2000b) Listening for rain: Theory and practice

Wind and rain are known to be significant sources of underwater noise. Previous researchers have developed algorithms to analyse the spectral content of underwater sound, classify according to source, and provide quantitative values. In this paper we describe the evaluation of such algorithms using acoustic data gathered in Loch Etive on the west coast of Scotland. The validation datasets are provided by a buoy-mounted anemometer and a tipping-bucket gauge. In summary, the detection of rain is found to be reliable, but the quantitative comparisons show significant scatter. 1