Estimating wind speed at an urban reference height

Different studies suggest reference heights z_ref for wind speed measurements in urban areas (e.g. z_ref = z_d + 10 m or z_ref = 1.25h, with z_d the zeroplane displacement and h the mean building height). However, available urban observations are usually not from that height. Therefore procedures are needed to estimate the wind speed at reference height from measurements taken at any other height. A recently suggested procedure of the European COST 715 Action (Rotach et al., 2004) uses three simple steps, developed with the knowledge of the urban roughness sublayer from a number of field and wind tunnel experiments:

(a) Estimation of the urban zeroplane displacement z_d and the height of the urban roughness sublayer z_* (interpreted as the height of maximum local Reynolds stress). Both values can be empirically calculated with morphometric data as input.

(b) Results from various experiments showed, that local Reynolds stress varies with height. Therefore, a parameterisation for the vertical profile of Reynolds stress is used to calculate a global friction velocity, u_*(z_*) (Rotach, 2001). Input is the local Reynolds stress measured any height above h, or a measurement from outside the city.

(c) Finally, the wind profile equation is numerically integrated with the corresponding loacal function u*(z) from step 2.

This procedure has been thoroughly tested with data from the Basel Urban Boundary Layer Experiment (BUBBLE). The year-long near-surface measurements at one of the towers allow the validation of the procedure with various input configurations, different measurement heights, as well as under different flow conditions. The estimated reference wind speeds are compared to measured wind speeds at reference height.

It is shown that in a height of approx. 1.5h highest values of local Reynolds stresses are found in the data, which is interpreted as z_*. The height of maximal local Reynolds stress is surprisingly constant for different flow situations.

Overall, the results of the procedure are encouraging, and most configurations result in reasonable estimates of the wind speed at the reference height, especially if any local measurement of Reynolds stress at the urban site is available. Larger sensitivity and errors are associated with flow directions that have larger inhomogeneities and a highly variable building height. It is not surprising, that the estimation of u_*(z_*) from a measurement outside the city results in poorest performance.

References

Rotach MW (2001): 'Simulation of urban-scale dispersion using a Lagrangian stochastic dispersion model', Boundary-Layer Meteorol., 99, 379-410.

Rotach MW, Batchvarova E, Berkowicz R, Brechler J, Christen A, Georgieva E, Janour Z, Krajny E, Middleton D, Osrodka L, Prior V, Soriano C (2004): 'Modification of flow and turbulence structure over urban areas', in: Fisher B., Rotach MW, Piringer M, Kukkonen J, Schatzmann M. (Eds): 'Meteorology applied to urban air pollution problems'. Final report of COST 715, in preparation.