EXPERIMENTS WITH STATISTICAL OBJECTIVE ANALYSIS

A series of experiments of height and wind field statistical objective analysis has been made, coming to the conclusions that the analysis scheme is effective; the statistical interpolation is better than the correction method; some parameters, for instance, the number of observations included in the grid estimate, the influence radius and the first guess field, are not sensitive to RMS differences between the analyzed and observed values; and the Gaussian correlation model is more effective than the damped cosine model.