The prediction of fog is a challenging task in operational weather forecast. Due to its dependency on small scale processes, numerical weather models struggle to deal with under scale features resulting in uncertainties on the fog forecast. Unawareness of the onset time and the duration of fog leads to disproportionate impact on open air activities, especially in aviation. Nevertheless, in a small sized country such as Portugal mainland, the fog varies greatly. The traffic of the two busiest Portuguese international airports of Porto and Lisbon is affected by the occurrence of fog in different times of the year. The fog occurrence at Porto is a predominant winter phenomenon and a summer one at Lisbon. A conceptual model supported by observational evidence associate the fog formation in the Tagus estuary followed by its slow advection towards the airport. At Porto the fog formation is highly dependent on local wind distribution, as an indication of the dominant role of local advection. Observational variables and their trend are local indicators of favouring conditions to the fog onset, like cooling, water vapour saturation and turbulent mixing. The dataset of 17 years of half hourly METAR from the airports of Porto and Lisbon are used to diagnose the pre fog conditioning. Two diagnostic models are proposed to assess pre fog conditions. The first model is adapted from the statistical method proposed by Menut et al (2014), which performs a diagnosis from key variables trend, such as temperature, wind speed, relative humidity and base height of very low clouds. Thresholds are defined from the METAR samples in the six-hours period prior to the formation of fog. Due to the local character of fog, the presented thresholds are the most appropriate ones for each airport. The predictability of fog is then assessed using observations. In the second approach will encompass networks such as the Multilayer Perceptron and also Recurrent Neural Networks (RNN), which are especially well suited for time series. By experimenting with different types of RNN, we will try to capture the connection between the temporal evolution of measured variables in the dataset and the fog onset. These experiments will include different time windows to measure its influence in the prediction performance. An ablation study will also be performed to measure the sensitivity of the models not only to the architecture but also to the measured variables in the dataset.