation (IPS) is the rapid variation in amplitude of radio waves caused by the interaction of the solar wind with the Earth’s ionosphere, a layer in the Earth’s upper atmosphere in which many of the atoms are ionized (the atom has lost electrons). The solar wind is the continual outflow of particles (mostly protons and electrons) from the Sun. The particles in the solar wind interact with the charged particles of the ionosphere, creating inhomogeneous pockets that refract the incoming radio waves, much as visible light is refracted by air pockets of different densities in the Earth’s atmosphere.

Features of the solar wind can be studied through observations of the scintillation of compact radio sources. For example, the speed of the solar wind can be determined by simultaneously observing the scintillation pattern of a compact radio source such as a quasar with two different radio telescopes some distance apart. The scintillation patterns observed at each site will be nearly identical, except for a small time lag. By measuring this time lag and the distance between the two observing sites, the velocity of the scintillation pattern and therefore the velocity of the solar wind, can be determined ⁽⁹⁾.

The direction of the solar wind can be determined as well through simultaneous observations at two different sites. This is accomplished by observing a single source for an extended period of time. As the Earth rotates, the orientation of the baseline between the two sites changes continuously throughout the duration of the observation. Since the degree of scintillation is greatest when the baseline is parallel to the flow of the solar wind, the direction of the solar wind can easily be determined ⁽¹⁰⁾.

Not only does the phenomenon of interplanetary scintillation provide information about the solar wind, it supplies information on the nature of the radio source exhibiting the phenomenon. In the mid-1960’s, Antony Hewish discovered the relationship between IPS and the size of radio sources ⁽¹¹⁾. He found that the angular diameter of compact radio sources between 0.1 – 1 arcsecond could be determined by analyzing the degree of scintillation present when observing the objects at longer radio wavelengths.

Armed with this knowledge, Hewish set out to search for quasars, compact radio sources with strong signals. Since compact sources scintillate much more than extended sources, scintillation was an excellent method for detecting quasars. It was during this search for quasars led by Hewish that a new breed of star was discovered by Jocelyn Bell Burnell.

Unlike IPS, which results in a random variation in radio signal fluctuation, Bell discovered a source that exhibited a series of equally spaced fluctuations or pulses, 1^1/3 seconds apart. No mechanism known of at the time could be responsible for causing an object the size of a star to vary in intensity at such a rapid rate, so Hewish believed it was likely a man-made signal. In the following months, however, Bell discovered a number of similar sources in different areas of the sky, ruling out the possibility that the signals were man-made ⁽¹²⁾. Bell had discovered pulsars, the rapidly spinning remnants of supernova explosions. This type of neutron star emits a narrow beam of radiation as it spins, that is detected as a pulsating radio source when the beam sweeps across the Earth. Thus, the variations in amplitude of the radio signal produced by pulsars are intrinsic in pulsars.

Much can be learned from the occultation of radio sources by the planets. In the mid-1980’s, the orientation of the magnetic fields of both Jupiter and Saturn were inferred from observations of radio scintillation during the occultations of the Voyager spacecraft ⁽¹³⁾. This was accomplished by analyzing the alignment of the fluctuations in amplitude and phase of Voyager’s radio signal (telemetry), caused by the planets’ ionospheres. Studies were also made of the occultation of Voyager by Titan, one of Saturn’s satellites. Radio scintillation measurements taken as the spacecraft passed behind Titan’s atmosphere revealed a high degree of signal fluctuations. The nature of these fluctuations suggested the presence of horizontal variations in atmospheric density, and atmospheric gravity waves, the cyclical rising of more-dense material above less-dense material followed by the restoration of the materials by gravity.

 
Conclusion

Who among us does not enjoy sitting outside on a warm summer night, perhaps in a grassy field far away from the glare of city lights, and gazing up into the dark night sky, virtually alive with twinkling stars. While the study of scintillation has gone on for millennia, it has only been in the past fifty years that technological advancements have enabled astronomers to correct the negative effects of the Earth’s atmosphere on the incoming light from celestial objects. Adaptive optics is now considered to be on the cutting edge of astronomic research. Through painstaking effort, astronomers have gained an increasing knowledge about the nature of scintillating optical and radio sources. As we continue to enjoy the phenomenon of scintillation, astronomers will continue to study this phenomenon, improving upon their understanding of the nature of scintillation and scintillating sources, and developing new techniques to minimize its effects.

References:

(1) Aboriginal Australia: http://www.aboriginalaustralia.com/

(2) Leonardo da Vinci: http://www.sumscorp.com/books/contin/p4c3.htm

(3) Lick Observatory Website: http://www.ucolick.org/~mountain/mthamilton/public/history/bldg_the_obs.html

(4) Osterbrock, D.E., Gustafson, J.R, Unruh, S.W, 1988, Eye on the Sky: Lick Observatory’s First Century, University of California Press

(5) Wright, H., 1994 (reprint), Explorer of the Universe, American Institute of Physics

(6) Baird, M., 1982, Weather Forecasting for Astronomy, Winmark Press

(7) Africano, J. et al, 1977, The Occultation of Epsilon Geminorum by Mars, Astrophysical Journal, Part 1, vol. 214, June 15, p. 934-945

(9) Department of Physics, University of Wales:  www.aber.ac.uk/~dphwww/research/ips.html

(10) Moran, P.J. et al, 1998, Annales Geophysicae, Abstract Volume 16 Issue 10, p 1259-1264

(11) Little, L.T. and Hewish, A., 1966, Interplanetary Scintillation and its Relation to the Angular Structure of Radio Sources, Monthly Notices of the Royal Astronomical Society, Vol. 134, p.221

(12) Burnell, S. J. B, 1977, Petit Four (After Dinner Speech), Annals New York Academy of Sciences, v. 302, pp. 685-689: 
http://cosmos.colorado.edu/astr1120/bell.html