CCD measurements of night sky brightness - Light pollution in italy

We invite advanced amateurs astronomers to collaborate in a scientific project of the International Dark-Sky Association - Italian Section already started in a number of countries. The project has the aim to collect a large number of measurements of night sky brightness in the primary astronomical photometrical bands together with extinction data. Amateurs which can have at their disposal a CCD device mounted on a small telescope and one, or more, standard U, B, V, R filters could make accurate measurements of sky brightness together with extinction in a large number of nights and, if their instruments are transportable, in many different sites. This will allow to obtain for each site the relation between atmospheric conditions and sky brightness, useful for many studies of light pollution. We would like to create many sub-groups in each country, each one with an experienced person who can give support and suggestions to the others, and discuss with them.

I. INTRODUCTION

We invite advanced amateurs astronomers to collaborate with us in a scientific project which has the aim to collect a large number of measurements of night sky brightness around the World in the primary astronomical photometrical bands together with extinction data. The project is part of scientific activities of the International Dark-Sky Association - Italian Section and is already started in a number of countries.

Many studies of light pollution and artificial sky brightness require large quantities of measurements of sky brightness that, in order to be useful, need to be associated to the knowledge of the atmospheric conditions during the measurements. The measure of vertical extinction is one of the simpler way to evaluate the aerosol content of the atmosphere.

We propose to advanced amateurs which can have at their disposal a CCD device mounted on a small telescope and one, or more, standard U, B, V, R filters to participate to our group of study making accurate measurements of sky brightness together with extinction in a large number of nights and, if their instruments are transportable, in many different sites. This will allow to obtain for each site the relation between atmospheric conditions and sky brightness. Measurements require registration of date, time, sky position in both celestial and altazimuthal coordinates. Could be useful if measurements will not be limited to zenith but will cover the entire hemisphere with one of the common sampling schemes.

Few experienced people who already made this kind of measurements will be happy to discuss the techniques to use and the results with who is willing to collaborate, so that results will be obtained with the same technique by all observers. We would like to create small sub-groups, each one with an experienced person who can give support and suggestions to the others, and discuss with them.

We plan to publish results in a professional journal with a paper at which all active observers will take part as authors, if resulting measurements will have an adequate level and will be in a sufficiently large number.

For information about sub-groups in your Country or any question, please contact Fabio Falchi falchifa@tin.it . Scientific discussions between observers will take place, if necessary, on Magnitude-6, mailing list http://www.gea.cesca.es/magnitude6/index.html . The more recent updates of this observation protocol will be published on the web site Light pollution in Italy http://www.pd.astro.it/cinzano/. Reports on the state of the work will be published on the IDA Newsletters.

II. OPERATING TECHNIQUE

Using a CCD, a small telescope and standard photometric filters you may follow this procedure to obtain the night sky brightness:

· It is not necessary that the night be particularly clear, in fact our aim is to obtain several brightness measurements taken in different transparency condition. The only request is that condition be constant during the measures and without clouds, veils and haze. The meteorological condition should remain constant and the Moon should be well under the horizon (h< -10°) during the measures.

· Possibly avoid sites with lighting installations at distance under about 500 m. Inside cities a unlighted park could be a good choice.

· Choose one or more sky zones, always including the zenith, where to measure the brightness. For example: the zenith, 8 zones at 45° altitude equally spaced in azimuth along the horizon and 12 zones at 20° altitude every 30° in azimuth.

· Choose several (a dozen, if possible) photometric standard stars (e.g. Johnson 1963 or, if you need fainter stars, Landolt 1992) having different altitude (from near the zenith to below 30° altitude). Particularly useful are the stars above 60° and those below 30°. Possibly choose a star near every zone where the brightness will be measured.

· Take exposures of standard stars and determine their altitudes (e.g. using a planetarium software later). Take exposures in the chosen sky zones. If the time used to make the exposures is so long to suspect a change in sky transparency you should measure again the standard stars counts in order to determine a second extinction coefficient. Then take the average of the two.

Take care to record the observing site latitude, longitude and altitude above sea level as accurately as possible (an accuracy better than 15” would be appreciated), the date, the duration and time of each exposure (check if the CCD camera control software does it automatically), alt-azimuth and equatorial coordinates of the measured sky zones and stars (you could do this later using a planetarium software and knowing exactly the sky zone or the star exposed and the time). Keep a note on the meteorological conditions.

b) Get reference flat frames and the dark frames for the exposure needed on the stars.

a) from every raw image subtract the dark frame taken with the same exposure time (the bias frame is assumed to be included in the dark frame)[2]. From the flat frame (It should be the average of several exposures) subtract the appropriate dark. Divide the dark-subtracted raw image by the flat normalised to the mean value (usually this normalisation is included automatically in the standard action of ‘divide by flat’ in the CCD control software)

b) On the star exposures measure the total counts of the star in an area surely covering the entire star image (often the outer portion of a star image seems black even if it contain counts). Subtract the sky counts on an equal area of the same image where there is only the background sky. Scale the counts to the time of one second.

c) In the sky images measure the total counts of the pixels in an area where no star are. Calculate the total sky area in arcsec² covered by these pixels. Scale the total sky counts in that area to that in a square arc second. Scale this number to one second exposure.

· Get the instrument photometric scale factor and extinction coefficient in this way (for every band studied):

a) For every standard star, calculate the air masses: x=1/cos z, where z is the zenith distance. In case of very low altitudes (z>70°) use this more accurate formula: x= (1 - 0.0012tan² z) sec z. Then calculate the variable y=m_cat+2.5log I_star, where m_cat is the magnitude of the star and I_star the counts in a second.

b) Graph y over x and obtain the best fit line y = a + bx . It is better to compute not only a and b, but their errors too (e.g.fig.1).

The photometric scale factor C is: C=a and the extinction coefficient[3] k is: k = - b

· Compute the brightness of the measured sky zones: m_sky = C – 2.5logI_sky, where I_sky is the sky count in a second in a square arc second.

· Report your measurements to your sub-group or to IDA-Italian Section (at the addresses of the authors) including all the necessary information, such as: the telescope aperture and CCD used, geographical position (latitude, longitude and altitude) of observing site, date and time, extinction coefficient in each band measured, alt-azimuth and equatorial coordinates of the measured sky zones and their brightness in each band measured. Please, include also an estimation of the magnitude of the fainter stars visible in the sky images and the size of the area used for the brightness measures. If available, give the solar cycle phase too.

FIGURE 1. In this example the extinction coefficient k =0.385 in magnitude per air mass and the photometric scale factor C = 16.19 magnitudes.

If you like to estimate the artificial sky brightness by yourself, you should transform the data from magnitudes/arcsec² to photonic radiance in ph s^-1 cm^-2 sr^-1 (or in luminance in cd/ m^-2, for V band only) and then subtract the natural sky brightness in the same units. Translation formulae have been given by Garstang (1986, 1989) in B band:

b [ph s^-1 cm^-2 sr^-1] = 10^{((41.438-V[mag/arcsec^2]) / 2.5)} or b [cd m^-2] = 10^{((12.603-V[mag/arcsec^2]) / 2.5)}

Formulae for other bands are under study. The natural sky brightness depends from many parameters, like e.g. the geographic position, the zone of the sky measured, the solar activity. It will be determined from measurements taken in the darkest sites. For a preliminary estimate we can assume approximately 22.8 mag/arcsec² in B and 21.7 mag/arcsec² in V.

IV. CORRECTION FOR DARKER SITES (OPTIONAL)

The sky brightness measured in darkest sites require a correction. Using a small field detector, such the CCD, some of the integrated star light is excluded from the measure of the sky background brightness. In fact we exclude from the counts all the stars brighter than the dimmest visible in the sky exposure (we choose a sky portion on the image where no stars are visible). In order to refer the sky brightness to that perceived with the naked eye, we need to take into account the light of the star fainter than sixth magnitude up to the star brighter than the dimmest on the CCD sky image. The correction is negligible where the natural brightness is a small fraction of the total (i.e.: where light pollution is high).

Even if integrated star light varies with the sky zone, depending on the galactic latitude, we can approximately correct our measurements adding the average contributes of the stars excluded from the measure. E.g. if the fainter stars included in the measurement of the sky counts are of magnitude 13, you have to add the contributes from magnitude 6 to 12. In table 1 the typical contributes at mean galactic latitude are given. They may be three times higher on the Milky Way and three times lower at the galactic poles. State clearly if you made this correction or not and always report the original uncorrected sky brightness.

V. REFERENCES

Johnson, H.L., 1963, in Basic Astronomical Data, ed. K. A. Strand, Univ. Chicago Press, p.204

Magnitude	Integrated star light (10^-6 cd/m²)
7	4.41
8	4.76
9	5.18
10	6.09
11	6.09
12	6.51
13	6.23
14	5.95
15	5.67
16	4.83
17	3.85
18	3.08
19	1.82
>19	3.15

TABLE 1. Typical contributes of stars of each magnitude interval to the integrated star light (obtained from data in Allen 1976).

[1] International Dark-Sky Association, 3225 N. First Avenue, Tucson, AZ 85719-2103, ida@darksky.org.

The IDA website is http://www.darksky.org

[2] If your CCD allows bias subtraction, you do not need that the dark be of the same exposure time. You can, subtract the bias, rescale it to the chosen exposure time and sum the bias again. However, to reduce the noise, the dark should be of a comparable exposure time to the chosen one. An average of more dark of the same exposure time would be better.

[3] In fact, the star apparent magnitude “under the atmosphere” is m_app = m_cat + k/cos z = C –2.5log I_star where k is the extinction coefficient in magnitude per air mass and C is the photometric scale factor. From this formula we obtain that C – k/cos z = m_cat +2.5log I_star, so that y = C – kx. The photometric scale factor is C = a and the extinction coefficient is k = - b

THE IDA WORLDWIDE PROJECT FOR CCD AMATEUR MEASUREMENTS OF NIGHT SKY BRIGHTNESS

Pierantonio Cinzano

Fabio Falchi

ABSTRACT

I. INTRODUCTION

II. OPERATING TECHNIQUE

IV. CORRECTION FOR DARKER SITES (OPTIONAL)

V. REFERENCES