Atmospheric Extinction from the CMT


From its observations of about 50 photometric standards per night (pre-1999), the Carlsberg Meridian Telescope provides nightly values of atmospheric extinction in V (r' from April 1999) for La Palma. The data is now split up into yearly files going back as far as 1984.

The data for 1999 and after differ from those taken between 1984 and 1998 because for the first time a CCD camera has been used and a restricted range of zenith distances were observed each night. The extinction value is generated for each frame by using an instrumental zero-point derived from the long-term performance of the CCD ie. an assumption is made that the dust-free extinction value is constant over the long term. The mean for each night is given in the files from 1999 to 2013. Only those CCD frames considered photometric in quality have been used. Each frame has an average of 30 - 40 calibrating stars on it.

La Palma Technical Note No 31 can be used to convert values of extinction at r' to other wavelengths. Simply use the values in Table 1 to determine a differential extinction between r' (or V for the old data) and the passband of interest and apply this to the CMT data.

In August 1993, Derek Jones used the JKT to measure the extinction. Taking into account a 0.03 magnitude offset, to allow for the effect of the Pinatubo eruption, the following broad-band values are appropriate for good dust-free nights:

Passband U B V R I Z
Effective wavelength 365 445 551 658 806 910 nm
Extinction 0.46 0.22 0.12 0.08 0.04 0.05 mag

These generally agree with what was found in La Palma Technical Note 10 (Derek Jones 1984), again with a 0.03 offset. Interpolating from the above, the following values for good quality dust-free nights for other bandwidths have been deduced:

Passband g' r' i'
Effective wavelength 480 625 767 nm
Extinction 0.19 0.09 0.05 mag

Please note that the CCD extinction values (1999+) have been recalculated on 25 June 2001. If using the data, please read the header comments. If a reanalysis is warranted and new values need to be placed in above files, the version number for those values will be incremented.

Click to enlarge Here is a plot showing the V extinction from 1984 to 2010 as measured by the Carlsberg Meridian Telescope. Although the data from 1999 onwards were measured using an r' filter the data has been converted to an equivalent V extinction for this plot. The effect of the eruption of Mount Pinatubo in June 1991 on the extinction is clearly visible. The number of high extinction values that occur every summer is normally caused by Saharan dust.
(Postscript version of the plot).

At the time of the JKT measurements of Derek Jones (see above) in 1984.5 and 1993.7 there was still dust in the atmosphere from various volcanic eruptions and the derived extinction values would not be dust-free values. Thus, an offset of 0.03 magnitudes had to be applied. This offset was partly derived from the graph above and partly from a more detailed analysis of the 1999 data.


Dafydd Wyn Evans / IoA / dwe @ ast.cam.ac.uk
Last update 2 January 2013