Blanco Seeing Statistics, 2000 and before

In the following, 'Night report' refers to the number selected by the observer in the night report among the proposed categories: <0.4; 0.4-0.6; 0.6-0.8; 0.8-1.0; 1.0-1.5; 1.5-2.0; 2.0-3.0; >3.0. Actually, I use and plot the average of these respective categories: 0.4; 0.5; 0.7; 0.9; 1.25; 1.75; 2.5; 3.0. The numbers usually reflect the wavelength of interest of the instrument used (i.e. for example f/14 night reports yield a seeing estimate at about 2 microns).

• Recent f/14 image quality statistics: histogram [2] and night by night [3] (18 nights from Dec99 to May00)
• Recent PF image quality statistics [4] (Feb-Apr 00)
• 4m f/8 image quality statistics [5] (corrected at zenith and 550nm, before/after active optics, 92 to 94, SPIE paper by Baldwin et al.)
• Old Tololo image quality statistics [6] (89-91 from Suntzeff, not corrected to zenith)
• FWHM seeing comparison [7] over 96-97-98-99 from the tower seeing monitor at Tololo. The seeing values are averaged over each calendar month (this is an arbitrary choice). Dots with a ? mean that the number of nights used to average was less than 15 during a single month, and the result isn't very reliable.
• FWHM seeing during 1996 [8]: year median is 0.56" (stdev = 0.28"). Percentiles: 10% at 0.35" and 90% at 1.00" (8034 samples)
• FWHM seeing during 1997 [9]: year median is 0.62" (stdev = 0.36"). Percentiles: 10% at 0.39" and 90% at 1.02" (4985 samples).
• FWHM seeing during 1998 [10]: year median is 0.85" (stdev = 0.45"). Percentiles: 10% at 0.40" and 90% at 1.37" (7853 samples).
• FWHM seeing during 1999 [11]: year median is 0.71" (stdev = 0.43"). Percentiles: 10% at 0.41" and 90% at 1.30" (9499 samples).

Note: all the seeing data presented above was taken with the Carnegie monitor. 9499 samples per year means an average of 26 per night, corresponding to an average of 3 hours continuous sampling per night.