Wind Speed Estimation Using Acoustic Underwater Glider in a Near-Shore Marine Environment

Abstract

This paper investigates the use of an acoustic glider to perform acoustical meteorology. This discipline consists of analyzing ocean ambient noise to infer above-surface meteorological conditions. The paper focuses on wind speed estimation, in a near-shore marine environment. In such a shallow water context, the ambient noise field is complex, with site-dependent factors and a variety of nonweather concurrent acoustic sources. A conversion relationship between sound pressure level and wind speed is proposed, taking the form of an outlier-robust nonlinear regression model learned with in situ data. This method is successfully applied to experimental data collected in Massachusetts Bay (MA, USA) during four glider surveys. An average error in wind speed estimation of 1.3 m • s −1 (i.e., average relative error of 14%) over wind speed values up to 17 m • s −1 is reported with this method, which outperformed results obtained with relationships from the literature. Quantitative results are also detailed on the dependence of wind speed error estimation on the environment characteristics, and on the classification performance of observations contaminated by acoustic sources other than wind. Passive acoustic-based weather systems are a promising solution to provide long-term in situ weather data with fine time and spatial resolutions. These data are crucial for satellite calibration and assimilation in meteorological models. From a broader perspective, this paper is the first step toward an operationalization of acoustic weather systems and their onboard embedding in underwater monitoring platforms such as gliders.

Multimodal 4DVarNets for the reconstruction of sea surface dynamics from SST-SSH synergies

“…The ever-increasing availability of observation data and numerical simulations also greatly contribute to the development and evaluation of learning-based and data-driven approaches as illustrated by the considered experimental setting based on an open data challenge 5 . We could apply and extend the proposed framework to other space-time geophysical products such as ocean colour [51,44], sea surface turbidity [50,42], sea and land surface temperature [5], sea surface currents [11,41]... Future challenges also involve joint calibration and interpolation issues for future satellite missions [22] as well as multimodal synergies between satellite data and other remote sensing and in situ data sources such as drifters [45,4], underwater acoustics data [8], moored buoys [23], argo profilers [13,17]... Especially, the latter might provide new ways to better monitor the interior of the ocean which cannot be directly observed from space.…”

TOSSIT: A low-cost, hand deployable, rope-less and acoustically silent mooring for underwater passive acoustic monitoring

“…Passive Acoustic Monitoring (PAM) in the ocean has become a standard technique across the oceanographic community. On top of historical military applications, PAM is now widely used for biological, geological and meteorological questions such marine mammal occurrence and population density estimation [1] , ocean ambient noise characterization [2] , [3] , soundscape measurements on coral reef to assess biological activities [4] , marine biodiversity assessment [5] , seismic monitoring [6] , acoustical meteorology [7] , estimation of water column or seafloor geoacoustic properties [8] , [9] .…”

“…The TOSSIT concept is of interest for all the marine applications that have historically relied on bottom-moored single-hydrophone measurements, notably bioacoustics [1] , [2] , [5] , [14] or soundscape studies [2] , [3] , [4] , [7] , [15] . Classically, most other ocean acoustics experiments involve the use of synchronized arrays of sensors to perform spatial and temporal filtering.…”

Introducing Temporal Correlation in Rainfall and Wind Prediction From Underwater Noise

“…For many years, this prediction was carried out using empirical equations linking the desired quantity, be it wind or rain, to the noise intensity measured at a precise frequency [3], [4], [5], [6], [7], [8], [9], [10], [11], [12], [13] or, in rare cases, at a few frequencies [14]. Due to their simplicity, these equations have proven to be effective tools in predicting wind and rain with good accuracy, especially in the most frequent ranges.…”