Modelling air-sea interaction in the coastal atmospheric boundary layer
Supervisors: Ian Brooks, Andrew Ross (Leeds) and Judith Wolf (National Oceanography Centre, Liverpool)
The wind drives much of the movement of the sea, forcing the deep ocean circulation as well as coastal circulation. The latter may lead to storm surges and flooding resulting in loss of life and economic losses for the UK. One of the most fascinating, and not yet fully solved, problems in oceanography is the interaction between the atmosphere and the ocean. The interface between two turbulent boundary layers is difficult to observe and complex to model. The details of the atmospheric and oceanic boundary layers cannot be solved explicitly in operational models and are thus usually treated in a semi-empirical way, generally using separate models for the atmosphere and ocean, without accounting for the exchange of momentum or other fluxes in a conservative way. Many processes are affected by the air-sea exchange of momentum, heat, water vapour, gases (e.g. CO2) and other matter (e.g. sea spray aerosol). Effects of buoyancy on the stability of the air and sea boundary layers are also important in certain conditions.
The air-sea momentum transfer is complex and has been subject to much controversy over many years. Wind waves play a central role in air-sea interface processes (see e.g. Csanady, 2001). The history of the development of theory and observations of the wind input and wave growth is discussed in Cavaleri et al. (2007) and Janssen (2004). The full theoretical development of the air-sea momentum transfer is an extremely difficult problem because it involves a turbulent airflow over a surface that varies in space and time (Janssen, 2007). The effective shear stress may be written at the interface as τa=ρau*a2 in the atmosphere and τw=ρwu*w2 in the water, where the subscripts a and w refer to variables in the atmosphere and water respectively, ρ is density and u* the friction velocity. The appropriate friction velocity is often parameterized by using a representative drag coefficient, but many assumptions may be made in deriving this parameter. When forcing an ocean model with atmospheric wind fields, the wind is commonly specified at a standard height of 10m above the sea surface, and the drag coefficient is related to the wind speed, but not the details of the wave field. This simplistic representation assumes fully developed waves and does not capture the variability of the wind-stress in rapidly changing winds, such as in storms. The drag coefficient is related to a roughness length - in the case of the sea surface it is clear that this roughness length must be related to the wave height (Janssen, 1989). The operation of a coupled atmosphere-ocean wave model at ECMWF since 1998 has demonstrated improvement in both wind and wave forecasting and hindcasting (Janssen, 2004), but this has not been implemented in other operational systems. Ultimately the goal is to have a fully consistent momentum and energy balance between the atmosphere and ocean, accounting for effects such as Stokes’ drift and Langmuir circulation which are often not explicitly calculated.
Over the open ocean, the atmospheric surface layer can often be expected to be in a state of equilibrium with the sea surface. Conversely in coastal regions this is often not the case and the underlying assumptions of the air-sea flux parameterisations are violated, biasing the modelled fluxes of momentum, heat, and moisture (e.g. Brooks, 2001; Brooks et al., 2003). Numerical models of storm surges (Lowe et al., 2001) and waves (Brown et al., 2010) consistently underestimate extreme values compared with observations when forced by long meteorological reanalyses (e.g. Uppala et al., 2005), especially in semi-enclosed sea areas. Figure 1 shows wind vectors in flow around a coastal headland along with contours of the wind stress as measured by an instrumented aircraft flying 30m above the surface, and estimated by a bulk flux algorithm from the mean conditions. There are very significant differences between the two.
- Figure 1. Arrows show the wind vector for near-surface flow around a coastal headland. Black lines are contours of the measured wind stress (N m-2) and grey lines the estimated wind stress from the NOAA-COARE bulk air-sea flux algorithm.
Ocean forecasting is critically dependent on accurate wind forcing, especially for extreme events where the nonlinear relationship of wind-stress to wind means that small errors in winds can lead to much larger effects on waves and surges. Also, the rapidly changing winds in a storm mean that the waves never reach a state of full development. The surface stress acting on the mean circulation will be the total wind-stress minus the stress driving the waves (Janssen et al., 2004) it is thus important to model wind-wave exchange accurately in order to correctly predict storm surges. In coastal areas there are strong changes between the atmospheric boundary layer over land and sea, with much higher roughness over land. Orographic interactions can also affect the wind field in this region, significantly affecting the nearshore oceanographic forecasts of waves, currents and mixing.
Atmosphere-ocean interaction is complicated by all the effects noted above. Assessing their influence requires use of an atmospheric model which can be easily modified and coupled with the hydrodynamic and wave models. The WRF model is ideal for this application, is widely used at Leeds, and is supported by NCAS. A particularly important aspect of WRF is its two-way coupling between nesting levels; i.e. not only do the large scales drive the small scales but there is feedback from the small scales to the large scales.
Wolf (2009) has described the development of coupled wave and surge models, including the POLCOMS-WAM development at NOC. Recently Brown and Wolf (2009) have implemented the Janssen (1991) wave-age dependence of the surface roughness used in the wind-stress in the POLCOMS model and demonstrated that this variable roughness avoided the need for local tuning of the Charnock coefficient in the wind-stress parameterisation. The logical next step is to include the feedback effect of the waves on the atmospheric boundary layer. The ultimate goal is to properly account for the air-sea momentum transfer such that the momentum lost from the atmospheric boundary layer correctly drives the ocean circulation via the wave field.
This PhD project will investigate the impact of model grid resolution on the surface winds, momentum transfer, wave field, and wind-driven storm surge, and the impact of feedback of the spatially varying wave field on the atmospheric boundary layer and near-surface winds. It will use the Liverpool Bay as a test bed; Liverpool Bay is part of the northern Irish Sea, lying between NW England and N Wales, in which an intensive oceanographic observation and modelling programme is taking place as part of the POL Coastal Observatory (http://coastobs.pol.ac.uk/). Long-term measurements of waves and currents are being made from moored instruments and shore-based radars. Measurements of the marine atmospheric boundary layer are being made at a tower on Hilbre Island, where the X-band radar is based which measures waves and bathymetry over a region of radius several kilometres. Detailed models of tides, wind-driven and density-driven currents, waves and sediments are being developed in this area and the coastal impacts of storm waves and surges have been studied over many years.
Model winds are routinely supplied from the UK Met Office mesoscale model (presently at 12km resolution) but the local effects of the land-sea boundary layer and the mountains of Snowdonia will have important local effects. Various local area wind models are available, including the PSU/NCAR mesoscale model known as MM5 (http://www.mmm.ucar.edu/mm5/) and the NCEP/NCAR WRF model which is now widely used by many research groups, and will be used in this study. These are limited-area, nonhydrostatic, terrain-following sigma-coordinate models designed to simulate or predict mesoscale atmospheric circulation. The project will investigate the capabilities of the mesoscale atmospheric model and implement it for the Liverpool Bay area. Details of the atmospheric boundary layer in the coastal ocean will be the primary focus of the study, investigating the air-sea momentum exchange to waves and currents, with impacts on coastal flooding and erosion in extreme storms. This model will be used to drive the POLCOMS-WAM hydrodynamic and wave model system and results will be compared with the available measurements to investigate the local variability of winds, waves and currents in Liverpool Bay.
This project will link to a new EU project – FIELD_AC – which aims to improve nearshore oceanographic forecasts using state-of-the-art models applied to four test areas; one of which is Liverpool Bay (Figure 2). As part of this project the Barcelona Supercomputing Center will be setting up mesoscale atmospheric models and their experience will be available to facilitate setting up the atmospheric model for Liverpool Bay. Supervisors at the University of Leeds Atmospheric Dynamics group working on boundary layer modeling will also assist in model setup and investigation.
The student will receive training in modelling techniques as well as basic research skills and have the opportunity to become familiar with data analysis of various instruments for measuring waves, currents, turbulence and suspended and bedload transport of sediment. A good degree in the physical sciences as well as mathematical aptitude is required. Prior experience of computer programming (particularly in Fortran and Matlab) is an advantage.
- Figure 2. Liverpool Bay and the Coastal Observatory: blue squares are regular CTD stations, yellow squares are fixed moorings or coastal tide gauges.
References
Brooks, I. M. (2001) Air-sea interaction and the spatial variability of surface evaporation ducts in a coastal environment. GRL, 28, 2009-2012.
Brooks, I., S. Söderberg, and M. Tjernström, 2003 The turbulence structure of the stable atmospheric boundary layer around a coastal headland: Aircraft observations and modelling results. BLM, 107, 531-559.
Brown J.M., and Wolf J. (2009). Coupled wave and surge modelling for the eastern Irish Sea and implications for model wind-stress. Continental Shelf Research, 29: 1329–1342, doi: 10.1016/j.cr.2009.03.004.
Brown, J. M., Souza, A. J., & Wolf, J. (2010). An 11-year validation of wave-surge modelling in the Irish Sea, using a nested POLCOMS-WAM modelling system. Ocean Modelling, doi: 10.1016/j.ocemod.2009.12.006
Cavaleri L., J.-H. Alves, F. Ardhuin, A. Babanin, M. Banner, K. Belibassakis, M, Benoit, M. Donelan, J. Groeneweg, T.H.C. Herbers, P. Hwang, P.A.E.M. Janssen, T. Janssen, I.V. Lavrenov, R. Magne, J. Monbaliu, M. Onorato, V. Polnikov, D. Resio, W.E. Rogers, A. Sheremet, J. McKee Smith, H.L. Tolman, G.van Vledder, J. Wolf, I. Young 2007 Wave modelling – the state of the art. Progress in Oceanography, 75, 603–674.
Csanady, G.T. (2001) Air-Sea Interaction: Laws and Mechanisms. Cambridge University Press. 239pp.
Janssen, P.A.E.M. (1989). Wave-induced stress and the drag of air flow over sea waves. Journal of Physical Oceanography, 19, 745–754.
Janssen, P.A.E.M., 2004: The interaction of ocean waves and wind. Camb. Univ. Press. Cambridge, 300 pp.
Janssen, P.A.E.M. 2007 Progress in ocean wave forecasting. Journal of Computational Physics 227 (2008) 3572–3594
Uppala, S.M., et al. 2005: The ERA-40 re-analysis. Quarterly Journal of the Royal Meteorological Society, 131, 2961-3012.doi:10.1256/qj.04.176
Wolf, J. (2009) Coastal Flooding – Impacts of coupled wave-surge-tide models. Natural Hazards, 9, 2, 241, doi:10.1007/s11069-008-9316-5