Estimation of aerodynamic roughness length and displacement height of an urban surface from single-level sonic anemometer data

Realistic specifications of the roughness length and zero-plane displacement are crucial in the boundarylayer parametrisation that plays an important role in numerical mesoscale models (Abdella and Mcfarlane 1996). Martano (2000) proposed a method for estimating the surface roughness length (z0) and the displacement height (d) from single-level sonic anemometer data. This method reduces the problem of finding joint values of both z0 and d to a simpler least-squares procedure for one variable. In this method, the standard deviation σS of the function S(z0, d), defined below, is minimised with respect to d, resulting in a direct estimate of d and z0, together with their statistical uncertainty. The function S(z0, d) = {kU/u* + ψ[(z d)/L] ψ(z0/L)} is a statistical quantity that depends on the data. Here, k is the von Karman constant (0.35), U is the horizontal wind speed at height z, u* = (-)1/2 is the friction velocity, L = u*T/(kgθ*) is the Monin-Obukhov scale height, and θ* = -/u*. In the above u, w and θ are fluctuations in horizontal wind speed, vertical wind speed, and potential temperature respectively, T is the absolute temperature and g is the gravitational acceleration. The function ψ[(z-d)/L,z0/L/] = ψ[(zAust. Met. Mag. 53 (2004) 21-28