Mapping the plastic legacy: Geospatial predictions of a microplastic inventory in a complex estuarine system using machine learning
Zenodo (CERN European Organization for Nuclear Research)2024
Score: 35
?
0–100 AI score estimating relevance to the microplastics field. Papers below 30 are filtered from public browse.
Researchers applied machine learning techniques to develop geospatial predictions of microplastic inventory in a complex estuarine system, overcoming the limitations of coarse ocean basin models by accounting for the intricate geomorphological and hydrodynamic conditions that govern sediment-associated microplastic distribution.
Study Type
Environmental
Sediments store the majority of the microplastic legacy in the aquatic environment. To predict microplastic distribution in sediments, models have been developed focussing on coarse ocean basin scales or suited to relatively simple geomorphological and hydrodynamic systems. However, such models reach their limits in more complex estuarine systems, leaving a significant lack of modelling capability for microplastic pollution in key ecosystems. We address this gap by employing machine learning techniques to predict spatial MP inventories deposited in the Schlei, Northern Germany. Notably, the complex hydrodynamic regime, influenced by narrowings and braided embayments, freshwater tributaries, wind-driven mixing, and brackish inflows, creates an interplay of fluvial and marine sedimentary processes, which poses non-trivial challenges for reliable modelling of sedimentary MP transportation. Our approach, termed NIXVEGS (Nested Iterative X-Validation-to-Ensemble-modelling through GridSearches), leverages machine learning, integrating model selection, rigorous validation, and ensemble techniques tailored for small datasets. We estimated 20 trillion MP particles or 14.5 tonnes (50-5000 µm) residing in upper sediments of the Schlei proper, representing the highest resolution picture of an MP inventory for an entire geomorphological region to date. Our modelling concept is founded on the idea of applying granulometric proxies to account for the hydrodynamic regime bias in observed MP concentrations. Estuarine sediments store some of the highest MP concentrations among all sediment types. As depositional systems at the terrestrial-aquatic and fluvial-marine interface, they are key indicator ecosystems for monitoring the evolution of (micro)plastic pollution. Optimising sampling strategies for sediment grain size coverage and spatio-temporal hydrodynamics, and using models such as NIXVEGS, will be important to understand the contribution of complex estuarine systems to the global microplastic budget. Also see: https://micro2024.sciencesconf.org/559755/document