Measurements of night sky brightness at the Brno observatory


The night sky brightness is an important characteristic of a astronomical site. Unfortunately, night sky brightness measurements can rarely be seen. This page aims to provide raw data which interested individuals or parties might convert to night sky brightness measurements. Its evolution is of special interest, not only because of the light pollution law in the Czech Republic.

Accidental reader might want to read several papers on night sky brightness:


The images were taken at the Nicholas Copernicus Observatory and Planetarium (longitude: 16° 35' 02.2" E, latitude: 49° 12' 14.9" N, 338 m above the sea level) using 0.4 m (f = 1750 mm) Newtonian telescope with SBIG ST-7 CCD camera (765x510 pixels) read out in 2x2 binning with resulting scale of 2.17 arcsec/pixel. The filter wheel is equipped with BV(RI)C filters and a clear hole. The vast majority of analysed images were taken through V(RI)C filters.

Standard dark-substract and flat-fielding techniques were applied, the photometry and matching was done by muniphot and munimat, parts of MUNIPACK. Raw images are fed to an automatic pipeline written in Perl. The images were processed using the following configuration file:

READ NOISE (ADU; 1 frame)  = 15.00
GAIN (e-/ADU; 1 frame)     = 2.30
LOW GOOD DATUM (in sigmas) = 7.00
HIGH GOOD DATUM (in ADU)   = 65535.00
FWHM OF OBJECT             = 3.00
LS (LOW SHARPNESS CUTOFF)  = 0.20
HS (HIGH SHARPNESS CUTOFF) = 1.00
LR (LOW ROUNDNESS CUTOFF)  = -1.00
HR (HIGH ROUNDNESS CUTOFF) = 1.00
WATCH PROGRESS             = 0.00
FITTING RADIUS             = 2.00
PSF RADIUS                 = 15.00
VARIABLE PSF               = 0.00
FRACTIONAL-PIXEL EXPANSION = 0.00
ANALYTIC MODEL PSF         = 1.00
EXTRA PSF CLEANING PASSES  = 5.00
PERCENT ERROR (in %)       = 0.75
PROFILE ERROR (in %)       = 5.00
IS       INNER SKY RADIUS  = 20.0
OS       OUTER SKY RADIUS  = 30.0
The treshold is different for each field and is given in the output file. The aperture radius for comparison star measurements was 7.0909 pixels while typical FWHM is between 2.5 and 3 pixels.


The output file contains following information:

ColumnDescription
01Field name.
02Right Ascension (J2000). Hours, minutes and seconds are separated by colons.
03Declination (J2000). Degrees, arcminutes and arcseconds are separated by colons.
04Geocentric Julian date.
05Sky brightness in counts per pixel. Acquired from the muniphot output.
06Error of sky brightness (in counts).
07Filter.
08Exposition in seconds.
09Instrumental magnitude of the comparison star. If no_match is present, the pipeline failed to identify stars in the image.
10Error of instrumental magnitude of the comparison star. If N/A is present, the pipeline failed to identify stars in the image.
11Absolute magnitude of the comparison star in the given filter taken from literature (mostly by Henden). If N/A is present, absolute magnitude of the star in the given filter isn't available.
12Identification of the comparison star using either GSC number or a MEDUZA chart letter. In one case, the comparison star is denoted using its equatorial coordinates.
13Number of detected stars in the image. If the total of stars gets over 200, the field is crowded. Should be treated with respect to column 14.
14Treshold used for image processing. When lower than 3.0, spurious star detections can occur.
15Observer. OP = Ondrej Pejcha, P = Petr Sobotka, JS = Jan Skalicky.


A short guide how to calculate sky magnitude from data on this page

This guide is largely based on email by F. Hroch posted to [CCD-CR] mailing list.

Assuming that the field is at the zenith, the intensity of the comparison star can be counted as follows:

I_star = 100.4(25 - column_09)

The intensity of sky brightness per square arcsecond is:

B_sky = column_05 / 2.172 (1 pixel = 2.17 arcsec)

The sky magnitude is then:

m_sky = column_11 - 2.5*log(B_sky/I_star)

The reader should perform his/her own extinction correction as observations of stars at different airmasses in single night are given. The filter colour coefficients are still subject to investigation, for V and IC are pretty small, anyway. If you want the most up-to-date values, please contact me.


Data policy: The measurements are "freeware", but please let me know if you download it or "even" use it.