THE CHALLENGE OF MODELLING AND FORECASTING FOG

Fog can severely impede human activities yet fog formation and evolution still prove very difficult to predict. The LANFEX campaign ("Local And Non local Fog Experiment", Sep 2014 - Mar 2016) will provide a unique set of longterm measurements of fog evolution from the Met Office site in Cardington and an area in south-west Shropshire using a distributed network of sensors, automated weather stations, meteorological towers and tether sondes. This project aims at simulating the evolution of fog by concentrating on local processes and the role of surface and large-scale synoptic forcing. The student will contribute to the analysis and interpretation of the longterm data and will use three numerical models to improve our understanding of the physics of fog evolution and the ability to predict fog operationally. The LANFEX data set will be used to compare with model runs and provide a detailed assessment of model performance which will facilitate improvements to the prediction of fog. In particular, it is intended to examine model physics and parameterisations related to initial fog formation and its deepening into thick fog, but all aspects of fog evolution including its breakup will be examined.
First, an established model of fog dynamics, thermodynamics and microphysics (Microphysical FOG model, MIFOG, Bott et al., 1992; von Glasow and Bott, 1999) will be used to simulate several case studies with and without fog development to test the model's performance under different cases. In addition to the synoptic forcing the following parameters are among those that play a key role in fog evolution: time of year, boundary layer height, presence/absence of clouds and tall vegetation, soil temperature and humidity, aerosol number and composition. All these parameters are included in MIFOG and the sensitivity of the various parameters will be investigated in detail. Importantly, MIFOG includes a detailed description of the microphysics of fog which has been shown in the past to play an important role for fog evolution.
A parameterised version of MIFOG, the PAFOG model (PArameterised FOG model, Bott and Trautmann, 2002), in which the detailed fog microphysics have been replaced with a parameterised version will then be used to test how good the performance of this computationally highly efficient model is compared to the more complex model MIFOG and the data. In order to link with the need of the CASE partner and main user of this research, the UK Met Office, the single column version of the Unified Model (UM) will be used and tested against the more specialised models and the field data to benchmark its performance and to provide further guidance to its future development. Promising parameterisations of PAFOG will be incorporated into the UM with the aim of improving the operational forecast of fog. It is furthermore expected that new insights about the dynamical, microphysical and thermodynamic evolution of fog will be gained during this project. Data from various UM runs will be used to drive the model and the student will also investigate differences in fog forecast between 1D and 3D models.
The risks for this project are very low as the field sites for the funded LANFEX project were carefully chosen and the duration of the experiment will ensure that enough fog events will be characterised. All numerical models as well as technical and scientific expertise are available and the commitment of all partners to the project is very strong.