Climate and Atmospheric Science (ICAS) PhD Projects
Modelling of turbulent atmospheric air flow within and above steep forested hills.
Supervisors: Dr Alan Gadian, Dr Sarah-Jane Lock and Dr Andrew Ross
Interactions between forest canopies and the atmosphere are immensely important, since forests are a major sink of CO2 from the atmosphere through photosynthesis, and are also an important source of volatile organic compounds (VOCs). Attempts to quantify these sources and sinks through in-situ flux measurements have proved difficult due to the effects of advection, even over quite moderate slopes - horizontal transport due to variations in the terrain cannot be ignored. Night-time drainage flows are particularly problematic. In recent years this has catalysed interest in understanding and modelling the dynamics of canopy-atmosphere interactions over hills and their impact on chemical transport. These problems are also important for a range of other applications. The presence of trees can modify the pressure drag over a hill. Correctly parameterising this is important for numerical weather prediction (NWP). Variations in wind strength, shear and turbulence caused by hills are a major cause of wind-throw damage to trees and are also of great importance to the wind energy industry. The impact of stable night-time conditions on canopy flows is not well modelled at present. In addition to causing night-time drainage flows, stable conditions can modify the winds and turbulence within forest canopies. Extremes of temperature can lead to frost damage to young plants or lead to drought and wilting.
Modelling all these phenomena requires an understanding of the turbulence within the forest canopy. This turbulence is significantly different from turbulence in the atmospheric boundary layer above and typical schemes are known not to perform particularly well within forest canopies. While the mean-flow may be reasonable, the details of the turbulent mixing are not. Despite this, current canopy modelling approaches have tended to use this approach for practical computational reasons. Large-eddy simulation (LES) offers an attractive alternative modelling approach where the larger scale turbulent eddies are explicitly resolved and only the small-scale turbulence is parameterised using a sub-grid scale (SGS) turbulence parameterisation scheme. This avoids issues with the use of mixing length closure schemes in canopies, and is also suitable for steep complex terrain, unlike mixing length schemes. These benefits come at high computational cost: the domain size must be large enough to contain the hill of interest, while the grid resolution must be small enough to accurately resolve the canopy and at least the larger scale turbulent eddies within it. With modern high performance computing facilities (such as those at Leeds) and new model developments this is now possible.
There are currently only a few studies in the literature on the use of LES for canopy flows over hills (including work at Leeds). This is due both to computational limitations and the lack of suitable models. This project aims to address these issues by extending a new model developed by the group at Leeds – the National Centre for Atmospheric Science (NCAS) Microscale Model. This model uses novel computational approaches to allow flow over very steep hills and to ensure the model scales well on high-performance parallel computers. The addition of a canopy model and a suitable LES SGS model will allow some exciting research questions to be addressed, both in terms of the development of novel numerical schemes and in answering some of the important science questions surrounding the atmosphere–canopy flows.
Fig 1: Large-eddy simulations of a) horizontal and b) vertical velocities within and above a forest canopy over a small hill.
There are a number of wind tunnel experiments conducted at ENFLO, University of Surrey and by collaborators at CSIRO in Canberra, Australia which will be used to validate the model developments. These experiments include both neutral and stable conditions. In addition there are also a number of datasets from field campaigns in the UK and elsewhere which may be used.
Fig 2: Wind tunnel experiments of flow over a hill at the ENFLO wind tunnel, University of Surrey.
Numerical questions will focus on the choice of SGS for use in a canopy. Much work has been done on LES studies of the atmospheric boundary-layer to develop more accurate SGS schemes, however little of this has been translated to canopy LES flows. A careful study of the merits of different schemes will be important as a foundation for future work in this area, particularly for stable conditions where there is currently no published work. Using results from LES simulations will also allow a more thorough assessment of the strengths and weaknesses of mixing length closure approaches to dealing with turbulence in canopies. From a practical and NWP point of view this is important as computationally we are still a long way from being able to run very high resolution LES models for long times or over large areas. LES simulations will be used to accurately determine the relative importance of advection and turbulent mixing in chemical transport within and above forest canopies. In particular, the effects of stability on the flow dynamics and on transport will be investigated. Finally, the ability of the model to cope with steep terrain will allow new research into flow over steep terrain which is either partially or fully forested.
This project offers an exciting opportunity to work within a highly experienced interdisciplinary group with extensive experience of atmospheric flow over hills and mountains, interactions with forest canopies, as well as expertise in numerical model development. The group is part of the internationally renowned Institute for Climate and Atmospheric Science within the School of Earth and Environment. It is planned that the student would have the opportunity to travel and work with collaborators in the US and Australia during their studentship.