Modelling floating debris’ beaching and drift nearshore -a case in Corsica & Sardinia, and the parametrization in the Litter TEP

The Litter -TEP (Thematic Exploitation Platform), which was developed by ARGANS Ltd, with a grant of CMEMS, aimed at forecasting litter introduction by rivers and marine drift on the European North-Western Shelf so as to help local coastal communities i. schedule beach cleansing and ii.  assess the potential origin of materials collected. It needed a litter beaching model, in addition to a drift model, for that. ARGANS benefited from a grant of IFREMER through the European interregional project MARITTIMO-SICOMAR plus, to study litter beaching processes on the Corsican shoreline, owing to the extensive survey performed in 2016-2017 by IFREMER and the localization of hot spots, i.e. locations with more than 10 litter pieces on a distance of 2-to-30m alongshore. After a gross analysis of data by CMEMS for winds, currents and waves, 3 areas were selected among the 6 main litter accumulation areas, i.e. La Maddalena, Capo di Feno, the Ajaccio Gulf, the Gulf of Propriano, Bastia shores and the Agriate Desert, to try to understand the reason for the location of the litter hot spots, but focusing exclusively on i. transport by waves and ii.a swash on the shore or ii.b picked up by longshore currents along the beach then swashed (ii.a) —without knowing the litter sources, as if the sources were disposed uniformly offshore linearly along the coast.

To get the transport component, the incoming waves were simulated with the spectral model SWAN, at a 25 m resolution, using inputs from WAVEWATCH III; to get the beaching per se, i.e. the surf zone dynamics that would deposit litter on the shore, we used a SWASH model that was nested in the former at a spatial resolution of 1 to 10 m. SWASH was originally discarded in favor of the XBeach model, a short-wave averaged and wave-group resolving model that we use for civil engineering calculation, because a computing-efficient model and its ore approximations fit the purpose (motions at the shore break are dominated by long wave). Yet, despite the possibility to action the ‘surf-beat’ mode of XBeach, allowing resolving the short wave variations on the wave group scale and getting the wave-driven currents (longshore current, rip currents), long(infragravity) waves, and runup and rundown of long waves (swash), we switched back to SWASH, as it does not consider a depth-averaged flow and seemed to resolve better the incident-band (short wave) runup on intermediate dissipation shores.

In the three AOIs, 67 hotspots were identified during the ground survey, and 90 hotspots were forecasted. Out of the 67s, 59 were forecasted: 42 at the right location and 17 with slight error which is probably due to the lack of proper VHR bathy-topography and sedimentological maps to perform the simulations. 8 surveyed hotspots were not foreseen, and 31 forecasted hotspots were not identified on ground. As such, the probability of detection was 88% and the probability of false alarms 32%. Better rates are expected using the new LITTO3D lidar surveys of Corsican nearshores, and a priori knowledge of litter sources.