Light pollution and visual astronomy

I have been an amateur astronomer for most of my life, ever since the late Giancarlo De Carlo, a world-class architect and a good friend of mine, sent me a very nice map of the Heavens he had gotten with a 1970 copy of the National Geographic magazine (see below). 

I was seven years old then, and the starry sky at night was glorious  to stare at with the unaided eye, even from the terrace of my house in the city of Venice. I quite well remember I could see the milky way back then… Now, a third of a century past, such a view would require a massive black-out: I am sure many youngsters have not ever seen our galaxy in all the glory it can display under a moonless, non-light-polluted sky, in fact.

Pollution of our atmosphere by artificial lights might look to you like a ridiculous issue if compared with the problems of our waters, air, and soil: I could not agree more. On the other hand, a more efficient use of luminous energy would have several beneficial effects to the environment. And one of these would be a happier bunch of amateur astronomers.

If you look at a map of the intensity of luminous pollution over north-eastern Italy (see above), you notice that there are by now really few regions close to where I live which have so far been spared from a intense pollution during the night. The picture has been obtained with Google Earth, which allows you to map the nighttime appearance of the Earth from satellites, and with it you get some indication on the amount of light pollution from cities and roads.

Better, however, is the map on the left [P. Cinzano, F. Falchi (University of Padova), C. D. Elvidge (NOAA National Geophysical Data Center, Boulder), copyright 2001 ISTIL, Thiene, Reproduced from www.lightpollution.it], which has been generated by taking into account the combined effect of known sources, secondary scattering, and terrain elevation – including screening from natural obstacles such as mountains – in a model which has been shown to rather accurately predict the darkness of a site at night. The model has however the drawback of a scarce spatial resolution (about 3 km on the ground). Small-scale screening effects are thus ignored, while they can make locally a lot of difference!

Cinzano’s maps are a valuable instruments to visual astronomers, who need to locate the most proficuous sites for deep sky observations – the best compromise between the darkness of the site and the ease of reaching it with their equipment. But they cannot tell the whole story, because of the importance of the screening effects above mentioned. Direct measurements on the ground have to be made.

Enter the Sky Quality Meter (SQM, see picture on the right), a XXIst century instance of the good-old exposimeter our dads used to take a picture. A well-calibrated SQM should tell, with an error not larger than about 10%, the visual magnitude of the sky per square arcsecond: a number on a logarithmic scale which describes the actual darkness of the sky. A reading of 22.0 on a SQM is a virtually perfect sky; a reading of 21.0 is a dark sky with moderate to low light pollution; a reading of 20.0 is already wanting for observing deep sky objects, and lower readings mean one has better watch TV. The SQM is used by pointing to the sky and pressing a button: duh! However, the device has a sensitivity in a rather wide cone (about 80 degrees), so that one has to be careful to avoid including in it parts of the surrounding scenery, which can significantly alter the result. Also, the inclusion of significant parts of the milky way may change the readings by as much as 50%.

So the SQM is a useful tool, but it is not enough. Expert amateurs know how to estimate the visual limiting magnitude of stars, by counting their number in well-defined patches of the sky. A visual magnitude near 7.0 is expected if the sky has a SQM reading of 22, and a rather linear relationship exists between those two numbers. However, the visual magnitude is affected by layers of clouds, even very thin ones, which – if not lighted from below- have no influence on the SQM readings. Transparency of the sky, that is, is a critical factor that has nothing to do with light pollution, although haze catches light impinging on it even from very far sources, and may cause a dramatic brightening of the background.

One way around the small spatial resolution and lack of sensitivity to small-scale screening of Cinzano’s map is to use it in combination with Google Earth to produce an aerial view of a site, overlaid with the color-coding of the terrain. By placing the view about 3000 meters above ground, one thus “looks” at the site from a point of the atmosphere which may receive or not receive light from far sources. You thus get to look at the parts of the surrounding terrain which contribute to lighting up a point at the zenith of the observing site, and the color coding does allow one to understand better how much light will affect the sky darkness. The picture below shows my favourite observing site in the Dolomites, Casera Razzo, from a point above it. You see that far south there indeed are strong sources of light -evidenced by yellow of the terrain, while green is mostly dark and blue is quite dark- but most of them are screened by near mountains (courtesy Mauro Da Lio – source here).

As you can see, amateur astronomers these days have all sorts of toys to play with to plan their observations and discover promising dark sites. In the end, however, the best judge is an instrumental test: you take your favourite telescope, bring it to the site on a clear night, and try a few faint objects. But the variability of sky clarity and observing conditions (among them, the turbulence of the atmosphere, which “defocuses” pointlike sources, decreasing the signal-to-noise and the detection threshold) mean that you will need multiple tests in each site. A full-time job!