Near-surface wind estimates using statistics from a planetary boundary-layer model

This paper shows the possibilities of a procedure for estimating near-surface wind statistics, by means of the numerical integration of a simple boundary-layer model with a second-order turbulent closure. Standard and easily available synoptic data are used as initial and boundary conditions. The development of this methodology is impelled by increasing requirements of a quick and precise knowledge of the wind characteristics in many regions of South America, which confronts the serious limitation of a reduced number of extended observational series, scattered over a vast continent. In order to evaluate the methodology, near-surface wind statistics from observed data at two locations are compared with model output statistics. Relative errors are about 0.2 for daily mean velocities and about 0.1 for weekly mean velocities, observed and computed time series being highly correlated in both cases. Calculated frequency distribution of wind directions is in good agreement with the observed one, and the absolute mean error in the daily mean wind direction is about 20 deg. Even though a wide variety of large-scale synoptic situations has been indirectly considered through boundary conditions, basic model output statistics resemble fairly well those observed at different levels between the surface and 100 m.