A Look Back on Two Decades of Doppler Sodar Comparison Studies

Abstract

This paper is a compilation of the results obtained from various Doppler sodar comparison experiments conducted over the last 20 years. These studies have attempted to quantify the uncertainties in sodar-derived values of the horizontal wind speed, wind direction, standard deviation of the vertical wind speed sw, and standard deviation of the horizontal wind direction sq. Doppler sodar configurations examined in these studies included bistatic, monostatic, and phased array. In most cases, reference measurements used for comparison were made by tower-based in situ sensors. Many investigators have used simple linear regressions and other statistical measures such as the correlation coefficient, bias (mean difference), comparability (root-mean-square difference), and precision (standard deviation) in an attempt to quantify those errors. The sodar-derived wind speed and wind direction are highly correlated against reference measurements (≈0.92) with a precision of 1.06 m s-1 and 21.5°, respectively, for the entire dataset. Correlations of sodar-derived values of sw were not quite as good (≈0.81) with an average precision of 0.18 m s-1. Past studies have shown that sw accuracies vary significantly from day (convective conditions) to night (stable conditions). Very few data values were available for sq, which had a poor correlation of 0.57 and a precision of 10.7°. The conclusions from many of these studies have shown that Doppler sodars can accurately obtain the mean wind speed and wind direction. Values of sw have larger uncertainties, while estimates of sq have errors that are consid-ered unacceptable for any practical use. Much of the observed scatter in sodar wind measurements can be attributed to a number of factors. These include, but are not limited to, instrument configuration, spatiotemporal variability, noise, and processing techniques.