The numerical weather prediction model WRF was used to predict the condition of fog formation. The parametrization of the model in terms of PBL schemes and Noah-MP land surface model was investigated and adjusted to the local conditions in the UAE. Maps of fog probability are generated using WRF outputs.
The climatology of fog formation was studied using years of SEVIRI images at a 15 minutes time intervals. The detection uses amongst others the brightness temperature difference between 3.9μm and 10.8μm channels. An enhanced brightness temperature gradient-based method is used for fog detection over the UAE. Satellite fog detection is done seamlessly in day and night times. Ancillary data from WRF outputs are used to differentiate between fog and Low Stratus to minimize false alarms. The high temporal resolution allow for the determination of event duration, start time and dissipation time. The approach links the brightness temperature gradient to dissipation time and visibility which allows for dissipation and visibility degradation nowcast.
The results of the analysis of the SEVRIR scenes indicate that fog formation occurred most frequently in desert areas, namely: Liwa, Um Az Zumul, and western part of Abu Dhabi Emirate. These areas have also the longest average event duration. Coastal regions and Northern Emirates showed less fog formation and shorter events, in comparison with other areas in the UAE. Fog frequency was the highest during the fall season, especially in November, where the mean monthly fog frequency was 26. The lowest fog occurrence was during the summer season. Amongst all summer months, July had the lowest monthly fog mean of 1.4 events. In the UAE, fog formation occurs mainly in early morning hours (~00:15 am). By looking into Abu Dhabi only, two types of events were observed. First, large-scale events, which starts usually early in the evening. While, local-scale events start approximately between 01:30 am to 02:00 am UTC. Moreover, long-time trend analysis showed an upward and downward trends in temperature and relative humidity, respectively.