The Prediction of Seeing and Other Relevant Climatological Conditions

The aim of exible scheduling is to improve the quality of observations and to guarantee eecient use of telescope time. Flexible scheduling implies the necessity of accurate forecasts and nowcasts (the latter term, widely used in aviation, implies near real-time analysis). The Astronomical Site Monitor (ASM) will constitute an interface between the VLT observer and the local environment. Seeing and meteorological data are continually collected at ESO's observational sites, and external meteorological data is being assessed. The climatological quantities of greatest importance for exible scheduling are seeing (at approximately one hour timescales) and water vapor and cloudiness (at approximately 24 hour timescales). A comprehensive range of ongoing studies are concerned with model-ing and prediction. We will describe work carried out, and activities which are being pursued. These include: nowcasting of seeing using statistical and neural network methods; cloud cover and water vapor prediction using observational satellite data; and the development of a meteorological local area model (LAM). The neural network approach even allows policy decisions relating to seeing quality to be incorporated.