Chadwick A. Moore Visual Estimations of Night Sky Brightness We can no longer avoid the issue of light pollution, because soon we will have nowhere left to go. The global problem of light pollution has never been so artfully expressed as in Woodruff T. Sullivan’s famous ‘Earth at Night’ satellite images.* Kosai and Isobe 1992 anagers of parks, preserves, refuges, and wilderness areas are becoming increasingly concerned over the loss in visibility of the night sky. Like the apocalyptic threat of Rachel Carson’s Silent Spring, those who cherish the night skies fear a night of no stars, with only the sallow glow of streetlights for inspira- Mtion. Encroaching city lights scatter light upward, bathing the otherwise dark sky and reducing the contrast to a point where stars are lost in urban glow. The effect of urban lighting, also known as light pollution, can reach surpris- ingly far. For example, from Death Valley National Park the lights of Las Ve- gas produce an obvious and obtrusive glow even though the city is 100 miles to the southeast. Los Angeles, 160 miles to the southwest, produces a dim but broad glow across the southern horizon (Moore and Duriscoe, in prep.). Astronomers at observatories contrast, and atmospheric extinction. were perhaps first to notice this However, less costly tools are avail- problem. As early as 1970, astrono- able to those interested in monitoring mers were scouring the USA for suit- their dark sky resource. The simplest able observing sites away from city and least costly monitoring methods lights; remaining opportunities were are visual estimations using the hu- few (Walker 1970; Garstang 1989). man eye. Today, astronomical observatories Under a pristine dark sky, per- employ multi-million-dollar equip- haps 14,000 stars in the celestial ment to measure sky brightness, sphere would be discretely visible _______________ * 46 The George Wright FORUM (Figure 1). Outdoor lighting tends to trast between the background and scatter light upward, brightening the fainter stars until they become invisi- background of space. This increase ble to the eye. Also lost with the stars in sky brightness reduces the con- are the diffuse objects in the sky— Volume 18 • Number 4 2001 47 nebulae, galaxies, comets, and the ever, what the eye lacks in precision, river of stars in our galaxy called the it makes up for in sensitivity and ease Milky Way. The visible loss of these of use. The scoptic (grayscale) vision faint and diffuse astronomical objects we use at night is surprisingly sensi- is what troubles amateur astronomers tive, able to detect as few as 200 so much. This group has been the photons per second falling on the most vocal in opposing light pollu- retina and transmitting a message to tion and promoting the conservation the brain (Russell 1917). The eye’s of dark night skies. rod cells are 1,000 times more sensi- Amateur astronomers and meteor tive than the color-detecting cone observers have made visual estima- cells (Carr et al. 1989b). Scoptic vi- tions of limiting magnitude for years. sion is most sensitive in the greens Limiting magnitude is a measure of and blues, and least sensitive to the the brightness of the faintest star one reds; thus the use of red-filtered can see. The astronomical magnitude flashlights to preserve night vision. scale increases with faintness. Mag- The star magnitudes used in this nitude zero represents bright stars method are measured in the “John- such as Vega, Antares, or Rigel, son V” spectrum, which closely while magnitude 7 stars would be matches the human eye’s scoptic vi- near the faint limit of most dark skies. sion, and are therefore an appropri- In exceptional cases, magnitude 8 ate analogue for brightness meas- stars have been observed with the urements. naked eye (Russell 1917; Bowen 1947). The number of visible stars, and the diversity of astronomical The visual estimation of limiting objects visible, decrease rapidly as magnitude is based on star counts of the limiting magnitude falls. It is not 25 established sample areas (similar uncommon for a remote area sur- to methods utilized by meteor ob- rounded by rapid urbanization to servers; Figure 2). Each area contains lose more than half the visible stars in a field of mapped stars with known a decade (Moore and Duriscoe, in brightness values. The observer prep.). scans the field using averted vision, The human eye is a somewhat trying to detect sequentially fainter imprecise instrument. Few people stars on the map. The faintest star have 20/20 vision without aid of observed becomes the sky’s limiting lenses, and the eye changes in light- magnitude (LM). By following the gathering capability and acuity with procedures for dark adaptation and age (Carr et al. 1989b). There is also counting, reasonable conformity can potential for bias in the eye’s central be attained between observers processing unit— the brain. How- (Blackwell 1946). 48 The George Wright FORUM Initially, this star count is con- fying some light sources while sup- ducted at the zenith (straight over- pressing others. The effect of local head). Counts can also be conducted weather upon sky brightness is an in quadrants of the sky, and at vari- interesting study in itself, but such ous angular altitudes above the hori- conditions should be avoided to zon. The process can take as little as produce a baseline inventory to track 30 minutes to arrive at a zenith LM long-term changes. number. Observers have used this methodology to produce brightness maps of different parts of the night sky, or to take single measurements Light scattered upward is not the on multiple nights to capture the only factor affecting an LM meas- range of variation associated with urement. Pollutants in the atmos- weather, seasonal changes, or atmos- phere can substantially increase the pheric scattering. Observations are extinction of light as it is transmitted conducted under cloudless, moon- through the atmosphere. Airborne less nights. Even distant clouds or particulates, in the absence of light ground fog skew the results, ampli- pollution, can substantially reduce Volume 18 • Number 4 2001 49 the faintest stars visible, even thought notice that the stars look sharpest the sky background may appear very and brightest in the late hours just dark (Garstang 1991). In this case, before dawn. This trend is mostly the visibility of stars and astronomi- due to atmospheric turbulence which cal objects are lost to light scattering settles and diminishes as the night and absorption, not due to decreased progresses and the land cools. This contrast. Air pollution compounds trend may also be the result of re- the scattering of light pollution, fur- duced light pollution as people turn thering the degradation of night sky off their porch lights, park their cars, visibility. Finally, both factors are and outdoor athletic events come to a affected by humidity in the atmos- close. phere (Carr et al. 1989b). The growth and size of aerosol particles in the atmosphere is related to As with many natural resource moisture. Therefore, higher humid- measurements, much of the chal- ities are expected to exacerbate both lenge can be separating the natural the scattering of existing light pollu- and human components. Natural sky tion as well as the absorption of star- brightness does exist; the human- light (Garstang 1991). Conditions of made component of sky brightness is greater scattering tend to brighten light pollution. Moonlight is the most nearby light sources while dimming obvious natural source, but can eas- far-off light sources (Carr et al. ily be avoided by sampling when the 1989b). The corollary to this phe- moon has set. Zodiacal light, the nomenon is that dry, high-altitude spike of illuminated dust particles dark-sky sites are more susceptible to circling the inner solar system, can far-off light sources. be a significant natural light source. The lower atmosphere is turbu- It is most obvious in spring and lent, producing the common effect of autumn, but will set a few hours after twinkling stars. Turbulence scatters sunset and rise a few hours before light and reduces the LM. Those dawn. Like the moon, zodiacal light precious few photons will be de- is easily avoided and simply results flected away from a single retinal cell, in a shorter observing window at and the eye will fail to detect a star, night (International Dark-Sky Asso- even though the night is dark and ciation 2000). pollution-free (Bortle 2001). There- Airglow is an important consid- fore, LM estimations will integrate a eration at dark-sky locations. This measure of atmospheric stability, results from the excitation of air when perhaps we are less interested molecules in the upper atmosphere in its effects than that of scattered that emit faint light. Airglow varies light or air pollution. Observers often with solar activity, and tends to be 50 The George Wright FORUM highest during the solar maximum class 1 skies are so rare now that few (which varies on an 11-year cycle; have ever seen them. the most recent was in 2001). Lastly, The Bortle scale definitions are galactic light and starlight can be a included and cross–referenced with significant enough light source that LMs (Figure 3). The Bortle scale is they can affect both the eye’s ability suitable for a wide range of condi- to see faint objects and the brightness tions, from the brightest urban areas of the sky itself. Star counts within (LM lower than 4) to the darkest the Milky Way are more difficult due sites (LM up to 8)—an advantage to the glowing background of the over the LM star count method.