EXCHANGE CHARACTERISTICS OF AN ANTHROPOGENICALLY MODIFIED LAGOON: AN EULERIAN-LAGRANGIAN MODELING CASE STUDY WITH AN EMPHASIS ON THE NUMBER OF PARTICLES

The transport pathways and exchange characteristics of the Kamil Abdüş Lagoon in Istanbul, Turkey, are simulated using a finite element model with a Lagrangian particle tracking module. The lagoon is in the process of being reconfigured. The simulations are performed using a draft configuration. The effect of winds and the number of particles on the e-folding time is simulated. Results show that the lagoon is strongly dominated by winds with a correlation coefficient of 0.897 between the wind and residual current magnitudes. The lagoon e-folds in 9.1 days under realistic winds and in 14.3 days when there is no wind with confidence levels of 5%. The Lagrangian study uses six simulations with particle numbers ranging between 65073 and 2730486. A methodology based on confidence levels is proposed. It is observed that approximately 784 000 particles are necessary to obtain 5% level of confidence. With a problematic history and new planning options, the lagoon has a potential to be used as an example setting, all-field study ground for anthropogenically engineered coastal ecosystems.

A Lagrangian study of interfaces at the edges of cumulus clouds

Interfaces at the edge of an idealised, non-precipitating, warm cloud are studied using Direct Numerical Simulation (DNS) complemented with a Lagrangian particle tracking routine. Once a shell has formed, four zones can be distinguished: the cloud core, visible shell, invisible shell and the environment. The union of the visible and invisible regions is the shell commonly referred to in literature. The boundary between the invisible shell and the environment is the Turbulent-NonTurbulent Interface (TNTI) which is typically not considered in cloud studies. Three million particles were seeded homogeneously across the domain and properties were recorded along individual trajectories. The results demonstrate that the traditional cloud boundary (separating cloudy and non-cloudy regions using thresholds applied on liquid condensate or updraft velocity) are some distance away from the TNTI. Furthermore, there is no dynamic difference between the traditional liquid-condensate boundary and the region extending to the TNTI. However, particles crossing the TNTI exhibit a sharp jump in enstrophy and a smooth increase in buoyancy. The traditional cloud boundary coincides with the location of minimum buoyancy in the shell. The shell pre-mixes the entraining and detraining air and analysis reveals a highly skewed picture of entrainment and detrainment at the traditional cloud boundary. A preferential entrainment of particles with velocity and specific humidity higher than the mean values in the shell is observed. Large-eddy simulation of a more realistic setup detects an interface with similar properties using the same thresholds as in the DNS, indicating that the DNS results extrapolate beyond their idealised conditions.

Relative dispersion in the Nordic Seas - new insights ten years later

<p>The POLEWARD experiment in the Nordic Seas (2007-2009) involved deployment of 150 drifters in the eastern Nordic Seas and has been the first large drifter pair experiment in the ocean (and one of the very few conducted up to date). The experiment yielded nearly 100 drifter pairs with initial separations 2km or less, which allowed us to elucidate several aspects of the relative dispersion (a proxy for tracer spreading and transport) at a basin scale, to quantify the role of mesoscale eddies in surface transport, and to further develop the relevant theoretical and analytical methods through a series of publications. Ten years ago however there were no modeling tools available to carry out a similar numerical Lagrangian study in this region resolving relevant scales of variability.</p><p>In this presentation, we will present an update on the relative dispersion of surface drifter pairs in the Nordic Seas, with over 400 pairs available. We will then compare the observed statistics to these derived from Lagrangian simulations (OpenDrift scheme) forced by output from a very high resolution regional ocean model (Regional Ocean Modeling System). The comparison is very favorable pointing to the ability of the ocean model to represent surface eddy stirring processes. We will also show analysis of the regional dispersion regimes using both drifter observations and model simulations, and consider the effect of including vertical motion in the Lagrangian simulations, which impacts their horizontal dispersion. We will also present statistics of the temperature differences on drifters pairs. These are underestimated by the model on daily time scales and deformation scales, which has implications for the model ability to simulate tracer processes on these scales. </p><p> </p>