Challenges for optical turbulence characterization and prediction at optical communication sites

Modelling of atmospheric optical turbulence has been of interest in astronomy for several decades, e.g. for site characterization and flexible scheduling. Nowadays, it is also considered for free-space optical communications, namely to conduct site selection and to design future optical communication systems. In this work, a general approach relying on numerical weather prediction simulations in order to perform optical turbulence prediction is presented. The approach makes use of the Weather Research and Forecasting model and raises several challenges. The latter, such as the choice of the C_n^2 models or the required temporal and spatial resolutions, are first discussed with regards to the literature. Then, optical turbulence prediction is conducted for the site of Redu, Belgium, illustrating the different challenges. These predictions are also compared with seeing measurements from a differential image motion monitor. The presented approach offers realistic seeing values that, however, do not follow rapid variations of the measured seeing. Origins of the discrepancies between measurements and predictions are to be found in the modelling of the boundary layer and motivate the use of a C_n^2 model relying on the turbulent kinetic energy. Further simulations and measurement campaigns at other optical communication sites are encouraged in order to refine some model parameters and compare statistically the prediction results.