ORCID Profile
0000-0003-4200-1450
Current Organisation
Universiteit Utrecht
Does something not look right? The information on this page has been harvested from data sources that may not be up to date. We continue to work with information providers to improve coverage and quality. To report an issue, use the Feedback Form.
Publisher: American Geophysical Union (AGU)
Date: 04-2021
DOI: 10.1029/2020JC017098
Abstract: Microplastic debris ending up at the sea surface has become a known major environmental issue. However, how microplastic particles move and when they sink in the ocean remains largely unknown. Here, we model microplastic subject to biofouling (algal growth on a substrate) to estimate sinking timescales and the time to reach the depth where particles stop sinking. We combine NEMO‐MEDUSA 2.0 output, that represents hydrodynamic and biological properties of seawater, with a particle‐tracking framework. Different sizes and densities of particles (for different types of plastic) are simulated, showing that the global distribution of sinking timescales is largely size‐dependent as opposed to density‐dependent. The smallest particles we simulate (0.1 μm) start sinking almost immediately around the globe and their trajectories take the longest time to reach their first sinking depth (relative to larger particles). In oligotrophic subtropical gyres with low algal concentrations, particles between 1 and 0.01 mm do not sink within the simulation time of 90 days. This suggests that in addition to the comparatively well‐known physical processes, biological processes might also contribute to the accumulation of floating plastic (of 1–0.01 mm) in subtropical gyres. Particles of 1 μm in the gyres start sinking largely due to vertical advection, whereas in the equatorial Pacific they are more dependent on biofouling. The qualitative impacts of seasonality on sinking timescales are small, however, localized sooner sinking due to spring algal blooms is seen. This study maps processes that affect the sinking of virtual microplastic globally, which could ultimately impact the ocean plastic budget.
Publisher: Copernicus GmbH
Date: 29-06-2020
DOI: 10.5194/NPG-2020-28
Abstract: Abstract. The detection of finite-time coherent particle sets in Lagrangian trajectory data using data clustering techniques is an active research field at the moment. Yet, the clustering methods mostly employed so far have been based on graph partitioning, which assigns each trajectory to a cluster, i.e. there is no concept of noisy, incoherent trajectories. This is problematic for applications to the ocean, where many small coherent eddies are present in a large fluid domain. In addition, to our knowledge none of the existing methods to detect finite-time coherent sets has an intrinsic notion of coherence hierarchy, i.e. the detection of finite-time coherent sets at different spatial scales. Such coherence hierarchies are present in the ocean, where basin scale coherence coexists with smaller coherent structures such as jets and mesoscale eddies. Here, for the first time in this context, we use the density-based clustering algorithm OPTICS (Ankerst et al., 1999) to detect finite-time coherent particle sets in Lagrangian trajectory data. Different from partition based clustering methods, OPTICS does not require to fix the number of clusters beforehand. Derived clustering results contain a concept of noise, such that not every trajectory needs to be part of a cluster. OPTICS also has a major advantage compared to the previously used DBSCAN method, as it can detect clusters of varying density. Further, clusters can also be detected based on density changes instead of absolute density. Finally, OPTICS based clusters have an intrinsically hierarchical structure, which allows to detect coherent trajectory sets at different spatial scales at once. We apply OPTICS directly to Lagrangian trajectory data in the Bickley jet model flow and successfully detect the expected vortices and the jet. The resulting clustering separates the vortices and the jet from background noise, with an imprint of the hierarchical clustering structure of coherent, small scale vortices in a coherent, large-scale, background flow. We then apply our method to a set of virtual trajectories released in the eastern South Atlantic Ocean in an eddying ocean model and successfully detect Agulhas rings. At larger scale, our method also separates the eastward and westward moving parts of the subtropical gyre. We illustrate the difference between our approach and partition based k-Means clustering using a 2-dimensional embedding of the trajectories derived from classical multidimensional scaling. We also show how OPTICS can be applied to the spectral embedding of a trajectory based network to overcome the problems of k-Means spectral clustering in detecting Agulhas rings.
Publisher: Copernicus GmbH
Date: 24-08-2021
Publisher: Elsevier BV
Date: 06-2023
Publisher: Copernicus GmbH
Date: 15-02-2022
Abstract: Abstract. Having descended through the water column, microplankton in ocean sediments is representative of the ocean surface environment, where it originated. Sedimentary microplankton is therefore used as an archive of past and present surface oceanographic conditions. However, these particles are advected by turbulent ocean currents during their sinking journey. So far, it is unknown to what extent this particle advection shapes the microplankton composition in sediments. Here we use global simulations of sinking particles in a strongly eddying global ocean model, and define ocean bottom provinces based on the particle surface origin locations. We find that these provinces can be detected in global datasets of sedimentary microplankton assemblages, demonstrating the effect provincialism has on the composition of sedimentary remains of surface plankton. These provinces explain the microplankton composition, in addition to, e.g., the ocean surface environment. Connected provinces have implications for the optimal spatial extent of microplankton sediment s le datasets that are used for palaeoceanographic reconstruction, and for the optimal spatial averaging of sediment s les over global datasets.
Publisher: Copernicus GmbH
Date: 03-03-2021
DOI: 10.5194/EGUSPHERE-EGU21-280
Abstract: & & Microplastic debris ending up at the sea surface has become a known major environmental issue. However, how microplastic particles move and when they sink in the ocean remains largely unknown. Here, we model microplastic subject to biofouling (algal growth on a substrate) to estimate sinking timescales and the time to reach the depth where particles stops sinking. We combine NEMO-MEDUSA 2.0 output, that represents hydrodynamic and biological properties of seawater, with a particle-tracking framework. Different sizes and densities of particles (for different types of plastic) are simulated, showing that the global distribution of sinking timescales is largely size-dependent as opposed to density-dependent. The smallest particles we simulate (0.1 & #956 m) start sinking almost immediately around the globe and their trajectories produce the longest time to reach their first sinking depth (almost 40 days as a global median). In oligotrophic subtropical gyres with low algal concentrations, particles between 1 mm and 10 & #956 m do not sink within the 90-day simulation time. This suggests that in addition to the comparatively well-known physical processes, biological processes might also contribute to the accumulation of floating plastic (of 1 mm to 10 & #956 m) in subtropical gyres. Particles of 1 & #956 m in the gyres start sinking largely due to vertical advection, whereas 0.1 & #956 m particles sink both due to biofouling and advection. The qualitative impacts of seasonality on sinking timescales are small, however, localised sooner sinking due to spring algal blooms is seen. This study maps processes that affect the sinking of virtual microplastic globally, which could ultimately impact the ocean plastic budget.& &
Publisher: Copernicus GmbH
Date: 17-06-2021
DOI: 10.5194/ISPRS-ANNALS-V-4-2021-217-2021
Abstract: Abstract. The Lagrangian analysis of particulate matter, biota and drifters, which are dispersed by turbulent fluid currents, is a cornerstone of oceanographic studies, covering erse study objectives. The results of Lagrangian simulations and observations is predominantly visualised by means of easy-access plotting interfaces and simple presentation techniques. We analysed over 50 publications from the years 2010–2020 with respect to their visual design to deduce common visualisation practices in the domain. In idual figures are analysed towards adherence to visualisation best-practices, algebraic visualisation guidelines and the IPCC visual style guide. In this article, we present the resulting best-practices and common pitfalls in the design of Lagrangian ocean visualisations. Based on this visual study, we highlight that raising awareness of established visual guidelines may have a higher impact on improving the visual quality of publications in oceanography than the vigorous development of more general-purpose visualisation tools.
Publisher: Copernicus GmbH
Date: 27-03-2022
DOI: 10.5194/EGUSPHERE-EGU22-5680
Abstract: & & Estimates of plastic quantities entering our oceans are not yet consistent with observed concentrations in the marine environment. This has led to the often-quoted statement that 99% of the marine plastics are missing. Here, we present a framework where the global transport of marine plastics is modelled over long time scales, in which the effects of different sources and sinks is investigated. Data assimilation techniques are used to inform unknown parameters regarding these sources and sinks, enabling us to quantify their role on the global plastic mass budget.& & & & State-of-the-art numerical models are included in the framework to capture for the first time the combined effect of marine plastic beaching, resuspension, biofouling, turbulent mixing, and fragmentation. The relative importance of different marine plastic sources is investigated, such as mismanaged coastal plastic waste, riverine outflow, and fishing activity. Unknown parameters are found by means of calibration to a large set of observational data of plastic concentrations in the ocean surface water, water column, ocean floor, and on coastlines.& & & & We show that with this framework, the global marine plastic mass budget can be closed. An overview is given of which environmental reservoirs are likely to contain most of the plastic mass, which sources are contributing to most of the pollution, and what the residence times of litter in the marine environment is. With the model calibration approach, we additionally get a better insight in the physics governing the transport of marine litter.& & &
Publisher: Copernicus GmbH
Date: 03-03-2021
DOI: 10.5194/EGUSPHERE-EGU21-1033
Abstract: & & Lagrangian simulations contribute to the study and comprehension of particulate-matter transport, its dissolution and dispersion in the oceans. Parcels is an open-source, Python-based module for Lagrangian ocean simulations. It is a known tool in the oceanographic community that has been applied to a variety of case studies, such as the tracing of microplastics, the backtracking of ocean floor plankton, and the migration of fish. In this module, particles are advected over time according to a selected flow field, where those particles can represent particulate-matter, biota or other objects with physical, hydrodynamic or biogeochemical properties. In this contribution, we present the substantial extensions of Parcels with respect to usability, physics modelling aspects of particle advection, and computational aspects of versatile, scalable and efficient simulations.& & & & Specifically, a suite of simple, concise notebook tutorials are tailored to novice user, covering step-by-step simulation setup instructions, whereas self-contained special-issue tutorials address advanced- and proficient user requirements. The considerable expansion of supported OGCM flow field input formats (e.g. MITgcm, POP and MOM5, among others) is a major interest in Parcels v2.2 for our steadily-growing user base.& & & & The new version further integrates previously-published physics methods into practical lagrangian particle simulations. As such, we implement an analytical advection scheme in addition to existing Runge-Kutta advection schemes. Furthermore, two-dimensional advection-diffusion is upgraded with the Milstein stochastic integration scheme and improved documentation. Those capabilities enable a more consistent modelling of diffusion- and uncertainty-dominated fluid transport processes.& & & & The case studies performed with previous versions indicate increased computational demands. Simulations are run over long decadal time scales as well as over day-periods with sub-second temporal increments, involving multiple basins and global scenarios, while also modelling increasingly complex particle processes. Overall, our developments respond to the big-data requirements of modern oceanographic studies, which include the aspects of (i) high record volume (i.e. large number of particles), (ii) high dimensionality in multi-variate records, (iii) high spatial resolution, (iv) high temporal resolution, (v) high scenario (i.e. case study) variability and (vi) the prevention of numerical error accumulation over long simulation time scales.& & & & The novel features of Parcels v2.2 are illustrated on distinct case studies within our contribution, in order to connect the technical features to their impact on particulate-matter ocean transport studies.& &
Publisher: Copernicus GmbH
Date: 19-01-2021
Abstract: Abstract. The detection of finite-time coherent particle sets in Lagrangian trajectory data, using data-clustering techniques, is an active research field at the moment. Yet, the clustering methods mostly employed so far have been based on graph partitioning, which assigns each trajectory to a cluster, i.e. there is no concept of noisy, incoherent trajectories. This is problematic for applications in the ocean, where many small, coherent eddies are present in a large, mostly noisy fluid flow. Here, for the first time in this context, we use the density-based clustering algorithm of OPTICS (ordering points to identify the clustering structure Ankerst et al., 1999) to detect finite-time coherent particle sets in Lagrangian trajectory data. Different from partition-based clustering methods, derived clustering results contain a concept of noise, such that not every trajectory needs to be part of a cluster. OPTICS also has a major advantage compared to the previously used density-based spatial clustering of applications with noise (DBSCAN) method, as it can detect clusters of varying density. The resulting clusters have an intrinsically hierarchical structure, which allows one to detect coherent trajectory sets at different spatial scales at once. We apply OPTICS directly to Lagrangian trajectory data in the Bickley jet model flow and successfully detect the expected vortices and the jet. The resulting clustering separates the vortices and the jet from background noise, with an imprint of the hierarchical clustering structure of coherent, small-scale vortices in a coherent, large-scale background flow. We then apply our method to a set of virtual trajectories released in the eastern South Atlantic Ocean in an eddying ocean model and successfully detect Agulhas rings. We illustrate the difference between our approach and partition-based k-means clustering using a 2D embedding of the trajectories derived from classical multidimensional scaling. We also show how OPTICS can be applied to the spectral embedding of a trajectory-based network to overcome the problems of k-means spectral clustering in detecting Agulhas rings.
Publisher: Copernicus GmbH
Date: 24-08-2021
DOI: 10.5194/ESD-2021-48
Abstract: Abstract. Having descended through the water column, microplankton in ocean sediments are representative for the ocean surface environment, where they originated from. Sedimentary microplankton is therefore used as an archive of past and present surface oceanographic conditions. However, these particles are advected by turbulent ocean currents during their sinking journey. So far, it is unknown to what extent this particle advection shapes the microplankton composition in sediments. Here we use global simulations of sinking particles in a strongly eddying global ocean model, and define ocean bottom provinces based on the particle surface origin locations. We find that these provinces can be detected in global datasets of sedimentary microplankton assemblages, demonstrating the effect provincialism has on the composition of sedimentary remains of surface plankton. These provinces explain the microplankton composition, together with e.g. ocean surface environment. Connected provinces have implications on the optimal spatial extent of microplankton sediment s le datasets that are used for palaeoceanographic reconstructions, and on the optimal spatial averaging of sediment s les over global datasets.
Publisher: Cold Spring Harbor Laboratory
Date: 22-09-2022
DOI: 10.1101/2022.09.20.508652
Abstract: Catch and distribution of tuna in the ocean are typically investigated with ocean basin-scale models. Due to their large scale, such models must greatly simplify tuna behaviour occurring at a scale below ∼100 km, despite interactions at this level potentially being important to both catch and distribution of tuna. For ex le, the associative behaviour of tuna with man-made floating objects, that are deployed by fishers to improve their catch rates (Fish Aggregating Devices FADs), are usually ignored or simplified. Here we present a model that can be used to investigate the influence of tuna dynamics below the ∼100 km scale on larger scales. It is an Agent-Based Model (ABM) of a hypothetical, tuna-like species, that includes their interactions with each other, free-floating FADs and prey. In this ABM, both tuna and FADs are represented by Lagrangian particles that are advected by an ocean flow field, with tuna also exhibiting active swimming based on internal states such as stomach fullness. We apply the ABM in multiple configurations of idealised flow and prey fields, alongside differing interaction strengths between agents. When tuna swimming behaviour is influenced equally by prey and FADs, we find that the model simulations compare well with observations at the ≲ 100 km scale. For instance, compared to observations, tuna particles have a similar stomach fullness when associated or non-associated to a FAD, tuna colonize at similar timescales at FADs after their deployment and tuna particles exhibit similar variations in continuous residence times. However, we find large differences in emergent dynamics such as residence and catch among different flow configurations, because the flow determines the time scale at which tuna encounter FADs. These findings are discussed in the context of directing future research, and an improved interpretation of tuna catch and other data for the sustainable management of these economically important species.
Location: Germany
No related grants have been discovered for Christian Kehl.