Revealing underwater plastics: Detection of suspended macroplastics using acoustic backscatter

Abstract

Listen

Translate article

<p>Plastic pollution has become a threat to both nature and humans. A substantial amount of plastic is transported in rivers towards the sea. Insight in the sources and transport in rivers, and how this is changing with conditions like discharge, is needed to plan effective mitigation strategies. Current measurement techniques mostly target floating items, recent studies however suggest that suspended plastics form a significant part of the total riverine plastic transport. Despite being successful, recently applied sampling techniques using several nets in the vertical are too invasive, expensive and labour intensive to apply on large spatial and temporal scale. A non-invasive and continuous measurement technique of suspended plastics, applicable on a large scale, is needed.</p><p> </p><p>Sonar has shown potential for plastic detection. During measurements, a high-frequency acoustic signal is transmitted. This signal is scattered by water and suspended particles. Using an Acoustic Doppler Current Profiler (ADCP), the chosen sonar device, flow velocity can be obtained using the frequency shift of the signal. Aside from the frequency shift, also the strength of the returned acoustic signal can be measured. This backscatter strength depends on the size, form and material of the sampled object. It is nowadays used to estimate suspended sediment concentrations, but also relatively large objects like fish, organic matter and macro-plastics can be recognised by high backscatter intensity. As ADCPs have been broadly used for decades, current and historical data from a large network of measuring devices are available for analysis of plastic transport and its fluctuations. We propose steps for a method to quantify the macroplastic concentration from ADCP data.</p><p> </p><p>During the study, a RDI StreamPro ADCP is horizontally mounted in a basin. Macroplastics differing in shape, size and polymer type are sampled 5 times within a period of 10 seconds, on three different distances (around 1, 3 and 5 meter) from the transducer. Almost all plastic items show a significantly higher backscatter intensity than the background signal, on all measurement distances. In contrast to net-based measurements, rotation is found to be an important aspect in the identification of items during ADCP measurements. To further develop the detection method, analysis of ADCP field data containing the usual background noise, combined with net measurements for validation, is needed. Expectation is that net measurements, used to make an estimate of the fraction organic material and plastic, are only needed periodically to make a good continuous estimate of macroplastic transport using ADCP data.</p>