Fuzzy astronomical seeing nowcasts with a dynamical and recurrent connectionist network

We assess a neural-based method for fuzzy astronomical seeing prediction, based on known meteorological variables at the same time-point. This multiple regression (or 'nowcasting') will allow the modern telescopes to be preset, a few hours in advance, in the most suited instrumental mode. The data used are extensive meteorological and seeing (observing quality) measurements partly made at Cerro Paranal in Chile, site of the Very Large Telescope (VLT). Exploratory data analysis was carried out to explore the internal relationships in the data. Then, a time- and space-recurrent network is used in combination with a novel but simple fuzzy coding approach to capture the temporal regularities of the seeing series. Such a connectionist network is endowed with an internal dynamic by means of arbitrary recurrent time-delayed connections. We devote considerable attention to the way we coded the input data. The performance of the connectionist model is appraised and the results are compared with a k-nearest neighbors discriminant analysis method.