Experimental issues in atmospheric boundary layer simulations: roughness length and integral length scale determination

Wall shear stress was measured directly in three atmospheric boundary layer (ABL) simulations using a specially constructed force balance. The roughness length of the simulations were determined using: (a) balance measured shear velocities; (b) shear velocities computed from X-wire Reynolds stress measurements; and (c) indirect procedures where accepted forms of turbulent boundary layer mean velocity profiles are fitted to the measured mean-velocity profiles. Comparison of balance measured shear velocities with those computed from the Reynolds stress measurements showed that the latter may be off by more than 15%, leading to large errors in roughness length values. The significant uncertainties associated with roughness length estimates obtained using indirect velocity-profile-fitting techniques in ABL simulations are documented. It is shown that the roughness length estimates computed from indirect procedures can be in error by factors of two or more. Experimental uncertainties associated with the determination of integral length scale are also presented.