{"references": ["Finnigan, J.J., \"Turbulence in plant canopies,\" Annu Rev Fluid\nMech, 2000. 32: p. 519-571.", "Lancaster, N. and A. Bass, \"Influence of vegetation cover on sand\ntransport by wind: Field studies at Owens Lake, California,\" Earth\nsurf. process. landforms, 1998. 23: p. 69-82.", "Putten, W.H.V.D., \"Establishment of Ammophilla Arenaria ( Marram\nGrass) From Culms, Seeds and Rhizomes,\" The Journal of Applied\nEcology, 1990. 27: p. 188-199.", "Mohan, M. and M.K. Tiwari, \"Study of momentum transfers within a\nvegetation canopy,\" Proc. Indian Acad. Sci. (Earth Planet. Sci.),\n2004. 113: p. 67-72.", "Poggi, D., G.G. Katul, and J.D. Albertson, \"Momentum transfer and\nturbulent kinetic energy budgets within a dense model canopy,\"\nBoundary-Layer Meteorology, 2004. 111: p. 589-614.", "Kim, H.G., V.C. Patel, and C.M. Lee, \"Numerical simulation of wind\nflow over hilly terrain,\" Journal of wind engineering and industrial\naerodynamics, 2000. 87: p. 45-60.", "Lun, Y.F., et al., \"Applicability of linear type revised k \u256c\u00c1 models to\nflow over topographic features,\" Journal of wind engineering and\nindustrial aerodynamics, 2007. 95: p. 371-384.", "Maurizi, A., J.M.L.M. Palma, and F.A. Castro, \"Numerical\nsimulation of the atmospheric flow in a mountainous region of the\nNorth of Partugal,\" Journal of wind engineering and industrial\naerodynamics, 1998. 74-76: p. 219-228.", "Ayotte, K.W., J.J. Finnigan, and M.R. Raupach, \"A second-order\nclosure for neutrally stratified vegetative canopy flows,\" Boundary-\nLayer Meteorol, 1999. 90: p. 189-216.\n[10] Green, S.R., \"Modelling turbulent air flow in a stand of widelyspaced\ntrees,\" Phoenics J., 1992. 5(294-312).\n[11] Katul, G.G., et al., \"One- and Two-equation models for canopy\nturbulence,\" Boundary-Layer Meteorol, 2004. 113(80-109).\n[12] Neary, V.S., \"Numerical solution of fully developed flow,\" Journal of\nengineering mechanics, 2003. 129: p. 558-563.\n[13] Sogachev, A. and O. Panferov, \"Modification of two-equation\nmodels to account for plant drag,\" Boundary-Layer Meteorol, 2006.\n121: p. 229-266.\n[14] Sherman, D.J., et al., \"Wind-blown sand on beaches: an evaluation of\nmodels,\" Geomorphology, 1998. 22: p. 113-133.\n[15] Stephen, M. and Z. Xin, \"Computational analysis of pressure and\nwake characteristics of an aerofoil in ground effect,\" Journal of fluid\nengineering, 2005. 127: p. 290-298.\n[16] Launder, B.E. and D.B. Spalding, \"The numerical computational of\nturbulent flows,\" Computer methods in applied mechanics and\nengineering, 1974. 3: p. 269-289.\n[17] Yakhot, V. and S.A. Orszag, \"Renormalization group analysis of\nturbulence: I. Basic theory,\" Journal of Scientific Computing, 1986.\n1(1): p. 1-51.\n[18] Shih, T.-H., et al., \"A New k-\u256c\u00c1 eddy-viscosity model for high\nReynolds number turbulent flows model development and\nvalidation,\" Computers fluids, 1995. 24(3): p. 227-238.\n[19] Menter, F.R., \"Two-equation eddy-viscosity turbulence models for\nengineering applications,\" AIAA Journal, 1994. 32(8): p. 1598-1605.\n[20] Wakes, S.J., et al., \"Using Computational Fluid Dynamics to\ninvestigate the effect of a Marram covered foredune; initial results,\"\nin Seventh International Conference on Modelling, Measurements,\nEngineering and Management of Seas and Coastal Regions. 2005:\nAlgarve, Portugal.\n[21] Li Liang, et al., \"Improved k-\u256c\u00c1 two-equation turbulence model for\ncanopy flow,\" Atmospheric Environment 2006. 40: p. 762-770.\n[22] Aynsley, R.M., W. Melbourne, and B.J. Vickery, \"Architectural\nAerodynamics,\" 1977: Applied Science.\n[23] Raupach, M.R. and R.H. Shaw, \"Averaging procedures for flow\nwithin vegetation canopies,\" Boundary-Layer Meteorol, 1982. 22: p.\n79-90.\n[24] Sanz, C., \"A note on k \u2212 \u256c\u00c1 modelling of vegetation canopy airflows,\"\nBoundary-Layer Meteorology, 2003. 108: p. 191-197.\n[25] Foudhil, H., Y. Brunet, and J.-P. Caltagirone, \"A fine-scale k-\u256c\u00c1 model\nfor atmospheric flow over heterogeneous landscapes,\" Environ Fluid\nMech, 2005. 5: p. 247-265.\n[26] Liu, J., et al., \"E-\u256c\u00c1 modelling of turbulent air flow downwind of a\nmodel forest edge,\" Boundary-Layer Meteorol 1996. 77: p. 21-44\n[27] Blocken, B., T. Stathopoulos, and J. Carmeliet, \"CFD simulation of\nthe atmospheric boundary layer: wall function problems,\"\nAtmospheric environment 2007. 41: p. 238-252.\n[28] Parsons, D.R., et al., \"Numerical of airflow over an idealised\ntransverse dune,\" Environmental modeling and software, 2004. 19: p.\n153-162.\n[29] Neff, D.V. and R.N. Meroney, \"Wind-tunnel modeling of hill and\nvegetation influence on wind power availability,\" Journal of wind\nengineering and industrial aerodynamics, 1998. 74-76: p. 335-343.\n[30] Cao, S. and T. Tamura, \"Experimental study on roughness effects on\nturbulent boundary layer flow over a two-dimensional steep hill,\"\nJournal of wind engineering and industrial aerodynamics, 2006. 94:\np. 1-19."]}

Turbulence modeling of large-scale flow over a vegetated surface is complex. Such problems involve large scale computational domains, while the characteristics of flow near the surface are also involved. In modeling large scale flow, surface roughness including vegetation is generally taken into account by mean of roughness parameters in the modified law of the wall. However, the turbulence structure within the canopy region cannot be captured with this method, another method which applies source/sink terms to model plant drag can be used. These models have been developed and tested intensively but with a simple surface geometry. This paper aims to compare the use of roughness parameter, and additional source/sink terms in modeling the effect of plant drag on wind flow over a complex vegetated surface. The RNG k-ε turbulence model with the non-equilibrium wall function was tested with both cases. In addition, the k-ω turbulence model, which is claimed to be computationally stable, was also investigated with the source/sink terms. All numerical results were compared to the experimental results obtained at the study site Mason Bay, Stewart Island, New Zealand. In the near-surface region, it is found that the results obtained by using the source/sink term are more accurate than those using roughness parameters. The k-ω turbulence model with source/sink term is more appropriate as it is more accurate and more computationally stable than the RNG k-ε turbulence model. At higher region, there is no significant difference amongst the results obtained from all simulations.