5 Discussion

The sites tested span from heavily polluted to dark skies. Caracas and Florence show obviously the brightest values. Nevertheless, in spite of the fourfold factor in population compared to Florence, the typical night sky at Caracas is 0.5 magnitude darker. This is due to the smaller number of lamps per capita, and also because of the moderately high altitude of Caracas (900 m). Mérida, Siena and S. Omero show almost the same light pollution level, once again, with a wide range of inhabitants. S. Omero has a remarkable high level, considering its small population. Located a few kilometers from the town of Teramo, this village is however surrounded by many polluting sources. Probably, its proximity to the Adriatic coast, heavily illuminated at night, contributes strongly to the high level found. Le Tolfe is a residential zone about 4 kilometers away from downtown Siena. Luciana, many years ago an inhabited small village near Florence, is the site of the Torre Luciana Observatory, a small facility operated by the Physics Department of the University of Siena. In spite of its proximity to Florence, and excluding the northern horizon, which is completely flooded by lights from this city, the zenith sky brightness is not so high. El Valle, located 6 kilometers from the town of Mérida, shows a sky approximately one magnitude brighter than the typical dark, natural sky. Finally, the least polluted skies were found on the Italian Alps, at La Balma lodge, and at the Venezuelan Astronomical Observatory in the Andes. Nevertheless, even at these sites, a mild level of light pollution is evident.

Thanks to earlier maps of sky brightness in Italy, calculated by Bertiau et al. (1973), it has been possible to estimate the increase of light pollution over 25 years at the sites tested. The maps produced by these authors show sky magnitudes in the B band. To allow a comparison, the colour index of the polluting light must be known. From photoelectric observations by Walker (1973), carried out in California and Arizona the same year, a colour index B-V = 1.2 can be calculated. Assuming a similar spectral distribution in Italy in 1973, B values from Bertiau were converted to V magnitudes and the comparison followed. This is shown in the last column of Table 2. It is clear that the increase of light pollution in inhabited centers is not directly related to the population growth over 25 years. Recently, Falchi & Cinzano (1998), by means of satellite images, synthesized a new light pollution map of the italian territory. The sky brightness values reported in Table 2 seem to confirm their forecasts.

In a developing country like Venezuela, light pollution is much lower for a given population and indeed common, incandescent lamps are still widely used in rural areas. Unfortunately, no previous sky brightness data is available and measurements at other inhabited centers are needed to determine a reliable luminosity-population relation in this country. However, comparing the values found in Caracas and Mérida with the results of Garstang (1986) for 12 Ontario cities, it is evident that city zenith brightness in Venezuela is nearly 2.5 times lower, indicating an artificial lighting output slightly over 500 lumens per capita.

Figure 2: Spectra of light polluted sky at selected sites in Italy and Venezuela. Wavelengths are given in Angströms. Known emission lines from artificial lighting are labelled HPS (High Pressure Sodium lamps) and Hg (Mercury Vapor lamps). All spectra were taken on Kodak T-MAX film processed for ISO 25000 and the original dispersion is 120 Å/mm

In conclusion, present night sky brightness levels in populated areas are due to the higher light efficiency of modern, inadequately shielded gas discharge lamps as well as their widespread use. On the other hand, the spectral survey carried out at heavily polluted sites shows a similar, if not equal, spectral distribution (Fig. 2), produced by Mercury and High Pressure Sodium lamps.