Dynamical controls on the life cycle of fogMet Office (Atmospheric Processes and Parametrisations – APP)Contact email: A.N.Ross@leeds.ac.uk
Forecasting the onset and duration of fog is an important, and difficult, forecasting problem. Fog causes significant disruption to road traffic and aviation and can be a significant factor in road traffic accidents. Receiving accurate and timely forecasts can help reduce the impact of such weather, minimising disruption, accidents and casualties. The current Met Office Unified Model (MetUM), along with many other models, has a systematic bias to producing fog too often and too quickly compared with observations, and a tendency for the fog to thicken much more rapidly than is observed. Similarly the model can often struggle to accurately capturing the timing of fog break-up. There is significant interest in understanding the processes controlling the formation, development and erosion of fog in order to better represent these in weather and climate models.
The Met Office has recently lead a major field campaign (LANFEX – the Local And Non-local Fog EXperiment) to study the fog lifecycle. Observations were made in two areas – one area of relatively flat terrain around the Met Office observation-based research unit at Cardington, Bedfordshire, and the other in more complex terrain around the Clun and Teme valleys in Shropshire (an area of small scale hills and valleys typical of the UK). Although radiation fog is often assumed to form in situ and is considered as a one-dimensional problem, the observations show that advection plays an important role at both field sites. This can either be large scale advection due to synoptic changes, or small scale advection due to local drainage flows within valleys. Even smaller scale gravity waves and the “sloshing” of the cold air in valleys appear to play a role in the spatial and temporal variability of fog. The effects can either be direct, through the transport of cold air, water vapour or fog droplets, or indirectly through changes in the turbulence and mixing. Figure 1 shows an example of valley fog in the Clun Valley during the previous COLPEX campaign (Price et al 2011).
Figure 1: Fog forming in the Clun valley during the COLPEX campaign. (Photo: Jeremy Price).
The observations suggest the importance of these dynamical processes, but the observations alone are often insufficient to really understand these processes due to the limited spatial distribution of the observations even in an intensive field campaign like LANFEX. There is also open questions about how well models capture these processes. Finally, work so has focussed on the formation and initial development of the fog, with much less focus on the dispersion of fog, although this is an equally important forecasting problem.
This project will make use of the extensive observations from the LANFEX experiment, combined with high resolution (100m) simulations using the MetUM to explore the different dynamical processes influence the fog lifecycle (see figure 2 for an example). There is also the possibility of using more idealised large-eddy simulations (with a resolution as high as 1m) conducted with MONC (Met Office-NERC cloud model) to study the more detailed turbulence processes affecting fog development.
Figure 2: Predicted visibility from the 100m MetUM simulation over the Clun Valley region for an IOP during LANFEX. The black circles show observed visibility at the main sites for the same time.
Key questions for the PhD will include
- How well do the 100m MetUM simulations capture the spatial patterns of valley flow and turbulence leading up to and during fog episodes and to what extent does advection play a role in fog formation?
- Does the model reproduce the observed sensitivity of fog formation to turbulence and through what mechanism(s) does enhanced turbulence suppress fog in the model?
- How does valley geometry and the presence of side valleys alter the turbulence and mixing of temperature and humidity, and can this explain why some valleys are frequently foggy while adjacent valleys remain fog-free?
- What impact do small scale waves and sloshing motions have on the variability of fog, and can they affect the fog development?
- What role do dynamical processes play in the dispersion or persistence of fog after sunrise?
This project will develop a better physical understanding of the role of dynamical processes in fog development in real, complex terrain and how well this is represented in models. Detailed heat and moisture budgets from the model will highlight which mechanisms control the development of fog. This mechanistic understanding, along with sensitivity tests using the model will provide insight into the predictability of fog, and identify which factors limit this predictability. The work will directly feed into improving future very high resolution model simulations of fog (e.g. Boutle et al, 2016, – 300m simulations to improve fog forecasting for Heathrow). The process understanding will also highlight the key processes which need to be represented in downscaling schemes, such as that being developed at the Met Office for operation forecasting (e.g. Sheridan, Vosper and Smith, 2017), or in parametrisations schemes for coarser resolution operational weather and climate models.
Potential for high impact outcome
High profile news stories, such as the cancellation or delay of more than 10,000 passengers at Heathrow during December 2016 and January 2017 show the economic and personal impact of fog. Improved forecasting can mitigate delays by allowing airlines to plan better to ensure aircraft and crews are in the right place to reduce delays and to provide better information to passengers. Road traffic accidents due to fog also pose a threat to human health as well as causing major disruption to road users (e.g. the Sheppey Bridge crash in September 2013 involving 130 vehicles and over 60 injuries). Improving the understanding and forecasting of fog is a challenge for many parts of the world with a number of groups working in this area, and so there is an expectation that this work can produce high impact journal outputs. The links with the Met Office will ensure that the findings are pulled through to operational improvements, for example feeding in to the current development of improved downscaling techniques.
The student will work under Dr Andrew Ross in the Dynamics and Clouds research group within ICAS. This project provides a high level of specialist scientific training in: (i) State-of-the-science application and analysis of high resolution numerical weather prediction (NWP) models; (ii) analysis of in-situ and remote sensing meteorological measurements from a range of observational platforms; (iii) numerical modelling and use of cutting-edge supercomputers. Co-supervision will involve regular meetings between all partners and extended visits for the student to work alongside researchers in the UK Met Office, under the joint supervision of Dr Jeremy Price (observations) and Dr Ian Boutle (modelling). The successful PhD student will have access to a broad spectrum of training workshops put on by the Faculty that include an extensive range of training workshops in numerical modelling, through to managing your degree, to preparing for your viva (http://www.emeskillstraining.leeds.ac.uk/).
The student should have a keen interest in the challenges of understanding and modelling the weather and a strong background in a relevant quantitative science (meteorology, maths, physics, engineering, environmental sciences). Experience of scientific programming / data analysis is desirable, but not essential.
The proposal has been agreed as a “Partnership Project” with the UK Met Office and the student will benefit from Met Office supervisors as well as academic supervisors at Leeds. Leeds has a strong record of close collaboration with the Met Office and is one of only 4 universities in the Met Office Academic Partnership. The project aligns with existing collaborations between Leeds and the Met Office on boundary layer meteorology in complex terrain and fog. The successful student will spend time working with the industry supervisors at the Met Office.
Boutle, I.A., Finnenkoetter A., Lock A.P., Wells H. (2016) The London Model: forecasting fog at 333m resolution. Quart. J. Royal Meteorol. Soc. 142, 360-371.
Price, J. et al (2011) COLPEX: Field and Numerical Studies over a Region of Small Hills, Bull. Am. Meteorol. Soc. 92, 1636-1650.
Sheridan P.,F. Vosper S.B., Smith S.A. (2017) A physical based downscaling for temperature in complex terrain. J. Appl. Meteol Climatol. Under review.
Related undergraduate subjects:
- Applied mathematics
- Atmospheric science
- Environmental science
- Natural sciences
- Physical science