Using ADCP Background Sound Levels to Estimate Wind Speed

Abstract

It is well known that ambient sound is generated by wind through the process of wave breaking and bubble injection. The resulting sound levels are highly correlated with wind speed and, even though the physical process is not fully understood, sound levels can be used to estimate wind speeds with accuracies comparable to other marine wind measurement techniques. It has been noted by several researchers that background sound levels in acoustic Doppler current profiler (ADCP) systems are correlated to wind speeds; however, conventional wisdom would suggest that this signal should be dominated by thermal noise. In this report, background sound levels in 75-, 150-, and 300-kHz ADCP systems have been investigated. Techniques required to convert raw data into absolute sound levels and to adjust these values to estimate representative surface sound levels are presented. Only the background sound levels in the 150-kHz ADCP retain a signal from the surface-generated ambient sound. For these systems, deployment-independent wind speed estimates can be made with an accuracy of 1.5 ± 1.5 m s?1 for wind speeds up to 15 m s?1; accuracies of ?0.1 ± 1.4 m s?1 can be achieved when using deployment-specific calibration constants. At higher wind speeds, significant signal attenuation occurs due to the presence of subsurface bubbles; a correction for this attenuation is applied. Wind speeds determined from background sound levels can be combined with the ability of upward-looking ADCPs to extract wind direction to provide wind vector data. There is also a wind speed?dependent signal in the near-surface backscatter levels of the 300-kHz systems. The 300-kHz signal is not associated with background sound levels and is most likely caused by near-surface backscatter.