index.html

    We have conducted a new campaign to measure the seeing at the site of the Observatorio Astronómico Nacional at San Pedro Mártir, this time with a DIMM instrument. The results obtained during 120 nights, over a period of almost three years, yield a median seeing of 0.60 arcsec and a first quartile of 0.48 arcsec. These measurements were made 8.3 m above the ground and with exposure times of 6 ms.

    We show that the seeing can be excellent and very stable for whole nights, with the best measurements yielding a median of 0.37 arcsec and a first quartile of 0.32 arcsec during more than eight hours of continuous observations.

    The current results are in very good agreement with our previous study of the site. The expected value of the median seeing 15 m above the ground and extrapolated to null integration time is 0.61 arcsec. Finally, San Pedro Mártir is compared with those major astronomical sites in the world where seeing has been measured with DIMM instruments. This comparison allow us to conclude that San Pedro Mártir is one of the best astronomical sites in the world.

    For this study a two-aperture DIMM from LHESA Electronique was used. This instrument, described by Vernin & Muñoz-Tuñon (1995), is very similar to the ones employed at La Palma. It consists of a 20 cm Schmidth-Cassegrain telescope on an equatorial mount with automatic guiding capabilities, a diaphragm with two 60 mm diameter apertures whose centers are separated by 140 mm, an optical wedge in one of these apertures, an intensified CCD camera, and a PC-compatible computer equipped with a frame grabber. The resulting double image of the same star is captured by the camera and sent to the computer via the frame grabber. By computing the variance of the diferential motion of the double image, in both the parallel and perpendicular directions with respect to the apertures, two independent values of seeing are obtained. The advantage of this diferential technique is that the erratic motion of the image, produced by the instrument (due to tracking errors, wind shaking and bad focusing among other things) is canceled out.

    The DIMM telescope and part of the electronics were located at the top of a concrete pedestal, at the same site where the Carnegie Monitor was installed during our previous campaign (Echevarría et al. 1998); the diaphragm of the telescope was placed 8.3 m above the ground. The computer and the rest of the electronics were located inside a small hut, well detached from the pedestal.

    Each data point originated from the processing of a sequence of 200, 6 ms exposure time frames. By including the associated processing time, we obtained a pair of airmass corrected seeing measurement every 14 seconds. In all cases, each data point was calculated as the average of the simultaneous seeing measurements. When these values differed by more than 12% -- the expected relative error for the DIMM (Muñoz-Tuñon, Vernin & Varela 1997) -- the data point was discarded.

    Measurements with zenith distances larger than 60 degrees were also discarded. Nights with less than 100 measurements or with fewer than one hour of observations (eight) were not considered in this analysis. Data from a total of 106 observation nights were collected from August 18, 2000 through June 23, 2003. Most of them were obtained simultaneously with our ongoing astronomical observing runs, which explains their non-uniform time-distribution. In addition, and in order to improve our seasonal coverage in Autumn, we have included -- with their kind permission -- 14 nights observed in December 2000 and October 2001 by Conan et al. (2002). These measurements were obtained with the same DIMM and at the same height. However, their December measurements were done with an integration time of 10 ms, not 6 ms, the time used in all our observing runs and in their October 2001 run. Consequently, each of the December 2000 seeing values was multiplied by 1.06, in order to convert it into its approximate 6 ms equivalent. This factor was estimated from our results with the interlaced exposures method described by Tokovinin (2002), and from the null exposure time corrections reported in Giovanelli et al. (2001) and Ilyasov (2002).

Click on the next image in order to get the plot and the seeing statistics for each measured night. You will also have access to all the raw data files.

Click on the next image to see the total seeing distribution and how it changes every month and during the four seasons.

If you want to download everything quickly, here you can get all the the zipped raw data (3.5MB) and graphs in gif (2MB) and pdf (7.3MB) format.

CONCLUSIONS

From an almost three-year long program, consisting of 120 nights of unevenly distributed DIMM measurements at the site of the Observatorio Astronómico Nacional in San Pedro Mártir, we conclude:

The overall median seeing and 1st quartile at 8.3 m and with 6 ms integration-time have values of 0.60 and 0.48 arcsec, respectively. Making San Pedro Mártir one of the best astronomical sites in the world as seen in the previous table.
It is not uncommon for the seeing to remain very stable for whole nights, the best results yielding a median of 0.37 arcsec for more than eight hours of continuous observations.
We found a substantial seasonal variation of seeing, in general accordance with Echevarría et al. (1998).
Summer, with a median of 0.55 arcsec, is excellent; Spring and Autumn, with median values around 0.62 arcsec, are very good, the latter slightly better than previously reported; Winter, with a median of 0.78, is not as good, and even worse than previously reported. These results might be influenced by low-number statistics, specially those for Autumn and Winter, seasons that should be better sampled.
Extrapolating our result for the yearly median seeing, we estimate a null exposure time value of 0.61 arcsec at 15 m above ground level.