Limiting the impact of light pollution on human health, environment and stellar visibility

https://doi.org/10.1016/j.jenvman.2011.06.029Get rights and content

Abstract

Light pollution is one of the most rapidly increasing types of environmental degradation. Its levels have been growing exponentially over the natural nocturnal lighting levels provided by starlight and moonlight. To limit this pollution several effective practices have been defined: the use of shielding on lighting fixture to prevent direct upward light, particularly at low angles above the horizon; no over lighting, i.e. avoid using higher lighting levels than strictly needed for the task, constraining illumination to the area where it is needed and the time it will be used. Nevertheless, even after the best control of the light distribution is reached and when the proper quantity of light is used, some upward light emission remains, due to reflections from the lit surfaces and atmospheric scatter. The environmental impact of this “residual light pollution”, cannot be neglected and should be limited too. Here we propose a new way to limit the effects of this residual light pollution on wildlife, human health and stellar visibility. We performed analysis of the spectra of common types of lamps for external use, including the new LEDs. We evaluated their emissions relative to the spectral response functions of human eye photoreceptors, in the photopic, scotopic and the 'meltopic’ melatonin suppressing bands. We found that the amount of pollution is strongly dependent on the spectral characteristics of the lamps, with the more environmentally friendly lamps being low pressure sodium, followed by high pressure sodium. Most polluting are the lamps with a strong blue emission, like Metal Halide and white LEDs. Migration from the now widely used sodium lamps to white lamps (MH and LEDs) would produce an increase of pollution in the scotopic and melatonin suppression bands of more than five times the present levels, supposing the same photopic installed flux. This increase will exacerbate known and possible unknown effects of light pollution on human health, environment and on visual perception of the Universe by humans. We present quantitative criteria to evaluate the lamps based on their spectral emissions and we suggest regulatory limits for future lighting.

Highlights

This work presents a new additional way to limit light pollution effects. We summarize light pollution consequences on health, environment and astronomy. We summarize the technical prescription to limit light pollution. We found that a limit to the short wavelength blue light should be implemented. New white LEDs pollute the blue part of the spectrum and should be regulated.

Introduction

Light pollution is the alteration of natural light levels in the night environment produced by introduction of artificial light. Due to the continuous growth of nighttime artificial lighting, this problem is increasingly debated and many localities have developed regulations to constrain the wasteful loss of light into the sky and environment.

The expanding use of light at night is due to the fact that humans are diurnal animals that are trying to extend activities into the usually dark hours. This increasing use is driven by what seems common sense, and by the lighting industry with justifications that at first may seem correct. With few exceptions, everything we build is lit at night. This includes streets, roads, bridges, airports, commercial and industrial buildings, parking lots, sport centers and homes. Outdoor lighting continues to expand as more infrastructure is built. Lighting levels are often set high with one or more secondary objectives in mind. For instance, building exteriors are often lit for a merely aesthetic effect. Shopping centers are typically heavily lit to attract shoppers and create a lively environment designed to stimulate spending. Lighting levels in public areas are often set high as a deterrent against crime, even though studies have not proven this to have any effect on crime rates (Marchant, 2004, Marchant, 2005, Marchant, 2006). Indeed the cores of our urban centers are bathed in light and the resulting light pollution can extend more than a hundred kilometers out from the city’s edge.

There is reliable evidence that this artificial extension of the day produces serious adverse consequences to human health and environment.

The impact of light pollution on the night sky has been described in depth by Cinzano, Falchi and Elvidge (Cinzano et al., 2001). In the First Atlas of Artificial Night Sky Brightness they showed that more than 60% of world population lives under light polluted skies (99% of the population of USA and Europe) and almost one-fifth of world terrain is under light polluted skies.

In regards to human, to date there are no doubts that exposure to light at night (LAN) decreases pineal melatonin (MLT) production and secretion and are not only a source for phase shift in daily rhythms. Apart of timing and exposure duration, the two light variables responsible for the suppression of MLT production are: 1) light intensity and 2) wavelength. Therefore, it seems that the combination of both variables should be considered for the threshold of LAN. Light intensity levels found to suppress MLT production are decreasing as research progresses. During the eighties of last century, it was shown that bright light at an order of thousands of lux was requested for abolishing the secretion (Lewy et al., 1980). The discovery of a novel photoreceptor, the Non Image Forming Photoreceptors (NIFPs), and the photopygment melanopsin gave an opportunity for a better understanding of light perception by humans and showed the effects of spectrum in the human high response to LAN exposure (Thapan et al., 2001, Brainard et al., 2001, Hankins and Lucas, 2002, He et al., 2003, Berman and Clear, 2008, Leonid et al., 2005). The results of a study (Cajochen et al., 2005), in which the impact of wavelength on humans was assessed by measuring melatonin, alertness, thermoregulation and heart rate draw the attention to the significant role of wavelength. It was shown that exposure of 2 h to monochromatic light at 460 nm in the late evening significantly suppressed melatonin secretion while under the same intensity, exposure timing and duration but at wavelength of 550 nm such effects were not observed. Already Wright et al. (2001) showed that even illuminance as low as 1.5 lux affects circadian rhythms. Moreover, recently it as shown that bedroom illumination, typical of most homes in the evening, is sufficient to reduce and delay MLT production (Gooley et al., 2011). From the results of these studies it can be noted that MLT suppression by LAN is wavelength depended and intensities can be much lower than those used several decades ago.

Alteration of the circadian clock may cause performance, alertness, sleep and metabolic disorders. Exposure to light at night suppresses the production of the pineal hormone melatonin, and since melatonin is an oncostatic or anti-carcinognenic agent, lower levels in blood may encourage the growth of some type of cancers (Glickman et al., 2002, Stevens et al., 2007, Kloog et al., 2008, Kloog et al., 2009, Bullough et al., 2006, Haim et al., 2010). MLT seems to have an influence on coronary heart disease (Brugger et al., 1995). LAN acts directly on physiology, or indirectly by causing sleep disorders and deprivation, that may have negative effects on several disorders such as diabetes, obesity and others (Haus and Smolensky, 2006, Bass and Turek, 2005). For a brief review of physiological, epidemiological and ecological consequences of LAN see Navara and Nelson (Navara and Nelson, 2007).

Therefore, the increase in light intensity on the one hand and the wide use of “environmentally friendly bulbs” with a short wavelength emission on the other, are probably having severe negative impact on health through the suppression of MLT production.

In the natural environment, animals and plants are exposed to light at night levels that vary from about 5 × 10−5 lux of the overcast sky, to 1 × 10−4 lux by the starry sky on a moonless night, to 2 × 10−2 lux at the quarter moon, to 0.1–0.3 lux during the week around full moon. The artificial light of a typical shopping mall, 10–20 lux, is up to 200 thousand times brighter than the illuminance experienced in the natural environment around new moon. No wonder that it has become apparent that light at night has strong environmental effects in behavioral, population and community ecology (in foraging, mating, orientation, migration, communication, competition, and predation) and effects on ecosystems. For a review of ecological consequences of light pollution see (Navara and Nelson, 2007, Rich and Longcore, 2004, Rich and Longcore, 2006, Longcore, 2010, Kempenaers et al., 2010). This strong evidence of the adverse effects of artificial light at night on animals and on human health should be balanced against the supposed positive effects on safety and security.

Fortunately it is possible (and also simple in theory, if those involved in lighting collaborate) to limit the light pollution effects and, at the same time, allow for the lighting that is usually perceived as a need by people. Practical ways to limit the effects of light pollution on the night sky and the night environment are well known and verified (Cinzano, 2002):

  • a)

    Full Shielding: Do not allow luminaires to send any light directly at and above the horizontal, with particular care to cut the light emitted at low elevations (in the range gamma = 90–135° above the downward vertical, i.e. 0–45° from the horizon plane). In practice, light in this range travels long distances through the atmosphere and enhances the additive property of light pollution (Cinzano and Castro, 2000, Luginbuhl et al., 2009), an effect that compounds the problem, especially in densely populated areas. An additional limitation on the light leaving the fixture downward (in the range gamma = 80–90° from the downward vertical, i.e. 0–10° below the horizon plane) should also be enforced. This is because the nearly-specular reflection of asphalt at grazing incidence considerably increases the amount of light at low angles above the horizontal (although this reflected light is much more subject to screening by surrounding vegetation and buildings). This limitation will also improve the comfort and visual performance of road users by lowering the direct glare from fixtures.

  • b)

    Limiting the Area of Lighting: Carefully avoid wasting downward light flux outside the area to be lit. Such waste is not only a main cause of increase of installed flux per unit surface (and in turn a main cause of increase in energy expense), but some of this light is also reflected upward from these surfaces. Even if Lambertian diffusion from horizontal surfaces is less effective in sending light at low elevations than direct emission by luminaires, nevertheless when the direct emission is eliminated, the diffuse reflection remains as an appreciable source of pollution.

  • c)

    Eliminate Over lighting: Avoid luminances or illuminances greater than the minimum required for the task, and dim lights when the application allows it.

  • d)

    Shut Off Lights When Not in Use1: It makes sense to turn the lights off when you leave the room, or for the lights to turn off automatically, but in outdoor lighting these options are rarely available (in Italy, for example, almost all the parking lots of shopping malls are lit all night long, and likewise for the industrial/artisan/commercial areas, whether or not there are workers at night).

  • e)

    Limit Growth in Installed Lighting: Limits to the increase of the new installed flux should be implemented. A 1% yearly increase could be allowed at first for each administrative area, followed by a halt in the increase of total installed flux, and then by a decrease. This does not mean that no new installation will be allowed, but simply that if you want to install new lights you have to decrease the flux in the existing overlighted areas.

To these basic prescriptions, some others could substantially improve lighting quality (e.g. a requirement that the lighting installation be designed by a professional lighting designer, although this might not be feasible in poorer countries nor advisable for smaller installations, provided they respect the code) or to take account of specific kinds of installations (e.g. signs or historical buildings). Most of these prescriptions are already implemented in some of the most advanced anti-light- pollution laws such as Lombardia (Italy) Regional Law n.17 of March 27, 2000 with its subsequent additions and modifications. Twelve other similar regional laws followed in Italy, and most Italian territory and population are now protected by these laws. Slovenjia adopted a similar law in year 2007. Falchi (2011) found that despite an almost doubling in the outdoor installed flux, in two studied sites in Lombardia, the artificial sky brightness did not increase over the last twelve years. This is probably due to the adoption of laws against light pollution in the surroundings of the sites. A full enforcement of the prescriptions could probably make a substantial improvement in the quality of the night sky and environment. In fact, the same research shows that in six studied sites, on average, 75% of the artificial sky brightness is produced by light escaping directly from fixtures and only 25% from the reflections off lighted surfaces. This implies that, all the rest being equal, a complete substitution of the installed fixtures with fully shielded ones could lower the artificial sky brightness to 1/4 of present levels. In two of the studied sites, more than 90% of the artificial sky brightness derived by direct light. These sites would presumably have a 90% decrease in light pollution as a result of retrofitting fixtures to fully shielded in the surrounding territory that produce light pollution, i.e. a circle of at least 100 km radius.

Nevertheless, even when the best control of the light distribution is reached and when the proper quantity of light is used, some upward light emission remains, due to reflection from the properly lighted surface. This is an unavoidable by-product of the lighting operation, even when measures a), b) and c) have been achieved: lighting is installed just to produce reflections of light. However, after the light has performed its useful function, it is then dispersed into the environment. Due to its near-Lambertian behavior, this reflection is frequently less effective at low elevations than at large elevations, so the effect on the night sky tends to be confined largely to the vicinity of the source. In any case, the environmental impact of this residual light pollution cannot be neglected.

Limitation of this residual pollution requires limits not only on “how” nighttime lighting is arranged according to prescriptions a) and b), but also “how much” nighttime lighting is made. Typically it has been proposed to limit the growth rates of installed flux in each city, or to limit the average density of installed light flux (e.g. installed flux per hectare or acre). However, following the example of the radio portion of the electromagnetic spectrum, there is an additional way to limit this residual pollution: by preferential use of light sources with spectral characteristics that have the least impact on star visibility and human and wildlife health, while maintaining a given degree of visibility in areas that need artificial lighting. This would allow reduction of the negative astronomical and biological effects without impairing essential night lighting.

This solution has been applied for decades whenever Low Pressure Sodium (LPS) and High Pressure Sodium (HPS) lamps have been requested in place of Mercury Vapor (MV) or incandescent lamps. The arrival of new LED light sources for nighttime outdoor lighting and widespread use of broad spectrum Metal Halide (MH) lamps even where they aren’t the best option enhances the need to define a more quantitative prescription, applicable to any kind of lamp and capable of giving precise indications to the lighting industry on the way to proceed in light source development or improvement (e.g. how to filter or tailor the spectrum of the emitted light).

The prescription should:

  • (i)

    be as effective as possible in protecting the night environment from the adverse effects of light pollution;

  • (ii)

    take account of existing nighttime lighting habits in order to minimize the impact on human activities:

  • (iii)

    allow easy identification of non-compliant light sources; and

  • (iv)

    allow easy measurement in the field, when needed.

In this paper we discuss the problem, we recognize two different quantitative parameters, we devise a prescription and we investigate how it could be enforced.

Section snippets

Methods

The possibility of limiting the residual light pollution, avoiding the need to limit nighttime outdoor lighting itself, is based on the different response with wavelength of the two main classes of eye receptors and the action spectrum of circadian rhythm disruption for rodents, monkeys and humans (Brainard and Hanifin, 2005). In a schematic way and for the purposes of this paper, we can distinguish the photopic response of cones and the scotopic response of rods. The eye response is fully

Measurements

Emission spectra were acquired using an ASD, Inc. FieldSpec 3 spectroradiometer equipped with an 8° field of view foreoptic. The instrument had been radiometrically calibrated and spectra were acquired in radiance (W/m2/μm/sr) mode over the 350–2500 nm range. Each lamp was warmed up prior to measurement and the spectra were acquired from one lamp at a time in a dark room. The measured light sources included the following classes 1) liquid fuel lamps, 2) pressurized fuel lamps, 3) incandescent,

Proposed limits

Residual light pollution is that produced by reflected light, after direct upward emission has been accurately minimized, over lighting has been avoided, and the flux wasted illuminating outside surfaces has been minimized. It would remain to be dealt with after laws or regulations have required zero direct emission above the horizontal by lighting fixtures, limited the luminance or illuminance to the minimum required by security rules, minimized as much as possible the fraction of light wasted

Conclusions

In this paper we analysed the different energy and luminous fluxes in the melatonin suppression action spectrum and in the scotopic band for several types of lamps. We found that huge differences exist in the blue emissions of the lamps, for the same photopic luminous flux. Due to the fact that night vision and our health are impaired more by blue light, we proposed two limits to be followed in the adoption of lamps for external use. The first should be used everywhere, as a standard, in order

Acknowledgments

We acknowledge Dr. Barry A.J. Clark and Dr. Jan Hollan for very interesting discussions and suggestions and consequent improvement of this work. We acknowledge also Dr. Steven W. Lockley and Dr. Paul Marchant for help given in their respective fields.

References (57)

  • J. Bass et al.

    Sleepless in America: a pathway to obesity and the metabolic syndrome?

    Arch. Intern. Med.

    (2005)
  • W. Benenson et al.

    Handbook of Physics

    (2002)
  • S.M. Berman et al.

    Past visual studies can support a novel human photoreceptor

    Light Eng.

    (2008)
  • G.C. Brainard et al.

    Photons, clocks, and consciousness

    J. Biol. Rhythm.

    (2005)
  • G.C. Brainard et al.

    Photic regulation of melatonin in humans: ocular and neural signal transduction

    J Biol. Rhythm.

    (1997)
  • G.C. Brainard et al.

    Action spectrum for melatonin regulation in humans: evidence for a novel circadian photoreceptor

    J. Neurosci.

    (2001)
  • J.D. Bullough et al.

    Of mice and women: light as a circadian stimulus in breast cancer research

    Canc Causes Contr

    (2006)
  • SI Brochure Appendix 2. Practical Realization of the Definition of the Candela

    (2007)
  • C. Cajochen et al.

    High sensitivity of human melatonin, alertness, thermoregulation, and heart rate to short wavelength light

    J. Clin. Endocrinol. Metab.

    (2005)
  • The CIE System of Physical Photometry

    (2001)
  • Commission Internationale de l’Eclairage Proceedings 1924

    (1926)
  • P. Cinzano

    In Light Pollution and the Protection of the Night Environment

  • P. Cinzano

    A laboratory for the photometry and radiometry of light pollution

    Memor. Soc. Astrom. Ital. Supplement

    (2003)
  • P. Cinzano

    A portable spectrophotometer for light pollution measurements

    Memor. Soc. Astrom. Ital. Supplement

    (2004)
  • P. Cinzano et al.

    Mem. SAIt - J. Ital. Astronom. Soc.

    (2000)
  • P. Cinzano et al.

    The first world atlas of the artificial night sky brightness

    Mon. Not. Roy. Astron. Soc.

    (2001)
  • J. Copas

    The downside of publication

    Significance

    (2005)
  • Cited by (427)

    • The reliability of satellite-based lighttrends for dark sky areas in Austria

      2023, Journal of Quantitative Spectroscopy and Radiative Transfer
    View all citing articles on Scopus
    View full text