The collection of detailed and accurate information about marine habitats and flora species is crucial for mapping, monitoring and management of marine and coastal environments. Remote sensing is widely used to collect information at marine environments, while in recent years the potential use of UAS for mapping is examined. The aim of this paper is the creation of a prediction model for the optimal flight windows of UAS, using the programming language R. The methodology examines several limitations of UAS data acquisition over coastal areas, related to environmental conditions, mainly due to weather and sea state. A theoretical protocol that summarizes the parameters that affect the quality of aerial data acquisition, was created. These parameters are related to the weather conditions (wind, temperature, clouds etc.) and oceanographic phenomena (waves, turbidity, sun glint etc.), prevailing in the study area during the UAV flight. The protocol for the collection of accurate and reliable geospatial information in coastal and marine areas using UAS will be a useful mapping tool for the coastal zone mapping. The produced prediction model will act as a versatile computation approach to different input variables and therefore can be used widely. The input variables of this model refer to weather conditions prevailing in the area of interest and measurements of oceanographic parameters. The result of the prediction model is the optimal flight windows for the collection of accurate and qualitative marine information, in a region of interest.