A statistical model of telescope image quality

Because of the variability of atmospheric conditions and the great number of possible observing configurations a realistic and reliable assessment of the effect of the local atmosphere on the telescope performance can only be done by a statistical model. Such a model will also allow us to study in a realistic manner the effect of operational criteria for the venting devices of the enclosure.

The model comprises the following elements:

1. The starting point is a large set of continuous night-time wind and natural seeing data measured on the site. Typically the set will have a record with the mean values for each period of 10 or 20 minutes.
2. A random zenithal distance between 0 and 60 and a random azimuth () with respect to the mean wind direction are associated to each data record. This assumption is confirmed by actual statistical data of telescope operations.
3. The parameterizations for seeing and dynamic deformations of the mirror and for the guiding error.
4. The characteristics of the air flow inside the enclosure at all its different orientations configurations.
5. The criteria driving the operation of the wind screen, louvers and the other venting devices.
6. A performance assessment criterion that relates the effect of a disturbing influence to the overall best performance that is achievable in the actual conditions. The notion of Central Intensity Ratio presented in section fulfills this condition as it assess the loss of image quality with respect to the limiting natural seeing.

In the next pages we will illustrate the use of this statistical model using the VLT case as an example.

• Parameterisation of seeing and guiding effects
  • Wind versus seeing for the primary mirror
  • Guiding errors
  • Computation of the CIR figure
• Computation procedure
• Results and discussion
• Synthesis